Photovoltaic power consumption scheduling system and method based on photovoltaic prediction and battery replacement prediction

Abstract

The application provides a scheduling system and method for realizing photovoltaic consumption based on photovoltaic prediction and battery replacement prediction, comprising: a data acquisition module for acquiring data in real time; a prediction module for predicting data in 96 future time periods based on a CNN (Convolutional Neural Network) and an LSTM (Long Short Term Memory) double network model, a trained photovoltaic power generation prediction model and a battery replacement frequency prediction model of a battery replacement station, and 96 historical data before a current time point; a decision center module for controlling the charging of batteries in a battery compartment of the battery replacement station; a charging control module for receiving instructions from the decision center module, and realizing the start and stop of charging machines, charging power setting; and an operation notification module for receiving instructions from the decision center module. The application realizes the balance between photovoltaic power generation and power supply of the battery replacement station, effectively utilizes the photovoltaic clean energy to the maximum extent, improves the vehicle power supply efficiency, and maximizes the operation benefit of the battery replacement station.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of new energy battery swap technology, in particular to a scheduling system and method for realizing photovoltaic consumption based on photovoltaic prediction and battery swap prediction. BACKGROUND

[0002] With the popularization and use of electric vehicles and intelligent electric vehicle battery swap stations, more and more electric vehicle battery swap stations will tend to use clean energy such as photovoltaic to supplement the power of the battery swap station. When supplementing the power of the existing battery swap station, the commonly used method is to directly use the mains power to charge at high power when the battery compartment is out of power or to realize time-of-use charging based on the price of the local power grid. Although this method is helpful for the operation of the battery swap station, it will cause a certain impact on the power grid. After the battery swap station is combined with photovoltaic to form a battery swap station microgrid, the power supplement structure of the battery swap station changes from single mains power supplement to mains power plus photovoltaic combination supplement. Although the battery swap station is supplemented by photovoltaic during the day, it is still supplemented by mains power at night. Moreover, the uncertainty of photovoltaic power generation will also have a certain impact on the power supplement and operation of the entire battery swap station. Therefore, a scheduling method and system are needed to achieve a balance among photovoltaic power generation, mains power, and battery swap station power supplement (operation). SUMMARY

[0003] The present application aims to provide a scheduling system and method for realizing photovoltaic consumption based on photovoltaic prediction and battery swap prediction to overcome the above problems or at least partially solve the above problems.

[0004] To achieve the above purpose, the technical solution of the present application is as follows:

[0005] The present application provides a scheduling system for realizing photovoltaic consumption based on photovoltaic prediction and battery swap prediction, which comprises:

[0006] A collection module for collecting data in real time, collecting the power generation of the photovoltaic system and the number of battery swaps in the battery swap station every 15 minutes, collecting the historical photovoltaic power generation and the historical number of battery swaps of the battery swap station from the previous year, the interval time being 15 minutes, collecting the historical charging capacity of each battery pack in the battery swap station, and collecting the number of available batteries and the number of unavailable batteries in the battery swap station;

[0007] A prediction module based on a CNN convolutional neural network and an LSTM long short-term memory network dual network model, which combines the historical data collected by the collection module every certain period of time to construct, iteratively optimize, and upgrade the existing photovoltaic power generation prediction model and the battery swap station battery swap number prediction model, and based on the trained photovoltaic power generation prediction model and the battery swap station battery swap number prediction model, the future 96 time periods are predicted based on the 96 historical data before the current time point, and the interval time is 15 minutes;

[0008] A decision center module combines the predicted photovoltaic power at 96 future time points and the number of battery replacements predicted by the prediction module and the historical average charging power of each battery pack to control the charging of the batteries in the battery compartment of the battery swap station.

[0009] A charging control module receives instructions from the decision center module to start and stop the charging of the charger and set the charging power.

[0010] An operation notification module receives instructions from the decision center module, and the instruction information includes the discount time point, the number of coupons, and the operation of issuing coupons to the driver side.

[0011] As a further scheme of the present application, the historical data collected by the collection module includes photovoltaic power and the number of battery replacements.

