Method, device and storage medium for evaluating thermal power in a geographical area based on supply and demand relationship

Abstract

The present invention relates to a method, device and storage medium for evaluating the thermal power of a geographical area based on supply and demand relationships, and is applied to the field of online car-hailing technology, including: obtaining online car-hailing order data and driver operation service trajectory data for a period of history, and dividing the time dimension at the same time, extracting the driver order income amount, driver order service time and driver order listening time in each time dimension, obtaining the historical IPH under different time dimensions, and finally fusing the historical IPH and real-time IPH under the same time dimension according to the set weight; the present application combines the history of fine-grained time intervals, and according to the appropriate weight distribution, it can take into account the historical periodic laws of different spatial regions and different time periods, and can also consider the current real-time sudden changes, thereby solving the deficiency of the single dimension of traditional thermal calculation schemes; at the same time, the calculation of thermal level based on IPH index is guided by driver income, which is more in line with driver wishes and service goals.

Description

Technical Field

[0001] The present invention relates to the field of online car-hailing technology, and in particular to a method, device and storage medium for evaluating the thermal power of a geographical area based on supply and demand relationships. Background Art

[0002] Geographic area thermal assessment (heat map for short) refers to a computational technology that evaluates the hot and cold conditions of a certain type of business scenario through the size and distribution of business data. It usually uses a gradient from cold to warm colors through visualization tools to represent changes from low values ​​to high values. It can show the concentration trend or anomalies of data in geographic space distribution. Heat maps are widely used in various fields, including geographic information systems, bioinformatics, network traffic analysis, etc. For example, in the online car-hailing industry, heat maps are usually used to show the density of passenger demand for cars in a certain area of ​​the city, which can be used to evaluate order hotspots.

[0003] There are many heat map implementation solutions in the industry, but they usually focus on a certain business scenario, collect data from relevant geographic areas, such as population density, traffic flow, commodity sales volume, and any other data related to geographic location, and then clean and calculate the collected data to ensure the accuracy and consistency of the data. The processed result data should contain geographic location information (such as longitude and latitude) and the values ​​to be displayed, and finally choose a suitable visualization tool to build a heat map. The existing methods cannot adapt to complex and diverse business scenarios due to factors such as the single data dimension and the limitation of the geographic area division logic by the platform or tools. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a method, device and storage medium for evaluating the thermal power of a geographic area based on supply and demand relationships, so as to solve the problem that the existing methods cannot adapt to complex and diversified business scenarios due to factors such as a single data dimension and the geographical area division logic being restricted by platforms or tools.

[0005] According to a first aspect of an embodiment of the present invention, a method for evaluating thermal power in a geographical area based on a supply-demand relationship is provided, the method comprising:

[0006] A method for evaluating heat in a geographical area based on supply and demand relationships, characterized in that the method comprises:

[0007] Obtain historical N days and current real-time online car-hailing order data and driver service trajectory data;

[0008] Divide each day of the N historical days into X time periods, each of which has a length of Y. Number the X time periods in chronological order. Use the time periods with the same number in the N historical days as a time dimension, for a total of X time dimensions, each of which includes N time periods with the same number.

[0009] Based on the online ride-hailing order data of N days in history, the order time of all passenger orders is obtained every day, and the passenger orders are classified into time periods with different numbers every day according to the order time;

[0010] Based on the online car-hailing order data and driver service trajectory data for the past N days, obtain the driver's order income amount and the corresponding driver order service time for each passenger order in each time dimension, and obtain the driver's order listening time in each time period;

[0011] Replace the outliers on the driver's order income amount, driver's order service time, and driver's order listening time in each time period for each passenger order;

[0012] Based on the order income amount of the replaced driver, obtain the total order income amount in each time dimension; based on the order service time of the replaced driver, obtain the total driver order service time in each time dimension; based on the order listening time of the replaced driver, obtain the total order listening time in each time dimension;

[0013] Select any time dimension and two adjacent future time dimensions as a time window, traverse X time dimensions, and obtain X time windows;

[0014] Obtain the historical IPH in the corresponding time window dimension based on the total order revenue amount, total driver order service time, and total order listening time in the three time dimensions of each time window;

[0015] Obtaining the real-time IPH in the current time dimension according to the current real-time online car-hailing order data and the driver's service trajectory data;

[0016] Set the historical data weight and real-time data weight respectively, and obtain the final fused IPH under the corresponding time window dimension based on the historical IPH, real-time IPH, historical data weight and real-time data weight under the same time dimension;

[0017] Get the IPH of X time dimensions of any city in history P days, and select the largest IPH as the maximum IPH of the city; perform thermal rendering of the corresponding time window dimension of all honeycomb areas of the city based on the final fused IPH of the corresponding time window dimension of all honeycomb areas of the city and the maximum IPH of the city.

[0018] Preferably,

[0019] The replacing of outliers for the driver's order income amount, the driver's order service time, and the driver's order listening time in each time period for each passenger order includes:

[0020] Obtain the service IPH corresponding to each passenger order according to the driver order income amount and the driver order service time of each passenger order;

[0021] Based on the Three Sigma Principle, abnormal service IPHs and abnormal driver listening time are screened out, and the driver order income amount and driver order service time corresponding to the abnormal service IPH are replaced with the preset standard driver order income amount and the preset standard driver order service time, and the abnormal driver listening time is replaced with the preset standard driver listening time.

[0022] Preferably,

[0023] The three sigma-based screening of abnormal service IPHs and abnormal driver listening time includes:

[0024] