An optimization control method, electronic device and storage medium for community charging piles

Abstract

A method for optimizing the control of community charging piles, an electronic device, and a storage medium, belonging to the field of power demand response. To achieve the optimized control of community charging piles, the present invention calculates the total grid load of the community; collects historical data on the electricity consumption of community infrastructure and equipment on a seasonal basis to establish a power load model for community infrastructure; collects historical data on the daily electricity consumption of residents, including the electricity consumption of each hour, as well as weather and whether it is a holiday on a daily basis to establish a power load model for the daily life of community residents; establishes a power load model for available charging piles; based on the number of electric vehicles connected to the charging piles collected in real time, according to the corresponding time, combines the power load of community infrastructure and the power load of residents to establish a charging scheduling prediction model for charging piles; designs a priority iterative optimization algorithm to optimize the scheduling of the charging scheduling prediction model for charging piles, dynamically adjusts the number of connected charging piles, and realizes that the power load is always within a reasonable range.

Description

Technical Field

[0001] The present invention belongs to the field of power demand response, and particularly relates to an optimization control method for community charging piles, an electronic device, and a storage medium. Background Art

[0002] In recent years, with the intensification of global warming, shortage of oil resources, and energy security issues, in order to achieve the goal of global low-carbon development, the vehicle types are undergoing rapid changes. The continuous progress of vehicle power battery technology, especially new energy vehicle models represented by pure electric vehicles, is becoming the main force driving this change. With the rise of electric vehicles, their accessory charging piles are also spread all over the world. There are not only the construction of large charging stations but also the installation of private charging piles. Nevertheless, the growth rate of charging piles has slowed down compared with the sales growth rate of new energy vehicles, and the "range anxiety" and "endurance pain points" of new energy vehicle owners have also been amplified. The existing charging piles cannot fully meet the needs of new energy vehicle owners, and this problem has not been effectively solved. The reason is that the load of the community power grid is limited, while the power consumption load of the charging piles is very large. Therefore, a large number of charging piles cannot be installed, forcing vehicle owners to go elsewhere to charge.

[0003] Generally speaking, if too many charging piles are installed in a community, when multiple charging piles charge electric vehicles simultaneously, the power consumption load will show a sharp increase. When the power consumption load exceeds the power grid load, there will be potential safety hazards such as tripping and even fires.

[0004] Common machine scheduling methods include: (1) First Come First Served, which assigns tasks to charging piles in the order of arrival. It is simple and easy to implement, but may result in a relatively long average waiting time; (2) Shortest Job First, which selects the charging pile with the shortest charging time to execute first, and can minimize the average waiting time, but may cause the waiting time of long tasks to be too long. (3) Priority Scheduling, which assigns a priority to each task and schedules according to the priority level, but may lead to long waiting times for low-priority tasks and there is a problem of priority inversion; (4) Round Robin Scheduling, which assigns tasks to machines according to time slices. Each task executes for one time slice and then switches to the next task, but may cause frequent context switches, increase system overhead, and the response time is unstable, which is not applicable to real-time systems. Summary of the Invention

[0005] The problem to be solved by the present invention is to achieve the optimization control of community charging piles, and a method for optimizing the control of community charging piles, an electronic device, and a storage medium are proposed.

[0006] To achieve the above object, the present invention is realized through the following technical solutions:

[0007] An optimization control method for community charging piles includes the following steps:

[0008] S1. Calculate the total power grid load of the community: Collect the rated capacity, rated voltage, and transformation ratio data of all transformers in the community, and then accumulate the actual loads of each transformer to obtain the total power grid load of the community, and set the total power grid load of the community as K;

[0009] S2. Taking seasons as units, collect the historical data of the electricity consumption of the community's infrastructure and equipment, and establish an electricity load model for the community's infrastructure;

[0010] S3. Taking days as units, collect the historical data of the electricity consumption of residents on a single day including each hour, as well as the weather and whether it is a holiday, and establish a prediction model for the daily electricity load of community residents;

[0011] S4. Based on the total power grid load of the community obtained in step S1, the electricity load model of the community's infrastructure obtained in step S2, and the prediction model for the daily electricity load of community residents obtained in step S3, establish an available electricity load model for charging piles;

[0012] S5. Based on the number of electric vehicles connected to the charging piles collected in real time, according to the corresponding time, combined with the electricity load of the community's infrastructure and the electricity load of residents, establish a prediction model for the charging scheduling of charging piles;

[0013] S6. Design a priority iterative optimization algorithm to optimize the charging scheduling prediction model obtained in step S6, and dynamically adjust the number of charging piles connected according to real-time data and prediction results to ensure that the electricity load is always within a reasonable range.

