WO2019105065A1 - 基于电子地图的充电请求发起时间预测方法和装置 - Google Patents

基于电子地图的充电请求发起时间预测方法和装置 Download PDF

Info

Publication number
WO2019105065A1
WO2019105065A1 PCT/CN2018/100030 CN2018100030W WO2019105065A1 WO 2019105065 A1 WO2019105065 A1 WO 2019105065A1 CN 2018100030 W CN2018100030 W CN 2018100030W WO 2019105065 A1 WO2019105065 A1 WO 2019105065A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
user activity
charging request
index
electronic map
Prior art date
Application number
PCT/CN2018/100030
Other languages
English (en)
French (fr)
Inventor
牛姣姣
卢雨昕
吴毅成
Original Assignee
蔚来汽车有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 蔚来汽车有限公司 filed Critical 蔚来汽车有限公司
Priority to EP18883405.5A priority Critical patent/EP3719735A4/en
Publication of WO2019105065A1 publication Critical patent/WO2019105065A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/40—Business processes related to the transportation industry
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply

Definitions

  • the present invention relates to the field of charging request initiation time prediction technology, and in particular, to an electronic map based charging request initiation time prediction method and apparatus.
  • the rational location and construction of electric vehicle charging and replacement infrastructure is becoming more and more important, directly affecting the electric vehicle owner's power-on experience, and the prediction of electric vehicle users' power demand directly affects charging.
  • the result of site selection and construction of the power exchange infrastructure such as the prediction of the time distribution of the charging request initiation.
  • the existing charging request initiation time prediction is based on the regularity of people's work schedule, the commute time rule, etc., and the prediction accuracy is low, the guidance precision for charging and replacing construction is low, and the user's power-on experience is poor.
  • the technical problem to be solved by the present invention is to provide an electronic map-based charging request initiation time prediction method and apparatus, wherein the user activity index of the electronic map and the number of charging requests are positively correlated, and therefore, in the selected area and the selected time
  • the user active index is statistically analyzed to obtain the distribution time of the charging request, which improves the accuracy and reliability of the charging request initiation time prediction.
  • an electronic map-based charging request initiation time prediction method including:
  • each POI point corresponding to a set of index data, where the index data includes a user activity index corresponding to different time points;
  • the user activity level data corresponding to the multiple POI points is optimized to obtain final charging request time distribution data, including the following steps:
  • Each type of the user activity level data is optimized to obtain a final charging request time distribution data.
  • obtaining a user activity index of a plurality of POI points on the electronic map includes the following steps:
  • a user activity index corresponding to each POI point at each of the points in time is obtained.
  • normalizing each set of index data to obtain user activity level data corresponding to each POI point includes the following processing:
  • classifying user activity level data corresponding to the plurality of POI points includes the following steps:
  • the user charging request origination location is divided into N categories, where N is an integer greater than or equal to 2;
  • the similarity is compared with a corresponding similarity threshold. If the similarity threshold is higher, the set of user activity data is divided into corresponding location categories.
  • the optimizing processing of each type of the user activity level data includes the following steps: performing cluster analysis on each group of user activity level data to obtain final charging request time distribution data of each type of location. .
  • each set of index data is stored in a form or a curve.
  • an electronic map-based charging request initiation time prediction apparatus including:
  • An obtaining module configured to acquire a user activity index of a plurality of POI points on the electronic map, where each POI point corresponds to a set of index data, where the index data includes a user activity index corresponding to different time points;
  • a processing module configured to normalize each set of index data to obtain user activity level data corresponding to each POI point
  • the optimization module is configured to optimize the user activity level data corresponding to the plurality of POI points to obtain final charging request time distribution data.
  • optimization module includes:
  • a classification sub-module configured to classify user activity level data corresponding to the multiple POI points
  • the optimization submodule is configured to optimize each type of the user activity level data to obtain a final charging request time distribution data.
  • the obtaining module includes:
  • a POI point obtaining unit configured to select a plurality of POI points corresponding to the area to be predicted on the map
  • a time dividing unit configured to set a time interval, and divide the time period to be predicted into multiple time points according to the time interval
  • a data obtaining unit configured to acquire a user activity index corresponding to each POI point at each of the time points.
  • processing module includes:
  • a coordinate system establishing unit for establishing a coordinate system with time as an abscissa and a user active index as an ordinate;
  • a curve obtaining unit configured to acquire a curve corresponding to each set of index data on the coordinate system
  • the normalization unit is configured to keep the shape of each curve unchanged, adjust the position corresponding to the ordinate, and make the curve integrals for each curve equal, thereby obtaining the user activity level data corresponding to each POI point.
  • classification submodule includes:
  • a location classification unit configured to classify a user charging request origination location into N categories, where N is an integer greater than or equal to 2;
  • a data preset unit configured to set a set of preset user index degree data for each type of location, and set a corresponding similarity threshold
  • a data comparison unit configured to compare the set of user activity level data with the preset user index degree data to obtain a similarity
  • a data classification unit configured to compare the similarity with a corresponding similarity threshold, and if the similarity threshold is higher, divide the set of user activity data into a corresponding location category.
  • optimization sub-module is further configured to perform cluster analysis on multiple sets of user activity level data of each class to obtain final charging request time distribution data of each type of location.
  • each set of index data is stored in a form or a curve.
  • a controller comprising a memory and a processor, the memory storing a computer program, the program being capable of implementing the steps of the method when executed by the processor.
  • a computer readable storage medium for storing computer instructions that, when executed by a computer or processor, implement the steps of the method.
  • the electronic map-based charging request initiation time prediction method and device can achieve considerable technical progress and practicability, and has extensive industrial use value, and has at least the following advantages:
  • the invention predicts the time distribution of the user charging request through the electronic map data, and obtains the prediction result based on the Internet big data, can more accurately fit the charging and discharging behavior of the user, and improves the accuracy and reliability of the charging request initiation time prediction, thereby It provides a basis for the rational location selection of electric vehicle charging and replacing infrastructure and guiding the construction of charging and replacing electricity, so that the location of charging and replacing infrastructure and the construction of charging and replacing electricity are more reasonable, thus enhancing the user's power-on experience.
  • FIG. 1 is a schematic diagram of a method for predicting a charging request initiation time based on an electronic map according to an embodiment of the present invention.