[0012] The scheduling method for photovoltaic consumption based on photovoltaic prediction and battery replacement prediction comprises the following steps:

[0013] Step one, the collection module collects data in real time, collects the power generation of the photovoltaic system and the number of battery replacements in the battery swap station every 15 minutes, collects the historical photovoltaic power and the historical number of battery replacements of the battery swap station in the past year, the interval time is 15 minutes, collects the historical charging power of each battery pack in the battery swap station, collects the number of available batteries and the number of unavailable batteries in the compartment, and based on the CNN convolutional neural network and the LSTM long short-term memory network double network model, every time interval, the historical data collected by the collection module is combined to construct, iterate, optimize and upgrade the existing photovoltaic power prediction model and the battery replacement number prediction model of the battery swap station, and based on the trained photovoltaic power prediction model and the battery replacement number prediction model of the battery swap station, the future 96 time periods are predicted based on the 96 historical data before the current time point.

[0014] Step two, the decision center module judges whether the photovoltaic power generation in the future 1h meets the charging of the battery replacement in the future 1h based on the input photovoltaic power n1 in the future 1h, the number of battery replacements n2 in the future 1h and the average charging power of each battery ɑ, that is, whether n1≥n2*α.

[0015] Step three, if the photovoltaic power generation in the future 1h does not meet the charging power of the battery replacement in the future 1h, it is judged whether the number of available batteries n4 in the compartment meets the number of battery replacements n2 in the future 1h, that is, whether n4≥n2.

[0016] If the number of available batteries in the warehouse n4 does not meet the number of battery replacement times in the future 1h n2, that is, n4 < n2, the number of batteries to be charged is obtained based on the difference between the number of available batteries in the warehouse n4 and the number of battery replacement times in the future 1h n2, that is, n2-n4, at this time, the photovoltaic and the power grid output together, and the photovoltaic power is obtained in real time by the acquisition module, the decision center module dynamically issues the charging power value to the charging control module, and the charging power of the charger is adjusted in real time by the charging control module to charge the battery to be charged;

[0017] If the number of available batteries in the warehouse n4 meets the number of battery replacement times in the future 1h n2, that is, n4≥n2, the number of unavailable batteries in the warehouse n5 is obtained at this time, and the photovoltaic output is directly obtained, and the photovoltaic power is obtained in real time by the acquisition module, the decision center module dynamically issues the charging power value to the charging control module, and the charging power of the charger is adjusted in real time by the charging control module to charge the unavailable battery.

[0018] Step four, if the photovoltaic power generation capacity in the future 1h meets the charging capacity of the battery replacement in the future 1h, that is, the photovoltaic power generation capacity in the future 1h is greater than the charging capacity of the battery replacement in the future 1h, the decision center module calculates the maximum power generation capacity that meets the number of battery replacements n6, obtains the number of available batteries in the warehouse n4 and the number of unavailable batteries in the warehouse n5, at this time n4+n5=n7, calculates the number of batteries to be charged: |n6-n4|, the decision center module calculates the number of battery replacements that the photovoltaic power generation meets in the future 1h, and issues instructions to the operation notification module, that is, the operation strategy: based on the value of |n6-n4| to issue battery replacement coupons, encourage users to replace batteries, increase the number of battery replacements in the future 1h, at this time, the photovoltaic output is completely obtained, the photovoltaic power is obtained in real time by the acquisition module, and the charging power of the charger is dynamically adjusted to charge the unavailable battery.

[0019] As a further scheme of the application, in the step one, the power generation capacity and the number of battery replacements are taken as the input of the prediction model, the prediction of the power generation capacity and the number of battery replacements at 96 time points in the future can be obtained, and the data of the first four sample points of the prediction value is taken as the input of the decision center module, that is, the photovoltaic power generation capacity n1 in the future 1h and the number of battery replacements n2 in the future 1h;

[0020] The average charging capacity of each battery in the history N days is calculated by the acquisition module, which is taken as the input of the decision center module;

[0021] The number of available batteries in the warehouse n4 and the number of unavailable batteries in the warehouse n5 are collected, which are taken as the input of the decision center module.

[0022] As a further scheme of the application, in the step four, the maximum power generation capacity that meets the number of battery replacements n6=n1%n2*ɑ, and n6≤n7, n7 is the maximum number of battery replacements in 1h.

[0023] The application provides a scheduling system and method for realizing photovoltaic consumption based on photovoltaic prediction and battery replacement prediction.