[0014] Furthermore, the specific implementation method of step S2 includes the following steps:

[0015] S2.1. Collect the electricity consumption of various infrastructure and equipment in the community, including the electricity consumption data of street lights, elevators, and water and power supply systems. At hourly intervals, including the power consumption and operating duration of each equipment, clean the collected data to remove outliers;

[0016] S2.2. Classify and statistically analyze the electricity consumption data after data cleaning in step S2.1 seasonally. Use the pandas library in python to classify the electricity consumption of equipment seasonally, so as to establish a hierarchical structure for the electricity load of the community's infrastructure, and divide each equivalent part into different levels according to its load characteristics;

[0017] S2.3. Establish a single electricity load model for the community's infrastructure, and the expression is:

[0018] Y1 = w1x1 + w2x2 + w3x3 + b

[0019] Among them, x1 is the current season, w1 is the current season eigenvalue, x2 is the current date, w2 is the current date eigenvalue, x3 is the current time point, w3 is the current time point eigenvalue, Y1 is the output power consumption load, and b is the bias;

[0020] When there are multiple samples, the total power consumption load of the community infrastructure is obtained, and the expression is:

[0021] Y = Xw

[0022] Among them, Y is the total power consumption load of the community infrastructure, X is the attribute matrix, and w is the eigenvalue matrix;

[0023] The mean square error loss function is used to calculate the loss mean of multiple samples to obtain the bias b to minimize the total loss of the training samples. The mean square error loss function l (i) (w, b) The expression is:

[0024]

[0025] Among them, is the predicted value of the output power consumption load of the corresponding infrastructure i, and y (i) is the true value of the output power consumption load of the corresponding infrastructure i.

[0026] Furthermore, the specific implementation method of step S3 includes the following steps:

[0027] S3.1. Collect the daily electricity consumption of community residents in a year, and collect the data of the weather and whether it is a holiday on that day. Fit the daily electricity consumption load of community residents with the time, weather, and whether it is a holiday on that day to obtain the prediction model of the daily electricity consumption load of community residents. The expression for the current time electricity consumption load H of community residents' daily life is:

[0028] H = r1k1 + r2k2 + r3k3 + c

[0029] Among them, k1 is the time on that day, r1 is the eigenvalue of the time on that day, k2 is the weather, r2 is the eigenvalue of the weather, k3 is whether it is a holiday, r3 is the eigenvalue of whether it is a holiday, and c is the bias of the prediction model of the daily electricity consumption load of community residents;

[0030] S3.2. Set to identify and judge outliers by judging whether the power consumption load of the i-th sample deviates from 2 times the standard deviation. The expression is:

[0031]

[0032] Among them, H i is the power consumption load of the i-th sample, μ is the mean of n samples, and σ is the standard deviation of the samples;

[0033] When the following conditions are met, d is considered i an outlier, and the expression is:

[0034] |H i - μ| > 2 * σ

[0035] Set the method of using the median to fill the missing values of the samples, and the expression is:

[0036] H filled = median(H1, H2, H3, …, H n )

[0037] where median(·) represents the median of the data;

[0038] S3.3. Combine the current electricity load R of the community residents observed by the smart meter now to perform modeling, and the expression is:

[0039] X T Xr' = X T R now

[0040] where X is the characteristic equation, and each row represents the attribute characteristics of a sample; r' is the regression coefficient vector; R now is the current electricity load of the community residents observed at the current moment;

[0041] Use the root mean square error between different characteristics and the electricity load in the electricity load prediction model of community residents' daily life as the first function, and through optimization algorithm fitting, make the minimum value of the first function the optimal goal to obtain the optimal characteristic parameters. The expression of the root mean square error MSE is:

[0042]

[0043] where n is the number of samples, R nowi is the current electricity load of the community residents observed at the current moment for the i-th sample, is the predicted daily electricity load of the community residents for the i-th sample; the obtained r' is the optimal parameter of the electricity load prediction model of community residents' daily life.

[0044] Furthermore, the specific implementation method of step S4 includes the following steps:

[0045] S4.1. Calculate the total available electricity load T(t), and the expression is:

[0046] T(t) = K - H - Y;

[0047] S4.2. Based on the load rate of the transformer usually being controlled between 75% and 85% of its rated capacity, the total electricity load is set to be controlled at 85% of the rated capacity, and a model of the available electricity load of the charging pile is established, with the expression being

[0048] T'(t) = K * 0.85 - H - Y

[0049] Among them, T'(t) is the available electricity load of the charging pile.