  • FIG. 2(a) is a schematic diagram showing a preset user index degree data curve corresponding to a residence area according to an embodiment of the present invention.
  • FIG. 2(b) is a schematic diagram of a preset user index degree data curve corresponding to a work site according to an embodiment of the present invention.
  • FIG. 2(c) is a schematic diagram of a preset user index degree data curve corresponding to a destination according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of a charging request initiation time prediction result based on an electronic map according to an embodiment of the present invention.
  • FIG. 4 is a schematic diagram of an electronic map-based charging request initiation time prediction apparatus according to an embodiment of the present invention.
  • an electronic map-based charging request initiation time prediction method includes:
  • Step S1 Obtain a user activity index of a plurality of POI (Point of Interest) points on the electronic map, and each POI point corresponds to a set of index data, where the index data includes a user activity index corresponding to different time points;
  • POI Point of Interest
  • the user activity index is based on the tens of billions of geographic location data, which is used to represent the data of the regional population's heat, which is proportional to the number of people in the region and the group activity.
  • the user activity index can be obtained directly through existing electronic map software.
  • the time distribution of the user initiated charging request is positively correlated with the user activity index. Therefore, the charging request initiation time distribution is obtained by acquiring the user active indication and performing statistical analysis.
  • acquiring a user activity index of a plurality of POI points on the electronic map includes the following steps:
  • Step S11 selecting a plurality of POI points corresponding to the area to be predicted on the map
  • the area to be predicted is Shanghai
  • a part of the area is selected on the map corresponding to the Shanghai area, including multiple POI points.
  • the initiation of the user charging request is related to the time and place.
  • the user charging request initiation time can be divided into two categories: working day and non-working day.
  • the place of origin can be divided into three categories: place of residence, place of work and destination, where the place of residence can be a community; the place of work can be an office building, a park, etc.; the destination can be a hospital, a school, a shopping mall, and the like.
  • the plurality of POI information points of the to-be-predicted area include a plurality of residences, a plurality of work places, and a plurality of destinations.
  • initiation time classification and the initiation location classification of the user charging request are not limited thereto, and may be classified according to specific prediction requirements and prediction targets.
  • Step S12 Set a time interval, and divide the time period to be predicted into multiple time points according to the time interval;
  • the time interval is set according to specific prediction requirements. For example, the time interval can be set to 1 hour, 30 minutes, 15 minutes, etc. The shorter the time interval, the more statistical data is obtained, and the statistical results are more accurate.
  • the time to be predicted is also set according to the predicted demand. It can be set to 24 hours, and the distribution time of the charging request in the area can be counted within one day. It can also be set from 8:00 am to 8:00 pm, set to 12 hours, and the time period is counted. The charging request time distribution in this area.
  • the following is an interval of 1 hour, the time to be predicted is 24 hours, and the example is from 0 to 0:
  • Step S13 Acquire a user activity index corresponding to each POI point at each of the time points.
  • the user activity index corresponding to each POI point at 0:00, 1 hour, 2 hours, 3 hours...22 hours, 23 hours, and 0 day 0 time can be obtained, and all acquired data are drawn through a table or through a coordinate system. The way to store.
  • Step S2 normalizing each set of index data to obtain user activity level data corresponding to each POI point;
  • the step S2 specifically includes the following steps:
  • Step S21 taking the time as the abscissa and the user active index as the ordinate to establish a coordinate system
  • Step S22 obtaining a curve corresponding to each set of index data on the coordinate system, thereby obtaining a plurality of curves
  • Step S23 Keep the shape of each curve unchanged, adjust the position corresponding to the ordinate, and make the curve integrals for each curve equal, thereby obtaining the user activity level data corresponding to each POI point.
  • Step S3 Optimize user activity level data corresponding to the plurality of POI points to obtain final charging request time distribution data.
  • step S3 can include:
  • the step S31 may include the following steps:
  • Step S311 dividing the user charging request initiation location into three categories: a place of residence, a work place, and a destination;
  • Step S312 setting a set of preset user index degree data for each type of location, and setting a corresponding similarity threshold
  • a set of preset user index degree data is set as a standard for each type of place, as shown in FIG. 2, and similarity comparison is performed with each set of data, that is, The classification can be completed.
  • Step S313 comparing each group of user activity level data with the preset user index degree data to obtain a similarity degree
  • step S314 the similarity is compared with a corresponding similarity threshold. If the similarity threshold is higher, the set of user activity data is divided into corresponding location categories.
  • each type of charging request origination location corresponds to multiple sets of user activity level data, that is, corresponding to multiple curves.
  • cluster analysis is performed on multiple sets of user activity level data of each class, and specifically: clustering algorithm: CLARA algorithm (Clustering LARge Applications, clustering method in large application), K-MEANS algorithm (K- The MEANS algorithm is an input cluster number k, and a database containing n data objects, and outputs k clusters satisfying the minimum variance criterion.
  • CLARA algorithm Clustering LARge Applications, clustering method in large application
  • K-MEANS algorithm K- The MEANS algorithm is an input cluster number k, and a database containing n data objects, and outputs k clusters satisfying the minimum variance criterion.
  • CLARANS algorithm Clustering Algorithm based on Randomized Search, large-scale application clustering algorithm based on random search
  • the charging request time distribution prediction data outputted in the distribution curve shown in FIG. 3 or in the form of a table related to time can be used.
  • the method of the invention predicts the time distribution of the user charging request by using the electronic map data, and obtains the prediction result based on the Internet big data, which can more accurately fit the charging and discharging behavior of the user, and improves the accuracy and reliability of the charging request initiation time prediction.
  • an electronic map-based charging request initiation time prediction apparatus includes an acquisition module 1, a processing module 2, and an optimization module 3, wherein the acquisition module 1 is configured to acquire user actives of multiple POI points on an electronic map. Index, each POI point corresponds to a set of index data, the index data includes a user activity index corresponding to different time points; the processing module 2 is configured to normalize each set of index data to obtain a user corresponding to each POI point. The activity level data is used by the optimization module 3 to optimize the user activity level data corresponding to the plurality of POI points to obtain the final charging request time distribution data.
  • the optimization module 3 further includes a classification sub-module and an optimization sub-module, wherein the classification sub-module is configured to classify user activity level data corresponding to the plurality of POI points; and the optimization sub-module is used for each type of the user activity level The data is optimized to obtain the final charging request time distribution data.