[0024] The method for realizing optimal scheduling of municipal power, photovoltaic power and battery replacement station based on big data technology and LSTM prediction algorithm, the balance of photovoltaic power generation and municipal power to battery replacement station based on photovoltaic power generation prediction and battery prediction required by battery replacement operation, avoids the direct use of municipal power in the peak period, avoids the impact on the power grid, maximizes the effective use of photovoltaic clean energy, improves the vehicle energy supplement efficiency, and realizes the maximization of the operation benefit of the battery replacement station. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0026] Figure 1 It is a working principle block diagram of the application.

[0027] Figure 2 It is a method flow block diagram of the application.

[0028] In the figure: 1, prediction module; 2, acquisition module; 3, decision center module; 4, charging control module; 5, operation notification module. DETAILED DESCRIPTION

[0029] The exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be accurately conveyed to those skilled in the art.

[0030] Referring to Figure 1 The scheduling system for realizing photovoltaic consumption based on photovoltaic prediction and battery replacement prediction provided by the embodiments of the application comprises:

[0031] The acquisition module 2 acquires data in real time, acquires the power generation of the photovoltaic system and the number of battery replacements of the battery replacement station in the time period every 15 minutes, and acquires the historical photovoltaic power generation and the historical number of battery replacements of the battery replacement station from the previous year, the interval time is 15 minutes, acquires the historical charging capacity of each battery pack in the battery replacement station, and acquires the number of available battery packs and the number of unavailable battery packs in the battery replacement station;

[0032] The prediction module 1 is based on a CNN convolutional neural network and an LSTM long short-term memory network double network model, and every time interval, the historical data collected by the collection module 2, including the photovoltaic power generation power and the battery replacement frequency, are used to construct, iteratively optimize and upgrade the existing photovoltaic power generation power prediction model and the battery replacement station battery replacement frequency prediction model. On the other hand, based on the trained photovoltaic power generation prediction model and the battery replacement station battery replacement frequency prediction model, the future 96 time periods of data are predicted by adding the 96 historical data before the current time point, with an interval of 15 minutes;

[0033] The decision center module 3 combines the photovoltaic power generation power and the battery replacement frequency predicted by the prediction module 1 at the future 96 time points, and the historical average charging power of each battery pack to control the charging of the battery in the battery compartment of the battery replacement station;

[0034] The charging control module 4 receives the instructions from the decision center module 3 to realize the charging start-stop and charging power setting of the charger;

[0035] The operation notification module 5 receives the instructions from the decision center module 3, and the instruction information includes the preferential time point, the coupon quantity, and the operation coupon issued to the driver side.

[0036] Referring to Figure 2 The photovoltaic prediction and battery replacement prediction are used to realize the scheduling method of photovoltaic consumption, which includes the following steps:

[0037] Step one, the collection module 2 collects data in real time, collects the photovoltaic system power generation power and the battery replacement station battery replacement frequency every 15 minutes, collects the historical photovoltaic power generation power and the historical battery replacement frequency of the battery replacement station for one year before the current time, and the interval time is 15 minutes. The historical charging power of each battery pack in the battery replacement station is collected, the number of available batteries in the battery replacement station and the number of unavailable batteries in the compartment are collected, and the CNN convolutional neural network and the LSTM long short-term memory network double network model are used to construct, iteratively optimize and upgrade the existing photovoltaic power generation power prediction model and the battery replacement station battery replacement frequency prediction model every time interval. Based on the trained photovoltaic power generation prediction model and the battery replacement station battery replacement frequency prediction model, the future 96 time periods of data are predicted by adding the 96 historical data before the current time point, and the prediction of the future 96 time points of power generation power and battery replacement frequency can be obtained. The data of the first four sample points of the prediction value are used as the input of the decision center module 3, that is, the photovoltaic power generation power n1 of the future 1h and the battery replacement frequency n2 of the future 1h. The average charging power a of each battery in the historical N days calculated by the collection module 2 is used as the input of the decision center module 3. The number of available batteries n4 in the compartment and the number of unavailable batteries n5 in the compartment are used as the input of the decision center module 3;

[0038] Step two, the decision center module 3 judges whether the photovoltaic power generation capacity in the future 1h meets the charging capacity of the battery to be replaced in the future 1h based on the input of the photovoltaic power generation capacity in the future 1h n1, the number of battery replacement in the future 1h n2 and the average charging capacity of each battery a, that is, whether n1≥n2*α;