[0050] Furthermore, the specific implementation method of step S5 includes the following steps:

[0051] S5.1. Set the relevant parameters for establishing the charging scheduling prediction model of the charging pile, including the available electricity load T'(t) of the charging pile, the number N of electric vehicles connected to the charging pile, the initial state of charge B of the i-th electric vehicle i , the battery capacity C of the i-th electric vehicle i , the rated charging power P of the i-th electric vehicle i , the charging efficiency η of the electric vehicle i ;

[0052] S5.2. Calculate the charging duration t of the i-th single electric vehicle i , and the expression is:

[0053]

[0054] S5.3. Real-time monitor the geographical location, usage situation, the number of connected electric vehicles, and the battery power of the electric vehicles of each charging pile; make a point-to-point division for each household to obtain the usage load data of the community charging piles in units of households;

[0055] S5.4. Combine the infrastructure electricity load model and the residential daily electricity load prediction model to establish a charging scheduling prediction model of the charging pile, with the expression being:

[0056] T′(t) ≥ P f1 + P f2 + … + P ff + P j1 + P j2 + … + P jj

[0057] Among them, P is the usage load data of the community charging piles in units of households extracted from each partition. For two partitions, f samples and j samples are respectively extracted according to the sorting.

[0058] Furthermore, the specific implementation method of step S6 includes the following steps:

[0059] S6.1. Set an index for local restriction. Taking one day as an example, set the vehicle charging completion ratio of the first partition's ordinary area (0,1) and premium area (0,2) among the vehicles that have completed charging as a:b, where a < b;

[0060] S6.2. Based on the charging pile charging scheduling prediction model constructed in step S5, perform calculations to obtain the expression:

[0061]

[0062] Among them, A is the theoretical value of the number of vehicles that have completed charging in 24 hours; a is the number of vehicles that have completed charging in the ordinary area of the first partition, and b is the number of vehicles that have completed charging in the premium area;

[0063] a ≤ f

[0064] b ≤ j;

[0065] S6.3. Set U(n,m) as the value of the partition threshold. U(n,m) is taken from one of the electric vehicle charging durations, that is, one of the values in t i Calculate the value of A one by one for all possible U(n,m) values by means of exhaustive enumeration and sort them. The maximum value obtained by solving is the optimal value;

[0066] Based on the selection of U(n,m) and the number of iterations, perform scheduling processing, so as to ensure that the grid load always meets the charging needs of each electric vehicle under the restricted load.

[0067] An electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the steps of the above-mentioned community charging pile optimization control method are implemented.

[0068] A computer-readable storage medium stores a computer program thereon. When the computer program is executed by a processor, the above-mentioned community charging pile optimization control method is implemented.

[0069] Advantages of the present invention:

[0070] An optimization control method for community charging piles according to the present invention describes an optimization control algorithm for community charging piles on the premise that each household in the community has an electric vehicle and a charging pile is installed. It can meet the charging needs of electric vehicles without changing the original power grid load. The method includes analyzing the power structure of the community; analyzing the comprehensive electricity consumption of the community, that is, daily life electricity consumption, infrastructure electricity consumption and charging pile electricity consumption; estimating and predicting the daily life electricity consumption of residents; estimating and predicting the infrastructure electricity consumption of the community; collecting the usage conditions of all charging piles in the community and the battery levels of electric vehicles in real time, and combining the predicted daily life electricity consumption and infrastructure electricity consumption of the community, and rationally scheduling the charging piles through an optimization control algorithm to achieve charging of electric vehicles.

[0071] An optimization control method for community charging piles according to the present invention is to collect the usage conditions of each charging pile and the battery levels of the electric vehicles connected to the charging piles in real time, combine the predicted daily life electricity consumption and infrastructure electricity consumption of the community, and adjust the number of charging piles connected in real time, so that the electricity load of the community is always below the power grid load. Under the original power grid load, through the scheduling of the optimization control algorithm, the charging capacity of the charging piles is maximally exerted.

[0072] An optimization control method for community charging piles according to the present invention can adjust the electricity load of the charging piles specifically through predicting the infrastructure electricity load and the daily life electricity load of community residents; by analyzing the battery levels of the connected electric vehicles, it can predict the charging time required, and connect them to the community charging pile system in the form of points. After analysis and calculation by the optimization control model, the allocated charging time period can be obtained to meet the charging needs of each electric vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 is a flowchart of an optimization control method for community charging piles according to the present invention;

[0074] Figure 2 is an infrastructure electricity load model;

[0075] Figure 3 is a predicted model for daily life electricity load of residents;

[0076] Figure 4 is an optimization prediction model for charging piles. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0077] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention, that is, the specific embodiments described are only a part of the embodiments of the present invention, rather than all of the specific embodiments. The components of the specific embodiments of the present invention usually described and shown in the drawings herein can be arranged and designed in various different configurations, and the present invention can also have other embodiments.