  • the user activity index is based on the tens of billions of geographic location data, which is used to represent the data of the regional population's heat, which is proportional to the number of people in the region and the group activity.
  • the user activity index can be obtained directly through existing electronic map software.
  • the time distribution of the user initiated charging request is positively correlated with the user activity index. Therefore, the charging request initiation time distribution is obtained by acquiring the user active indication and performing statistical analysis.
  • the acquisition module 1 includes a POI point acquisition unit, a time division unit, and a data acquisition unit, wherein the POI point acquisition unit is configured to select a plurality of POI points corresponding to the area to be predicted on the map; for example, the area to be predicted is the Shanghai area. Then select a part of the map on the map corresponding to the Shanghai area, including multiple POI points.
  • the initiation of the user charging request is related to the time and place.
  • the user charging request initiation time can be divided into two categories: working day and non-working day.
  • the place of origin can be divided into three categories: place of residence, place of work and destination, where the place of residence can be a community; the place of work can be an office building, a park, etc.; the destination can be a hospital, a school, a shopping mall, and the like.
  • the plurality of POI information points of the to-be-predicted area include a plurality of residences, a plurality of work places, and a plurality of destinations.
  • initiation time classification and the initiation location classification of the user charging request are not limited thereto, and may be classified according to specific prediction requirements and prediction targets.
  • the time dividing unit is configured to set a time interval, and divide the time period to be predicted into a plurality of time points according to the time interval; wherein the setting of the time interval is set according to a specific prediction requirement, for example, the time interval may be set to 1 Hours, 30 minutes, 15 minutes, etc. The shorter the time interval, the more statistics are obtained and the more accurate the statistical results.
  • the time to be predicted is also set according to the predicted demand. It can be set to 24 hours, and the distribution time of the charging request in the area can be counted within one day. It can also be set from 8:00 am to 8:00 pm, set to 12 hours, and the time period is counted.
  • the charging request time distribution in this area The following is an interval of 1 hour, the time to be predicted is 24 hours, and the example is from 0 to 0:
  • the data acquisition unit is configured to acquire a user activity index corresponding to each POI point at each of the time points.
  • the user activity index corresponding to each POI point at 0:00, 1 hour, 2 hours, 3 hours...22 hours, 23 hours, and 0 day 0 time can be obtained, and all acquired data are drawn through a table or through a coordinate system. The way to store.
  • the processing module 2 includes a coordinate system establishing unit, a curve acquiring unit, and a normalization unit, wherein the coordinate system establishing unit is configured to establish a coordinate system with time as the abscissa and a user active index as the ordinate; the curve obtaining unit is configured to acquire Each group of index data corresponds to a curve on the coordinate system; the normalization unit is used to keep the shape of each curve unchanged, adjust the position corresponding to the ordinate, and make the curve integrals for each curve equal, thereby obtaining each POI The user activity level data corresponding to the point. By normalizing each set of index data, the calculation process can be simplified and the accuracy of the prediction results can be improved.
  • the classification sub-module includes a location classification unit, a data preset unit, a data comparison unit, and a data classification unit, wherein the location classification unit is configured to classify the user charging request origination into three categories: a place of residence, a work place, and a destination.
  • the data preset unit is configured to set a set of preset user index degree data for each type of location, and set a corresponding similarity threshold; since each type of charging request originating place has a large difference in charging time distribution, each type of location is Set a set of preset user index degree data as a standard, as shown in Figure 2, and compare the similarity with each set of data to complete the classification.
  • the data comparison unit is configured to compare the set of user activity level data with the preset user index degree data to obtain a similarity; the data classification unit is configured to compare the similarity with a corresponding similarity threshold, if Above the similarity threshold, the set of user activity data is divided into corresponding location categories.
  • the user activity level data is divided into three categories by the classification sub-module, and each type of charging request initiation location corresponds to multiple sets of user activity degree data, that is, corresponding to multiple curves.
  • the optimization sub-module is further configured to perform cluster analysis on multiple sets of user activity level data of each class to obtain final charging request time distribution data of each type of location.
  • the optimization sub-module performs cluster analysis on multiple sets of user activity level data of each class, and specifically adopts a clustering algorithm: CLARA algorithm (Clustering LARG Applications, clustering method in large-scale application), K- MEANS algorithm (K-MEANS algorithm is the input cluster number k, and the database containing n data objects, output k clusters satisfying the minimum standard of variance), CLARANS algorithm (Clustering Algorithm based on Randomized Search, based on random search Large application clustering algorithm) and so on. Thereby the final charging request time distribution data for each type of location is obtained.
  • CLARA algorithm Clustering LARG Applications, clustering method in large-scale application
  • K- MEANS algorithm K-MEANS algorithm is the input cluster number k, and the database containing n data objects, output k clusters satisfying the minimum standard of variance
  • CLARANS algorithm Clustering Algorithm based on Randomized Search, based on random search Large application clustering algorithm
  • the final optimization sub-module may predict the data by the charging request time distribution outputted by the distribution curve shown in Fig. 3 or the time-dependent table form.
  • the device of the invention predicts the time distribution of the user charging request through the electronic map data, and obtains the prediction result based on the Internet big data, which can more accurately fit the charging and discharging behavior of the user, and improves the accuracy and reliability of the charging request initiation time prediction.
  • the invention also provides a controller comprising a memory and a processor, the memory storing a computer program, the program being capable of implementing the steps of the method when executed by the processor.
  • the present invention also provides a computer readable storage medium for storing computer instructions that, when executed by a computer or processor, implement the steps of the method.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • Marketing (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Quality & Reliability (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Primary Health Care (AREA)
  • Development Economics (AREA)
  • Water Supply & Treatment (AREA)
  • Public Health (AREA)
  • Game Theory and Decision Science (AREA)
  • Navigation (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种基于电子地图的充电请求发起时间预测方法和装置，所述方法包括：获取电子地图上多个POI点的用户活跃指数，每个POI点对应一组指数数据（S1），所述指数数据包括不同时间点对应的用户活跃指数；将每组指数数据进行归一化处理，得到每个POI点对应的用户活跃程度数据（S2）；将所述多个POI点对应的用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据（S3）。通过电子地图数据对用户充电请求时间分布进行预测，基于互联网大数据得到预测结果，能够更加真实的拟合用户的充换电行为，提高了充电请求发起时间预测的精度和可靠性，从而为电动汽车的充换电基础设施的合理选址及指导充换电建设提供基础，提升用户的加电体验。