[0039] Step three, if the photovoltaic power generation capacity in the future 1h does not meet the charging capacity of the battery to be replaced in the future 1h, it is judged whether the number of available batteries in the warehouse n4 meets the number of battery replacement in the future 1h n2, that is, whether n4≥n2;

[0040] If the number of available batteries in the warehouse n4 does not meet the number of battery replacement in the future 1h n2, that is, n4

[0041] If the number of available batteries in the warehouse n4 meets the number of battery replacement in the future 1h n2, that is, n4≥n2, the number of unavailable batteries in the warehouse n5 is obtained, at this time the photovoltaic power is directly output, the photovoltaic power generation capacity is obtained by the acquisition module 2 in real time, the charging power value is dynamically issued by the decision center module 3 to the charging control module 4, the charging power of the charging machine is adjusted by the charging control module 4 in real time, and the unavailable battery is charged;

[0042] Step four, if the photovoltaic power generation capacity in the future 1h meets the charging capacity of the battery to be replaced in the future 1h, that is, the photovoltaic power generation capacity in the future 1h is greater than the charging capacity of the battery to be replaced in the future 1h, the decision center module 3 calculates the maximum number of battery replacement n6 that the power generation capacity meets (n6 = n1%n2*ɑ, and n6≤n7, n7 is the maximum number of battery replacement in 1h), the number of available batteries in the warehouse n4 and the number of unavailable batteries in the warehouse n5 are obtained, at this time n4+n5=n7, the number of batteries to be charged is calculated: |n6-n4|, the number of battery replacement that the photovoltaic power generation meets in the future 1h is calculated by the decision center module 3, and the instruction is issued to the operation notification module 5, that is, the operation strategy: based on the value of |n6-n4| to issue a battery replacement coupon to encourage users to replace the battery and increase the number of battery replacement in the future 1h, at this time the photovoltaic power is completely output, the photovoltaic power generation capacity is obtained by the acquisition module 2 in real time, the charging power of the charging machine is dynamically adjusted, and the unavailable battery is charged.

[0043] The above merely provides an example of the present application, and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall fall into the scope of claims of the present application.

Claims

1. A scheduling method for photovoltaic power consumption based on photovoltaic prediction and battery replacement prediction, characterized in that, The method comprises the following steps: Step one, the data acquisition module (2) collects data in real time, collects the power generation of the photovoltaic system and the number of battery swap times of the battery swap station in the time period every 15 minutes, collects the historical photovoltaic power generation and the historical number of battery swap times of the battery swap station in the past year from the current time, the interval time is 15 minutes, collects the historical charging capacity of each battery pack in the battery swap station, collects the number of available battery packs and the number of unavailable battery packs in the battery swap station, based on the CNN convolutional neural network and the LSTM long short-term memory network double network model, every time interval combines the historical data collected by the data acquisition module (2) to construct, iterate, optimize and upgrade the existing photovoltaic power generation prediction model and the battery swap station battery swap times prediction model, and based on the trained photovoltaic power generation prediction model and the battery swap station battery swap times prediction model, 96 historical data before the current time point are added to realize the prediction of the data in the future 96 time periods; Step two, the decision center module (3) judges whether the photovoltaic power generation in the future 1h meets the charging of the battery swap battery in the future 1h based on the input photovoltaic power generation n1 in the future 1h, the number of battery swaps n2 in the future 1h and the average charging capacity of each battery ɑ, that is, whether n1≥n2*α; Step three, if the photovoltaic power generation in the future 1h does not meet the charging capacity of the battery swap battery in the future 1h, whether the number of available battery packs n4 in the battery swap station meets the number of battery swaps n2 in the future 1h is judged, that is, whether n4≥n2; If the number of available battery packs n4 in the battery swap station does not meet the number of battery swaps n2 in the future 1h, that is, n4<n2, the difference between the number of available battery packs n4 and the number of battery swaps n2 in the future 1h is obtained, that is, n2-n4, at this time, the photovoltaic and the power grid output together, the photovoltaic power generation is obtained in real time by the data acquisition module (2), the charging power value is dynamically issued to the charging control module (4) by the decision center module (3), the charging power of the charging machine is adjusted in real time by the charging control module (4), and the battery to be charged is charged; If the number of available battery packs n4 in the battery swap station meets the number of battery swaps n2 in the future 1h, that is, n4≥n2, the number of unavailable battery packs n5 in the current battery swap station is obtained, at this time, the photovoltaic output is directly obtained, the photovoltaic power generation is obtained in real time by the data acquisition module (2), the charging power value is dynamically issued to the charging control module (4) by the decision center module (3), the charging power of the charging machine is adjusted in real time by the charging control module (4), and the unavailable battery is charged. Step four, if the future 1h photovoltaic power generation power meets the future 1h battery charging power, that is, the future 1h photovoltaic power generation power is greater than the future 1h battery charging power, the decision center module (3) calculates the maximum power generation power that meets the number of battery replacement n6, obtains the number of available batteries in the warehouse n4, and the number of unavailable batteries in the warehouse n5, at this time n4+n5=n7, calculates the number of batteries to be charged: |n6-n4|, the decision center module (3) calculates the number of battery replacements that the photovoltaic power generation meets in the future 1h, and sends instructions to the operation notification module (5), that is, the operation strategy: based on the value of |n6-n4|, issue a battery replacement coupon to encourage users to replace the battery and increase the number of battery replacements in the future 1h, at this time, the photovoltaic output is completely used, the acquisition module (2) acquires the photovoltaic power in real time, and dynamically adjusts the charging power of the charger to charge the unavailable batteries.

2. The scheduling method for realizing photovoltaic consumption based on photovoltaic prediction and battery replacement prediction, according to claim 1, characterized in that, In the step one, the power generation power and the number of battery replacements are taken as the input of the prediction model, the prediction of the power generation power and the number of battery replacements at 96 time points in the future can be obtained, and the data of the first 4 samples of the prediction value are taken as the input of the decision center module (3), that is, the photovoltaic power generation power n1 in the future 1h and the number of battery replacements n2 in the future 1h; The average charging power ɑ of each battery in the history of N days is calculated by the acquisition module (2) as the input of the decision center module (3); The number of available batteries in the warehouse n4 and the number of unavailable batteries in the warehouse n5 are acquired as the input of the decision center module (3).

3. The scheduling method for realizing photovoltaic consumption based on photovoltaic prediction and battery replacement prediction, according to claim 1, characterized in that, In the step four, the maximum power generation power that meets the number of battery replacements n6=n1%n2*ɑ, and n6≤n7, n7 is the maximum number of battery replacements in 1h.

4. A scheduling system for scheduling a photovoltaic power consumption based on the photovoltaic prediction and the battery replacement prediction according to any one of claims 1-3, characterized in that, It includes: The acquisition module (2) acquires data in real time, acquires the photovoltaic system power generation power and the number of battery replacements in the battery replacement station in the time period every 15 minutes, and acquires the historical photovoltaic power generation power and the historical number of battery replacements of the battery replacement station from the current time in the past 1 year, the interval time is 15 minutes, acquires the historical charging power of each battery pack in the battery replacement station, and acquires the number of available batteries in the battery replacement station and the number of unavailable batteries in the warehouse; The prediction module (1) is based on a double network model of CNN convolutional neural network and LSTM long short term memory network, every time interval combines the historical data collected by the acquisition module (2), constructs, iteratively optimizes and upgrades the existing photovoltaic power generation power prediction model and the number of battery replacements in the battery replacement station, and based on the trained photovoltaic power generation prediction model and the number of battery replacements in the battery replacement station, adds the 96 historical data before the current time point to realize the prediction of the data in the future 96 time periods, and the interval time is 15 minutes; The decision center module (3) combines the photovoltaic power generation power and the number of battery replacements at 96 time points predicted by the prediction module (1) and the historical average charging power of each battery pack to realize the control of the charging of the batteries in the battery replacement station; The charging control module (4) receives the instructions from the decision center module (3) to realize the start and stop of the charger and the setting of the charging power. An operation notification module (5) receives instructions from the decision center module (3), and the instruction information includes a preferential time point, a number of coupons, and the operation of issuing preferential coupons to the driver side.

5. The scheduling system for realizing photovoltaic power consumption based on photovoltaic prediction and battery swap prediction, according to claim 4, characterized in that, The historical data collected by the collection module (2) includes photovoltaic power generation power and battery replacement times.

Images