[0078] Therefore, the following detailed description of the specific embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents the selected specific embodiments of the present invention. Based on the specific embodiments of the present invention, all other specific embodiments obtained by those skilled in the art without creative efforts fall within the scope of protection of the present invention.

[0079] To further understand the content, features and effects of the present invention, the following specific embodiments are exemplified and are accompanied by Figure 1 -Accompanying Figure 4 The detailed description is as follows:

[0080] Embodiment 1:

[0081] An optimized control method for community charging piles includes the following steps:

[0082] S1. Calculate the total grid load of the community: Collect the rated capacity, rated voltage, and transformation ratio data of all transformers in the community, and then accumulate the actual loads of each transformer to obtain the total grid load of the community, and set the total grid load of the community as K;

[0083] S2. Collect historical data on the electricity consumption of community infrastructure and equipment on a seasonal basis, and establish a community infrastructure electricity load model;

[0084] Further, the specific implementation method of step S2 includes the following steps:

[0085] S2.1. Collect the electricity consumption of various infrastructure and equipment in the community, including the electricity consumption data of street lights, elevators, and water supply and power supply systems. At hourly intervals, including the power consumption and operating duration of each device, clean the collected data to remove outliers;

[0086] S2.2. Classify and statistically analyze the electricity consumption data after data cleaning in step S2.1 seasonally. Use the pandas library in python to classify the seasonal electricity consumption of equipment, thereby establishing a hierarchical structure of the community infrastructure electricity load, and dividing each equivalent part into different levels according to its load characteristics;

[0087] Build an equivalent model for grading the electricity load of the community infrastructure. The grading equivalent model includes parameters for different seasons; group devices or facilities with similar electricity consumption patterns into an equivalent part, that is, equivalent multiple devices into a whole, representing the electricity load of this type of device within this season.

[0088] S2.3. Build an electricity load model for a single community infrastructure, and the expression is:

[0089] Y1 = w1x1 + w2x2 + w3x3 + b

[0090] Among them, x1 is the current season, w1 is the eigenvalue of the current season, x2 is the current date, w2 is the eigenvalue of the current date, x3 is the current time point, w3 is the eigenvalue of the current time point, Y1 is the output electricity load, and b is the bias;

[0091] For example, if there are m samples now, and each sample has 3 eigenvalues, which are as mentioned above; then substituting all sample points into the equation gives:

[0092] Y = w1x 1,1 + w2x 1,2 + w3x 1,3 + b

[0093] Y = w1x 2,1 + w2x 2,2 + w3x 2,3 + b

[0094] …

[0095] Y = w1x m,1 + w2x m,2 + w3x m,3 + b

[0096] When the number of samples is multiple, the total electricity load of the community infrastructure is obtained, and the expression is:

[0097] Y = Xw

[0098] Among them, Y is the total electricity load of the community infrastructure, X is the attribute matrix, and w is the eigenvalue matrix;

[0099] Use the mean square error loss function to calculate the mean loss of multiple samples to obtain the bias b, in order to minimize the total loss of the training samples. The expression of the mean square error loss function l (i) (w, b) is:

[0100]

[0101] Among them, is the predicted value of the output electricity load corresponding to infrastructure i, y (i)is the true value of the output power consumption load corresponding to the infrastructure i.

[0102] Use the above sub - prediction models for fitting, and obtain a hierarchical equivalent prediction model through classification; and use historical data for model evaluation to ensure the accuracy and reliability of the model.

[0103] S3. Taking days as units, collect the historical data of the daily electricity consumption of residents including each hour, as well as the weather and whether it is a holiday, and establish a prediction model for the daily electricity consumption load of community residents;

[0104] Furthermore, the specific implementation method of step S3 includes the following steps:

[0105] S3.1. Collect the daily electricity consumption of community residents in a year, and collect the data of the weather and whether it is a holiday on the same day. Fit the daily electricity consumption load of community residents with the current time, weather, and whether it is a holiday on the same day to obtain a prediction model for the daily electricity consumption load of community residents. The expression for the current - time electricity consumption load H of community residents' daily life is:

[0106] H = r1k1 + r2k2 + r3k3 + c

[0107] Among them, k1 is the current time of the day, r1 is the eigenvalue of the current time of the day, k2 is the weather, r2 is the eigenvalue of the weather, k3 is whether it is a holiday, r3 is the eigenvalue of whether it is a holiday, and c is the bias of the prediction model for the daily electricity consumption load of community residents;

[0108] Establish the above - mentioned relationship model, fit and analyze the weather data, whether it is a holiday data and the current electricity consumption load to improve the correlation. Here, a linear regression model similar to requirement 2 is still used for predicting model construction; perform feature engineering processing on the collected data, including outlier processing, missing - value processing, and feature extraction to ensure the quality and usability of the data;

[0109] S3.2. Set the identification and judgment of outliers by judging whether the electricity consumption load of the i - th sample deviates from 2 times the standard deviation. The expression is:

[0110]

[0111] Among them, H i is the electricity consumption load of the i - th sample, μ is the mean of n samples, and σ is the standard deviation of the samples;

[0112] When the following conditions are met, then H i is considered an outlier. The expression is:

[0113] |H i - μ|>2 * σ

[0114] Set the method of using median filling to handle the missing values of the samples, and the expression is:

[0115] H filled = median(H1, H2, H3, …, H n )

[0116] where median(·) represents the median of the data;

[0117] S3.3. Combine the current electricity load R of the community residents observed by the smart meter now to perform modeling, and the expression is:

[0118] X T Xr' = X T R now

[0119] where X is the feature equation, and each row represents the attribute features of a sample; r' is the regression coefficient vector; R now is the current electricity load of the community residents observed at the current moment;

[0120] Use the root mean square error between different features and the electricity load in the daily electricity load prediction model of the community residents as the first function, and through optimization algorithm fitting, make the minimum value of the first function the optimal goal to obtain the optimal feature parameters. The expression of the root mean square error MSE is:

[0121]

[0122] where n is the number of samples, R nowi is the current electricity load of the community residents observed at the current moment for the i-th sample, is the predicted daily electricity load of the community residents for the i-th sample; the obtained r' is the optimal parameter of the daily electricity load prediction model of the community residents.

[0123] Use the above relationship model, combine the electricity load situation Y now measured by the smart meter in real time, establish a daily electricity load prediction model for the community residents, and use historical data for model evaluation to ensure the accuracy and reliability of the model.

[0124] S4. Based on the total grid load of the community obtained in step S1, the electricity load model of the community infrastructure obtained in step S2, and the daily electricity load prediction model of the community residents obtained in step S3, establish a model for the available electricity load of the charging piles;

[0125] Furthermore, the specific implementation method of step S4 includes the following steps:

[0126] S4.1. Calculate the total available electricity load T(t), and the expression is:

[0127] T(t) = K - H - Y;

[0128] At this time, it cannot be said that the above is the available power load of the charging pile. Because in general, the load rate of the transformer is usually controlled between 75% and 85% of its rated capacity, which is more appropriate. The load rate within this range can ensure that the transformer has a certain redundant capacity during long-term operation, improving its stability and reliability. At the same time, the load rate within this range can also ensure that the transformer operates normally within its rated capacity, extending its service life. So, if the total power load is set to be controlled at 85% of the rated capacity, then the available power load of the charging pile can be obtained as follows:

[0129] S4.2. Based on the fact that the load rate of the transformer is usually controlled between 75% and 85% of its rated capacity, so the total power load is set to be controlled at 85% of the rated capacity, and a model for the available power load of the charging pile is established, with the expression

[0130] T'(t) = K * 0.85 - H - Y

[0131] where T'(t) is the available power load of the charging pile.

[0132] S5. Based on the number of electric vehicles connected to the charging pile collected in real time, combined with the corresponding time, the infrastructure power load of the community and the residential power load, a prediction model for the charging scheduling of the charging pile is established;

[0133] Furthermore, the geographical locations, usage conditions, the number of electric vehicles connected, and the battery levels of the electric vehicles of each charging pile are monitored in real time; the above data are divided point-to-point for each household; thus, the load data P of the community charging piles per household is obtained i ; using the above data, combined with the infrastructure power load model and the residential daily power load prediction model, the priority iterative optimization algorithm is adopted, and historical data is used for model evaluation to ensure the accuracy and reliability of the model.

[0134] First, clarify the goal, that is, to ensure that the grid load is always below the limit load. Second, consider the charging efficiency, that is, to meet the charging needs of each electric vehicle as much as possible.

[0135] Then, first consider dividing the priority according to the charging time required by the battery. The higher the charging time required by the battery, the higher the priority, and the lower the corresponding charging pile node priority for the battery with less charging time required. However, charging cannot be carried out only based on the priority. As mentioned in the above background technology: If only the priority is used for evaluation, then there may be a situation where the task with a lower priority has too long a waiting time, that is, the charging pile fails to charge in time, and even the vehicle may not be charged when the owner picks up the car. Therefore, the charging piles are dynamically divided according to the charging time required by the electric vehicle battery, that is, divided into a general area and a premium area. The general area still charges according to the priority, while the premium area is dedicated to serving tasks with a lower priority, that is, the charging piles with a shorter charging time required by the electric vehicle battery.

[0136] With the above introduction, next, the charging piles can be sorted from more to less according to the charging time required by the electric vehicle and assigned priorities from low to high. Furthermore, the electric vehicles with a charging time less than U(0,0) are divided into the premium area (0,2), and correspondingly, the electric vehicles with a charging time greater than U(0,0) are divided into the general area (0,1). Further, the general area and the premium area can be further divided into their general area sub-areas and premium area sub-areas, and the division is continuously carried out until one electric vehicle occupies one sub-area. The purpose of doing this is to avoid the phenomenon of local optimal solutions during iteration. In this way, by comparing the optimal values calculated for each part, the overall optimal value is taken to determine the optimal solution. Thus, the partition result and specific data can be obtained.

[0137] Charging is carried out based on the above data. At the same time, it is detected whether the grid load exceeds the set threshold or the load fluctuation is greater than 15% and whether there is a new electric vehicle connected: (1) Whether the grid load exceeds the set threshold. If not, the detection is looped; if it exceeds or the fluctuation is too large, it jumps back to the partition step to re-partition and schedule the access situation of electric vehicles. (2) Whether there is a new electric vehicle connected. If the number of connected vehicles is the same as or less than the original number, the number of connected vehicles is repeatedly detected; if it is more than the original number, it is judged whether there is a charging vehicle reaching the partition threshold or being fully charged. If not, it enters the detection of the number of connected vehicles and waits while detecting; if so, it jumps back to the start step for calculation.

[0138] Furthermore, the specific implementation method of step S5 includes the following steps:

[0139] S5.1. Set the relevant parameters for establishing the charging scheduling prediction model of the charging pile, including the available power load T'(t) of the charging pile, the number N of electric vehicles connected to the charging pile, the initial state of charge B of the i-th electric vehicle i , the battery capacity C of the i-th electric vehicle i , the rated charging power P of the i-th electric vehicle i, charging efficiency η of electric vehicle i ;

[0140] S5.2. Calculate the charging duration t of the i-th single electric vehicle i , and the expression is:

[0141]

[0142] S5.3. Real-time monitor the geographical location, usage status, number of connected electric vehicles, and battery power of each charging pile; conduct point-to-point division for each household to obtain the charging pile usage load data of the community in units of households;

[0143] S5.4. Combine the infrastructure power consumption load model and the residential daily power consumption load prediction model to establish a charging pile charging scheduling prediction model, and the expression is:

[0144] T′(t)≥P f1 +P f2 +…+P ff +P j1 +P j2 +…+P jj

[0145] where P is the charging pile usage load data of the community in units of households extracted from each partition. For two partitions, f samples and j samples are extracted respectively according to the sorting.

[0146] S6. Design a priority iterative optimization algorithm to optimize the charging scheduling prediction model of the charging pile obtained in step S6. According to the real-time data and prediction results, dynamically adjust the number of connected charging piles to keep the power consumption load within a reasonable range at all times.

[0147] The calculation of the optimal value mentioned above is the optimal value of each part of the partition. The previous goal is to ensure that the grid load is always below the limit load and to meet the charging needs of each electric vehicle as much as possible. If only judged by the number of charged vehicles at this time, the reverse problem of the previous priority will inevitably occur, that is, the task with lower priority is executed first, which is contrary to the priority algorithm, so this indicator cannot be used to judge. Then an indicator needs to be set at this time for local restriction;

[0148] Furthermore, the specific implementation method of step S6 includes the following steps:

[0149] S6.1. Set an indicator for local restriction. Taking one day as an example, set the charging completion ratio of the vehicles involved in the ordinary area (0,1) and the high-class area (0,2) of the first partition among the charged vehicles as a:b, where a < b;

[0150] S6.2. Based on the charging scheduling prediction model constructed in step S5, calculations are performed to obtain the expression:

[0151]

[0152] Among them, A is the theoretical value of the number of vehicles with completed charging in 24 hours; a is the number of vehicles with completed charging in the ordinary area of the first partition, and b is the number of vehicles with completed charging in the premium area;

[0153] a ≤ f

[0154] b ≤ j;

[0155] S6.3. Set U(n,m) as the value of the partition threshold. U(n,m) is taken from one of the electric vehicle charging durations, that is, one of the values in t i For all possible values of U(n,m), the value of A is calculated one by one through an exhaustive method and sorted. The minimum value obtained by solving is the optimal value;

[0156] Based on the selection of U(n,m) and the number of iterations, scheduling processing is performed, so as to ensure that the grid load always meets the charging requirements of each electric vehicle under the limit load.

[0157] Embodiment 2:

[0158] An electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the method for optimizing the control of community charging piles described in Embodiment 1 are implemented.

[0159] The computer device of the present invention may be a device including a processor and a memory, such as a single-chip microcomputer including a central processing unit. Moreover, when the processor is used to execute the computer program stored in the memory, the steps of the above-mentioned method for optimizing the control of community charging piles are implemented.