Description

基于电子地图的充电请求发起时间预测方法和装置 技术领域
本发明涉及充电请求发起时间预测技术领域，尤其涉及一种基于电子地图的充电请求发起时间预测方法和装置。
背景技术
随着电动汽车的普及和发展，电动汽车的充换电基础设施的合理选址及建设越发重要，直接影响着电动汽车车主的加电体验，而对电动汽车用户加电需求的预测直接影响充换电基础设施的选址及建设的结果，例如充电请求发起时间分布的预测。然而，现有的充电请求发起时间预测多基于人们作息时间规律，上下班时间规律等进行粗略预测，预测精确度低，对充换电建设的指导精度低，用户加电体验差。
随着互联网的发展，人们可获得的数据也越来越多，因此可应用能够反映充电请求发起时间规律的数据来预测充电请求发起时间分布，来提高充电请求发起时间预测的精度和可靠性。
发明内容
本发明所要解决的技术问题在于，提供一种基于电子地图的充电请求发起时间预测方法和装置，电子地图的用户活跃指数和充电请求数呈正相关，因此，通过在选定区域和选定时间内对用户活跃指数进行统计分析，得出充电请求的发起时间分布，提高了充电请求发起时间预测的精度和可靠性。
根据本发明的一方面，提供了一种基于电子地图的充电请求发起时间预测方法，包括：
获取电子地图上多个POI点的用户活跃指数，每个POI点对应一组指数数据，所述指数数据包括不同时间点对应的用户活跃指数；
将每组指数数据进行归一化处理，得到每个POI点对应的用户活跃程度数据；
将所述多个POI点对应的用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
进一步的，所述将所述多个POI点对应的用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据，包括以下步骤：
将所述多个POI点对应的用户活跃程度数据进行分类；
对每一类所述用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
进一步的，获取电子地图上多个POI点的用户活跃指数包括以下步骤：
选取待预测的地区在地图上对应的多个POI点；
设定时间间隔，根据所述时间间隔将待预测时间段分为多个时间点；
获取每个POI点在每个所述时间点对应的用户活跃指数。
进一步的，将每组指数数据进行归一化处理，得到每个POI点对应的用户活跃程度数据包括以下处理：
以时间为横坐标，用户活跃指数为纵坐标建立坐标系；
获取每组指数数据在所述坐标系上对应的曲线；
保持每条曲线形状不变，调整纵坐标所对应位置，使对每条曲线的曲线积分相等，从而得到每个POI点对应的用户活跃程度数据。
进一步的，将所述多个POI点对应的用户活跃程度数据进行分类包括以下步骤：
将用户充电请求发起地点分为N类，其中，N为大于等于2的整数；
为每类地点设置一组预设用户指数程度数据，并设置对应的相似度阈值；
将所述每组用户活跃程度数据与所述预设用户指数程度数据进行对比，得到相似度；
将所述相似度与对应的相似度阈值进行对比，若高于所述相似度阈值，则将该组用户活跃程度数据划分到对应的地点类别中。
进一步的，所述对每一类所述用户活跃程度数据进行优化处理，包括以下步骤：对每类的多组用户活跃程度数据进行聚类分析，得到每类地点的最终的充电请求时间分布数据。
进一步的，所述每组指数数据采用表格或曲线的方式进行存储。
根据本发明的另一方面，提供了一种基于电子地图的充电请求发起时间预测装置，包括：
获取模块，用于获取电子地图上多个POI点的用户活跃指数，每个POI点对应一组指数数据，所述指数数据包括不同时间点对应的用户活跃指数；
处理模块，用于将每组指数数据进行归一化处理，得到每个POI点对应的用户活跃程度数据；
优化模块，用于将所述多个POI点对应的用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
进一步的，所述优化模块包括:
分类子模块，用于将所述多个POI点对应的用户活跃程度数据进行分类；
优化子模块，用于对每一类所述用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
进一步的，所述获取模块包括：
POI点获取单元，用于选取待预测的地区在地图上对应的多个POI点；
时间划分单元，用于设定时间间隔，根据所述时间间隔将待预测时间段分为多个时间点；
数据获取单元，用于获取每个POI点在每个所述时间点对应的用户活跃指数。
进一步的，所述处理模块包括：
坐标系建立单元，用于以时间为横坐标，用户活跃指数为纵坐标建立坐标系；
曲线获取单元，用于获取每组指数数据在所述坐标系上对应的曲线；
归一化单元，用于保持每条曲线形状不变，调整纵坐标所对应位置，使对每条曲线的曲线积分相等，从而得到每个POI点对应的用户活跃程度数据。
进一步的，所述分类子模块包括：
地点分类单元，用于将用户充电请求发起地点分为N类，其中，N为大于等于2的整数；
数据预设单元，用于为每类地点设置一组预设用户指数程度数据，并设置对应的相似度阈值；
数据对比单元，用于将所述每组用户活跃程度数据与所述预设用户指数程度数据进行对比，得到相似度；
数据分类单元，用于将所述相似度与对应的相似度阈值进行对比，若高于所述相似度阈值，则将该组用户活跃程度数据划分到对应的地点类别中。
进一步的，所述优化子模块还用于，对每类的多组用户活跃程度数据进行聚类分析，得到每类地点的最终的充电请求时间分布数据。
进一步的，所述每组指数数据采用表格或曲线的方式进行存储。
根据本发明又一方面，提供一种控制器，其包括存储器与处理器，所述存储器存储有计算机程序，所述程序在被所述处理器执行时能够实现所述方法的步骤。
根据本发明又一方面，提供一种计算机可读存储介质，用于存储计算机指令，所述指令在由一计算机或处理器执行时实现所述方法的步骤。
本发明与现有技术相比具有明显的优点和有益效果。借由上述技术方案，本发明一种基于电子地图的充电请求发起时间预测方法和装置可达到相当的技术进步性及实用性，并具有产业上的广泛利用价值，其至少具有下列优点：
本发明通过电子地图数据对用户充电请求时间分布进行预测，基于互 联网大数据得到预测结果，能够更加真实的拟合用户的充换电行为，提高了充电请求发起时间预测的精度和可靠性，从而为电动汽车的充换电基础设施的合理选址及指导充换电建设提供基础，使充换电基础设施选址及充换电建设更加合理，从而提升用户的加电体验。
上述说明仅是本发明技术方案的概述，为了能够更清楚了解本发明的技术手段，而可依照说明书的内容予以实施，并且为了让本发明的上述和其他目的、特征和优点能够更明显易懂，以下特举较佳实施例，并配合附图,详细说明如下。
附图说明
图1为本发明一实施例提供基于电子地图的充电请求发起时间预测方法示意图。
图2(a)为本发明一实施例提供的居住地对应的预设用户指数程度数据曲线示意图。
图2(b)为本发明一实施例提供的工作地对应的预设用户指数程度数据曲线示意图。
图2(c)为本发明一实施例提供的目的地对应的预设用户指数程度数据曲线示意图。
图3为本发明一实施例提供的基于电子地图的充电请求发起时间预测结果示意图。
图4为本发明一实施例提供的基于电子地图的充电请求发起时间预测装置的示意图。
主要附图标记说明：
1：获取模块
2：处理模块
3：优化模块
具体实施方式
为更进一步阐述本发明为达成预定发明目的所采取的技术手段及功效,以下结合附图及较佳实施例，对依据本发明提出的一种基于电子地图的充电请求发起时间预测方法和装置的具体实施方式及其功效，详细说明如后。
如图1所示，一种基于电子地图的充电请求发起时间预测方法，包括：
步骤S1、获取电子地图上多个POI(信息点，Point of Interest)点的用户活跃指数，每个POI点对应一组指数数据，所述指数数据包括不同时间点对应的用户活跃指数；
其中，用户活跃指数是基于百亿级地理位置数据，用于表征区域人群 热度情况的数据，与区域内人群数量、群体活跃度成正比。所述用户活跃指数可直接通过现有的电子地图软件获取。用户发起充电请求的时间分布与所述用户活跃指数呈正相关，因此，通过获取用户活跃指示并进行统计分析，从而得到充电请求发起时间分布。
具体地，获取电子地图上多个POI点的用户活跃指数包括以下步骤：
步骤S11、选取待预测的地区在地图上对应的多个POI点；
例如，待预测的地区为上海地区，则在上海地区对应的地图上选择一部分区域，包含多个POI点。
用户充电请求的发起与时间及地点均有关系，用户充电请求发起时间可以分为两大类：工作日和非工作日。发起地点可以分为三类：居住地、工作地和目的地，其中，居住地可以为小区；工作地可以为写字楼、园区等；目的地可以为医院、学校、商场等。上述待预测区域的多个POI信息点中包含多个居住地、多个工作地和多个目的地。
需要说明的是，用户充电请求的发起时间分类和发起地点分类并不限于此，可根据具体的预测需求和预测目标进行分类。
步骤S12、设定时间间隔，根据所述时间间隔将待预测时间段分为多个时间点；
其中，时间间隔的设定根据具体预测需求进行设定，例如，时间间隔可以设为1小时、30分钟、15分钟等，时间间隔越短，得到的统计数据越多，统计结果也越精确。待预测的时间也根据预测需求进行设定，可以设为24小时，统计一天内该区域的充电请求时间分布，也可以设为早8点到晚8点，设为12小时，统计该时间段内该区域充电请求时间分布。
以下以时间间隔为1小时，待预测时间为24小时，从0点到次日0点为例进行说明：
步骤S13、获取每个POI点在每个所述时间点对应的用户活跃指数。由此可得到每个POI点在0时、1时、2时、3时…22时、23时、次日0时对应的用户活跃指数，将所有获取的数据通过表格或通过坐标系描绘曲线的方式进行存储。
以下以通过坐标系描绘曲线的方式进行说明。
步骤S2、将每组指数数据进行归一化处理，得到每个POI点对应的用户活跃程度数据；
所述步骤S2具体包括以下步骤：
步骤S21、以时间为横坐标，用户活跃指数为纵坐标建立坐标系；
步骤S22、获取每组指数数据在所述坐标系上对应的曲线，从而得到多个曲线；
步骤S23、保持每条曲线形状不变，调整纵坐标所对应位置，使对每条 曲线的曲线积分相等，从而得到每个POI点对应的用户活跃程度数据。
通过对每组指数数据进行归一化处理，能够简化计算流程，提高预测结果的准确度。
步骤S3、将所述多个POI点对应的用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
步骤S2得到的用户活跃程度数据包括所有用户充电发起地点对应的用户活跃程度数据，因此可按照用户充电发起地点的类型将用户活跃程度数据进行分类，分别进行每一类地点的充电请求发起时间预测。因此步骤S3可包括：
S31、将所述多个POI点对应的用户活跃程度数据进行分类；
所述步骤S31可包括以下步骤：
步骤S311、将用户充电请求发起地点分为3类：居住地、工作地和目的地；
步骤S312、为每类地点设置一组预设用户指数程度数据，并设置对应的相似度阈值；
由于每一类充电请求发起地点的充电时间分布规律差异很大，因此为每类地点设置一组预设用户指数程度数据作为标准，如图2所示，与每组数据进行相似度对比，即可完成分类。
步骤S313、将所述每组用户活跃程度数据与所述预设用户指数程度数据进行对比，得到相似度；
步骤S314、将所述相似度与对应的相似度阈值进行对比，若高于所述相似度阈值，则将该组用户活跃程度数据划分到对应的地点类别中。
通过上述步骤，将所有用户活跃程度数据分为三类，每一类充电请求发起地点对应多组用户活跃程度数据，即对应多条曲线。
S32、对每一类所述用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
具体地，对每类的多组用户活跃程度数据进行聚类分析，具体可采用聚类算法中的：CLARA算法(Clustering LARge Applications，大型应用中的聚类方法)、K-MEANS算法(K-MEANS算法是输入聚类个数k，以及包含n个数据对象的数据库，输出满足方差最小标准的k个聚类)、CLARANS算法(Clustering Algorithm based on Randomized Search,基于随机搜索的大型应用聚类算法)等。从而得到每类地点的最优充电请求时间分布数据。
最终可以以图3所示的分布曲线或者跟时间相关的表格形式输出的充电请求时间分布预测数据。
本发明所述方法通过电子地图数据对用户充电请求时间分布进行预 测，基于互联网大数据得到预测结果，能够更加真实的拟合用户的充换电行为，提高了充电请求发起时间预测的精度和可靠性，从而为电动汽车的充换电基础设施的合理选址及指导充换电建设提供基础，使充换电基础设施选址及充换电建设更加合理，从而提升用户的加电体验。
如图4所示，一种基于电子地图的充电请求发起时间预测装置，包括获取模块1、处理模块2、优化模块3，其中，获取模块1用于获取电子地图上多个POI点的用户活跃指数，每个POI点对应一组指数数据，所述指数数据包括不同时间点对应的用户活跃指数；处理模块2用于将每组指数数据进行归一化处理，得到每个POI点对应的用户活跃程度数据；优化模块3用于将多个POI点对应的用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。优化模块3又包括分类子模块和优化子模块，其中，分类子模块用于将所述多个POI点对应的用户活跃程度数据进行分类；优化子模块用于对每一类所述用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
其中，用户活跃指数是基于百亿级地理位置数据，用于表征区域人群热度情况的数据，与区域内人群数量、群体活跃度成正比。所述用户活跃指数可直接通过现有的电子地图软件获取。用户发起充电请求的时间分布与所述用户活跃指数呈正相关，因此，通过获取用户活跃指示并进行统计分析，从而得到充电请求发起时间分布。
获取模块1包括POI点获取单元、时间划分单元和数据获取单元，其中，POI点获取单元用于选取待预测的地区在地图上对应的多个POI点；例如，待预测的地区为上海地区，则在上海地区对应的地图上选择一部分区域，包含多个POI点。
用户充电请求的发起与时间及地点均有关系，用户充电请求发起时间可以分为两大类：工作日和非工作日。发起地点可以分为三类：居住地、工作地和目的地，其中，居住地可以为小区；工作地可以为写字楼、园区等；目的地可以为医院、学校、商场等。上述待预测区域的多个POI信息点中包含多个居住地、多个工作地和多个目的地。
需要说明的是，用户充电请求的发起时间分类和发起地点分类并不限于此，可根据具体的预测需求和预测目标进行分类。
时间划分单元用于设定时间间隔，根据所述时间间隔将待预测时间段分为多个时间点；其中，时间间隔的设定根据具体预测需求进行设定，例如，时间间隔可以设为1小时、30分钟、15分钟等，时间间隔越短，得到的统计数据越多，统计结果也越精确。待预测的时间也根据预测需求进行设定，可以设为24小时，统计一天内该区域的充电请求时间分布，也可以设为早8点到晚8点，设为12小时，统计该时间段内该区域充电请求时间 分布。以下以时间间隔为1小时，待预测时间为24小时，从0点到次日0点为例进行说明：
数据获取单元用于获取每个POI点在每个所述时间点对应的用户活跃指数。由此可得到每个POI点在0时、1时、2时、3时…22时、23时、次日0时对应的用户活跃指数，将所有获取的数据通过表格或通过坐标系描绘曲线的方式进行存储。
以下以通过坐标系描绘曲线的方式进行说明。
所述处理模块2包括坐标系建立单元、曲线获取单元和归一化单元，其中，坐标系建立单元用于以时间为横坐标，用户活跃指数为纵坐标建立坐标系；曲线获取单元用于获取每组指数数据在所述坐标系上对应的曲线；归一化单元用于保持每条曲线形状不变，调整纵坐标所对应位置，使对每条曲线的曲线积分相等，从而得到每个POI点对应的用户活跃程度数据。通过对每组指数数据进行归一化处理，能够简化计算流程，提高预测结果的准确度。
所述分类子模块包括地点分类单元、数据预设单元、数据对比单元和数据分类单元，其中，地点分类单元用于将用户充电请求发起地点分为3类：居住地、工作地和目的地。数据预设单元用于为每类地点设置一组预设用户指数程度数据，并设置对应的相似度阈值；由于每一类充电请求发起地点的充电时间分布规律差异很大，因此为每类地点设置一组预设用户指数程度数据作为标准，如图2所示，与每组数据进行相似度对比，即可完成分类。数据对比单元用于将所述每组用户活跃程度数据与所述预设用户指数程度数据进行对比，得到相似度；数据分类单元用于将所述相似度与对应的相似度阈值进行对比，若高于所述相似度阈值，则将该组用户活跃程度数据划分到对应的地点类别中。通过所述分类子模块将所有用户活跃程度数据分为三类，每一类充电请求发起地点对应多组用户活跃程度数据，即对应多条曲线。
所述优化子模块还用于，对每类的多组用户活跃程度数据进行聚类分析，得到每类地点的最终的充电请求时间分布数据。
具体地，所述优化子模块对每类的多组用户活跃程度数据进行聚类分析，具体可采用聚类算法中的：CLARA算法(Clustering LARge Applications，大型应用中的聚类方法)、K-MEANS算法(K-MEANS算法是输入聚类个数k，以及包含n个数据对象的数据库，输出满足方差最小标准的k个聚类)、CLARANS算法(Clustering Algorithm based on Randomized Search,基于随机搜索的大型应用聚类算法)等。从而得到每类地点的最终的充电请求时间分布数据。
最终优化子模块可以以图3所示的分布曲线或者跟时间相关的表格形 式输出的充电请求时间分布预测数据。
本发明所述装置通过电子地图数据对用户充电请求时间分布进行预测，基于互联网大数据得到预测结果，能够更加真实的拟合用户的充换电行为，提高了充电请求发起时间预测的精度和可靠性，从而为电动汽车的充换电基础设施的合理选址及指导充换电建设提供基础，使充换电基础设施选址及充换电建设更加合理，从而提升用户的加电体验。
本发明还提供一种控制器，其包括存储器与处理器，所述存储器存储有计算机程序，所述程序在被所述处理器执行时能够实现所述方法的步骤。
本发明还提供一种计算机可读存储介质，用于存储计算机指令，所述指令在由一计算机或处理器执行时实现所述方法的步骤。
以上所述，仅是本发明的较佳实施例而已，并非对本发明作任何形式上的限制，虽然本发明已以较佳实施例揭露如上，然而并非用以限定本发明,任何熟悉本专业的技术人员，在不脱离本发明技术方案范围内,当可利用上述揭示的技术内容作出些许更动或修饰为等同变化的等效实施例,但凡是未脱离本发明技术方案的内容，依据本发明的技术实质对以上实施例所作的任何简单修改、等同变化与修饰，均仍属于本发明技术方案的范围内。