[0160] The so-called processor may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0161] The memory may mainly include a program storage area and a data storage area. Among them, the program storage area may store an operating system, application programs required for at least one function (such as a sound playback function, an image playback function, etc.); the data storage area may store data created according to the use of the mobile phone (such as audio data, phone book, etc.). In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0162] Embodiment 3:

[0163] A computer-readable storage medium stores a computer program thereon, and when the computer program is executed by a processor, it implements the method for optimizing the control of a community charging pile described in Embodiment 1.

[0164] The computer-readable storage medium of the present invention may be any form of storage medium readable by the processor of a computer device, including but not limited to non-volatile memory, volatile memory, ferroelectric memory, etc. A computer program is stored on the computer-readable storage medium. When the processor of the computer device reads and executes the computer program stored in the memory, the steps of the above method for optimizing the control of a community charging pile can be implemented.

[0165] The computer program includes computer program code, and the computer program code may be in the form of source code, object code, an executable file, or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disc, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc. It should be noted that the content included in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0166] It should be noted that relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the phrase "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0167] Although the present application has been described above with reference to specific embodiments, various improvements can be made thereto and components thereof can be replaced with equivalents without departing from the scope of the present application. In particular, as long as there is no structural conflict, the various features in the specific embodiments disclosed in the present application can be combined with each other in any way, and the reason for not exhaustively describing the situations of these combinations in this specification is only to save space and resources. Therefore, the present application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. An optimization control method for community charging piles, characterized in that, It includes the following steps: S1. Calculate the total grid load of the community: Collect the rated capacity, rated voltage, and transformation ratio data of all transformers in the community, and then accumulate the actual loads of each transformer to obtain the total grid load of the community, and set the total grid load of the community as K; S2. Take seasons as units, collect historical data on the electricity consumption of the community's infrastructure and equipment, and establish an electricity load model for the community's infrastructure; S3. Take days as units, collect historical data on the electricity consumption of residents per day including each hour, as well as weather and whether it is a holiday, and establish a prediction model for the daily electricity load of community residents; S4. Based on the total grid load of the community obtained in step S1, the electricity load model of the community's infrastructure obtained in step S2, and the prediction model for the daily electricity load of community residents obtained in step S3, establish an available electricity load model for charging piles; S5. Based on the number of electric vehicles connected to the charging piles collected in real time, according to the corresponding time, combined with the electricity load of the community's infrastructure and the electricity load of residents, establish a prediction model for the charging scheduling of charging piles; The specific implementation method of step S5 includes the following steps: S5.

1. Set relevant parameters for establishing a charging scheduling prediction model for charging piles, including the available power load T′(t) of the charging pile, the number N of electric vehicles connected to the charging pile, the initial state of charge B of the i-th electric vehicle i , the battery capacity C of the i-th electric vehicle i , the rated charging power P of the i-th electric vehicle i , the charging efficiency η of the electric vehicle i ; S5.

2. Calculate the charging duration t of the i-th single electric vehicle i , and the expression is: S5.

3. Real-time monitor the geographical location, usage status, number of electric vehicles connected, and battery power of each charging pile; make point-to-point division for each household to obtain the charging pile usage load data of the community in units of households; S5.

4. Combine the electricity load model of the infrastructure with the prediction model for the daily electricity load of community residents to establish a prediction model for the charging scheduling of charging piles, and the expression is: T′(t)≥P f1 +P f2 +…+P ff +P j1 +P j2 +…+P jj where P is the charging pile usage load data of the community in units of households extracted from each partition. For two partitions, f samples and j samples are extracted respectively according to the sorting; S6. Design a priority iterative optimization algorithm to optimize the scheduling of the charging scheduling prediction model obtained in step S5, and dynamically adjust the number of charging piles connected according to real-time data and prediction results to keep the electricity load within a reasonable range at all times; The specific implementation method of step S6 includes the following steps: S6.

1. Set an index for local limitation. Taking one day as an example, set the charging completion ratio of the vehicles involved in the ordinary area (0,1) and the high-class area (0,2) of the first partition among the vehicles that have completed charging as a:b, where a < b; S6.

2. Based on the charging scheduling prediction model constructed in step S5, perform calculations to obtain the expression: where A is the theoretical value of the number of vehicles that have completed charging in 24 hours; a is the number of vehicles that have completed charging in the ordinary area of the first partition, and b is the number of vehicles that have completed charging in the high-class area; a ≤ f b ≤ j; S6.

3. Set U(n,m) as the value of the partition threshold. U(n,m) is taken from one of the electric vehicle charging durations, i.e., t i Among the values in, calculate the A value one by one for all possible U(n,m) values by means of exhaustive enumeration and sort them. The maximum value obtained by solving is the optimal value; Based on the selection of U(n,m) and the number of iterations, perform scheduling processing to ensure that the grid load always meets the charging requirements of each electric vehicle under the limited load.

2. The optimized control method for a community charging pile according to claim 1, wherein, The specific implementation method of step S2 includes the following steps: S2.

1. Collect the electricity consumption of various infrastructure and equipment in the community, including the electricity consumption data of street lights, elevators, and water supply and power supply systems. At hourly intervals, including the power consumption and operating duration of each device, clean the collected data to remove outliers; S2.

2. Classify and statistically analyze the electricity consumption data after data cleaning in step S2.1 by season. Use the pandas library in Python to classify the electricity consumption of equipment by season, thereby establishing a hierarchical structure for the electricity load of community infrastructure, and dividing each equivalent part into different levels according to its load characteristics; S2.

3. Establish a single-community infrastructure electricity load model, with the expression: Y1 = w1x1 + w2x2 + w3x3 + b' where x1 is the current season, w1 is the eigenvalue of the current season, x2 is the current date, w2 is the eigenvalue of the current date, x3 is the current time point, w3 is the eigenvalue of the current time point, Y1 is the output electricity load, and b′ is the bias; When the number of samples is multiple, obtain the total electricity load of the community infrastructure, with the expression: Y = X′w where Y is the total electricity load of the community infrastructure, X′ is the attribute matrix, and w is the eigenvalue matrix; The mean loss of multiple samples is calculated using the mean squared error loss function to obtain the bias b', so as to minimize the total loss of the training samples. The mean squared error loss function l (i) (w, b′) is expressed as: Among them, is the predicted value of the output electricity load corresponding to infrastructure i, and y (i) is the true value of the output electricity load corresponding to infrastructure i.

3. The optimization control method for a community charging pile according to claim 2, characterized in that, The specific implementation method of step S3 includes the following steps: S3.

1. Collect the daily electricity consumption of community residents in a year, and collect data on the weather and whether it is a holiday on that day. Fit the daily electricity load of community residents with the current time, weather, and whether it is a holiday on that day to obtain a prediction model for the daily electricity load of community residents' daily life. The expression for the current-time electricity load H of community residents' daily life is: H = r1k1 + r2k2 + r3k3 + c where k1 is the current time on that day, r1 is the eigenvalue of the current time on that day, k2 is the weather, r2 is the eigenvalue of the weather, k3 is whether it is a holiday, r3 is the eigenvalue of whether it is a holiday, and c is the bias of the prediction model for the daily electricity load of community residents' daily life; S3.

2. Set the method for identifying outliers by judging whether the electricity load of the i-th sample deviates from 2 times the standard deviation, with the expression: where H i is the electrical load of the i-th sample, μ is the mean of n samples, and σ is the standard deviation of the samples; When the following conditions are met, H i is considered an outlier, and the expression is: |H i -μ| > 2*σ Set the method for handling missing values of samples by filling them with the median, with the expression: H filled = median(H1, H2, H3, …, H n ) where median(·) represents the median of the data; S3.

3. Combine the current electricity consumption load R of community residents actually observed by the smart meter now for modeling, and the expression is: X T Xr′ = X T R now Among them, X is the characteristic equation, and each row represents the attribute characteristics of a sample; r′ is the regression coefficient vector; R now is the current electricity consumption load of the community residents actually observed; Use the root mean square error between different features and the electricity load in the prediction model for the daily electricity load of community residents as the first function, and through optimization algorithm fitting, make the minimum value of the first function the optimal target to obtain the optimal feature parameters. The expression for the root mean square error MSE is: where n is the number of samples, and R nowi is the current moment's residential electricity load of the i-th sample actually observed, is the predicted daily electricity load of the residents in the i-th sample; the obtained r' is the optimal parameter of the prediction model for the daily electricity load of the residents in the community.

4. The optimized control method for community charging piles according to claim 3, characterized in that, The specific implementation method of step S4 includes the following steps: S4.

1. Calculate the total available electricity load T(t), with the expression: T(t) = K - H - Y; S4.

2. Based on the fact that the load rate of the transformer is usually controlled between 75% and 85% of its rated capacity, so set the control of the total electricity load to 85% of the rated capacity, and establish a model for the available electricity load of charging piles, with the expression: T′(t) = K * 0.85 - H - Y where T′(t) is the available electricity load of the charging pile.

5. An electronic device, characterized in that, It includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of a method for optimizing the control of community charging piles according to any one of claims 1 - 4.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a method for optimizing the control of community charging piles according to any one of claims 1 - 4.

Images