Claims (16)

  1. 一种基于电子地图的充电请求发起时间预测方法，其特征在于：包括：
    获取电子地图上多个POI点的用户活跃指数，每个POI点对应一组指数数据，所述指数数据包括不同时间点对应的用户活跃指数；
    将每组指数数据进行归一化处理，得到每个POI点对应的用户活跃程度数据；
    将所述多个POI点对应的用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
  2. 根据权利要求1所述的基于电子地图的充电请求发起时间预测方法，其特征在于：
    所述将所述多个POI点对应的用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据，包括以下步骤：
    将所述多个POI点对应的用户活跃程度数据进行分类；
    对每一类所述用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
  3. 根据权利要求1或2所述的基于电子地图的充电请求发起时间预测方法，其特征在于：
    所述获取电子地图上多个POI点的用户活跃指数包括以下步骤：
    选取待预测的地区在地图上对应的多个POI点；
    设定时间间隔，根据所述时间间隔将待预测时间段分为多个时间点；
    获取每个POI点在每个所述时间点对应的用户活跃指数。
  4. 根据权利要求1或2所述的基于电子地图的充电请求发起时间预测方法，其特征在于：
    所述将每组指数数据进行归一化处理，得到每个POI点对应的用户活跃程度数据包括以下处理：
    以时间为横坐标，用户活跃指数为纵坐标建立坐标系；
    获取每组指数数据在所述坐标系上对应的曲线；
    保持每条曲线形状不变，调整纵坐标所对应位置，使对每条曲线的曲线积分相等，从而得到每个POI点对应的用户活跃程度数据。
  5. 根据权利要求2所述的基于电子地图的充电请求发起时间预测方法，其特征在于：
    所述将所述多个POI点对应的用户活跃程度数据进行分类包括以下步骤：
    将用户充电请求发起地点分为N类，其中，N为大于等于2的整数；
    为每类地点设置一组预设用户指数程度数据，并设置对应的相似度阈值；
    将所述每组用户活跃程度数据与所述预设用户指数程度数据进行对比，得到相似度；
    将所述相似度与对应的相似度阈值进行对比，若高于所述相似度阈值，则将该组用户活跃程度数据划分到对应的地点类别中。
  6. 根据权利要求2所述的基于电子地图的充电请求发起时间预测方法，其特征在于：
    所述对每一类所述用户活跃程度数据进行优化处理，包括以下步骤：对每类的多组用户活跃程度数据进行聚类分析，得到每类地点的最终的充电请求时间分布数据。
  7. 根据权利要求1或2所述的基于电子地图的充电请求发起时间预测方法，其特征在于：
    所述每组指数数据采用表格或曲线的方式进行存储。
  8. 一种基于电子地图的充电请求发起时间预测装置，其特征在于：包括：
    获取模块，用于获取电子地图上多个POI点的用户活跃指数，每个POI点对应一组指数数据，所述指数数据包括不同时间点对应的用户活跃指数；
    处理模块，用于将每组指数数据进行归一化处理，得到每个POI点对应的用户活跃程度数据；
    优化模块，用于将所述多个POI点对应的用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
  9. 根据权利要求8所述的基于电子地图的充电请求发起时间预测装置，其特征在于：
    所述优化模块包括:
    分类子模块，用于将所述多个POI点对应的用户活跃程度数据进行分类；
    优化子模块，用于对每一类所述用户活跃程度数据进行优化处理，得到最终的充电请求时间分布数据。
  10. 根据权利要求8或9所述的基于电子地图的充电请求发起时间预测装置，其特征在于：
    所述获取模块包括：
    POI点获取单元，用于选取待预测的地区在地图上对应的多个POI点；
    时间划分单元，用于设定时间间隔，根据所述时间间隔将待预测时间段分为多个时间点；
    数据获取单元，用于获取每个POI点在每个所述时间点对应的用户活 跃指数。
  11. 根据权利要求8或9所述的基于电子地图的充电请求发起时间预测装置，其特征在于：
    所述处理模块包括：
    坐标系建立单元，用于以时间为横坐标，用户活跃指数为纵坐标建立坐标系；
    曲线获取单元，用于获取每组指数数据在所述坐标系上对应的曲线；
    归一化单元，用于保持每条曲线形状不变，调整纵坐标所对应位置，使对每条曲线的曲线积分相等，从而得到每个POI点对应的用户活跃程度数据。
  12. 根据权利要求9所述的基于电子地图的充电请求发起时间预测装置，其特征在于：
    所述分类子模块包括：
    地点分类单元，用于将用户充电请求发起地点分为N类，其中，N为大于等于2的整数；
    数据预设单元，用于为每类地点设置一组预设用户指数程度数据，并设置对应的相似度阈值；
    数据对比单元，用于将所述每组用户活跃程度数据与所述预设用户指数程度数据进行对比，得到相似度；
    数据分类单元，用于将所述相似度与对应的相似度阈值进行对比，若高于所述相似度阈值，则将该组用户活跃程度数据划分到对应的地点类别中。
  13. 根据权利要求9所述的基于电子地图的充电请求发起时间预测装置，其特征在于：
    所述优化子模块还用于，
    对每类的多组用户活跃程度数据进行聚类分析，得到每类地点的最终的充电请求时间分布数据。
  14. 根据权利要求8所述的基于电子地图的充电请求发起时间预测装置，其特征在于：
    所述每组指数数据采用表格或曲线的方式进行存储。
  15. 一种控制器，其包括存储器与处理器，所述存储器存储有计算机程序，所述程序在被所述处理器执行时能够实现权利要求1至7中任一项权利要求所述的方法的步骤。
  16. 一种计算机可读存储介质，用于存储计算机指令，所述指令在由一计算机或处理器执行时实现如权利要求1至7中任意一项权利要求所述的方法的步骤。
PCT/CN2018/100030 2017-11-28 2018-08-10 基于电子地图的充电请求发起时间预测方法和装置 WO2019105065A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP18883405.5A EP3719735A4 (en) 2017-11-28 2018-08-10 ELECTRONIC CARD BASED METHOD AND DEVICE FOR PREDICTING A CHARGE INQUIRY START TIME

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711216929.8A CN109840792B (zh) 2017-11-28 2017-11-28 基于电子地图的充电请求发起时间预测方法和装置
CN201711216929.8 2017-11-28

Publications (1)

Publication Number Publication Date
WO2019105065A1 true WO2019105065A1 (zh) 2019-06-06

Family

ID=66665207

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/100030 WO2019105065A1 (zh) 2017-11-28 2018-08-10 基于电子地图的充电请求发起时间预测方法和装置

Country Status (3)

Country Link
EP (1) EP3719735A4 (zh)
CN (1) CN109840792B (zh)
WO (1) WO2019105065A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113626396B (zh) * 2021-07-29 2024-03-12 北京达佳互联信息技术有限公司 一种聚类检测方法、装置、电子设备、存储介质及产品

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104778263A (zh) * 2015-04-23 2015-07-15 储盈新能源科技（上海）有限公司 一种电动车充电站系统仿真数据挖掘方法
CN106130110A (zh) * 2016-07-15 2016-11-16 华北电力大学 基于分层概率选择出行地的电动出租车充电站定容方法
CN107146013A (zh) * 2017-04-28 2017-09-08 国网北京市电力公司 一种基于灰色预测和支持向量机的分类型电动汽车需求时空分布动态预测方法
CN107316097A (zh) * 2017-05-18 2017-11-03 蔚来汽车有限公司 电动汽车充电需求预测方法及系统

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011164771A (ja) * 2010-02-05 2011-08-25 Motion:Kk 充電スタンドの運用管理サーバおよび充電スタンドの運用管理システム
CN103065199A (zh) * 2012-12-18 2013-04-24 广东电网公司电力科学研究院 电动汽车充电站负荷预测方法
US10121158B2 (en) * 2013-04-26 2018-11-06 General Motors Llc Optimizing vehicle recharging to limit use of electricity generated from non-renewable sources
US20150032516A1 (en) * 2013-07-24 2015-01-29 International Business Machines Corporation Managing electric vehicle (ev) charging station usage
CN105119743B (zh) * 2015-08-07 2017-05-10 百度在线网络技术（北京）有限公司 用户行为意图的获取方法及装置
CN106897919A (zh) * 2017-02-28 2017-06-27 百度在线网络技术（北京）有限公司 用车类型预测模型建立、信息提供方法及装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104778263A (zh) * 2015-04-23 2015-07-15 储盈新能源科技（上海）有限公司 一种电动车充电站系统仿真数据挖掘方法
CN106130110A (zh) * 2016-07-15 2016-11-16 华北电力大学 基于分层概率选择出行地的电动出租车充电站定容方法
CN107146013A (zh) * 2017-04-28 2017-09-08 国网北京市电力公司 一种基于灰色预测和支持向量机的分类型电动汽车需求时空分布动态预测方法
CN107316097A (zh) * 2017-05-18 2017-11-03 蔚来汽车有限公司 电动汽车充电需求预测方法及系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3719735A4 *

Also Published As

Publication number Publication date
CN109840792A (zh) 2019-06-04
EP3719735A4 (en) 2021-04-21
EP3719735A1 (en) 2020-10-07
CN109840792B (zh) 2021-06-29

Similar Documents

Publication Publication Date Title
Cheng et al. Prediction-based task assignment in spatial crowdsourcing
Chen et al. Spatiotemporal data model for network time geographic analysis in the era of big data
US7392247B2 (en) Method and apparatus for fusing context data
US20170068902A1 (en) Modeling of Geospatial Location Over Time
JP2019512764A (ja) ユーザ地理的ロケーションのタイプを識別するための方法および装置
EP3364309B1 (en) Account mapping method and device based on address information
Rao et al. Modeling electric taxis' charging behavior using real-world data
US20190130333A1 (en) Systems and methods for determining an optimal strategy
CN111831899B (zh) 导航兴趣点推荐方法、装置、服务器及可读存储介质
JP6787624B2 (ja) 音声指紋の照合のための方法、マシン可読媒体及びシステム
US11853259B2 (en) Indexing documents in a nested hierarchy of directories
CN110569321A (zh) 基于城市地图的网格划分处理方法、装置和计算机设备
CN107533676A (zh) 固定分配内容选择基础设施中的响应延迟减少
Cheng et al. FROG: A fast and reliable crowdsourcing framework
CN111639092A (zh) 一种人员流动分析方法、装置、电子设备及存储介质
WO2019105065A1 (zh) 基于电子地图的充电请求发起时间预测方法和装置
Nespoli et al. User behavior clustering based method for ev charging forecast
CN109213940B (zh) 大数据下实现用户位置计算的方法、存储介质、设备及系统
CN107103037B (zh) 一种地理区域的社会功能的识别方法与终端设备
Putri et al. A distributed system for finding high profit areas over big taxi trip data with MognoDB and spark
Wang et al. Reachability-driven influence maximization in time-dependent road-social networks
He et al. Multiple routes recommendation system on massive taxi trajectories
CN112528138A (zh) 社区活动信息准确推荐方法、终端设备及可读存储介质
CN116433053B (zh) 数据处理方法、装置、计算机设备和存储介质
CN105488194B (zh) 一种相似用户识别方法及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18883405

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2018883405

Country of ref document: EP

Effective date: 20200629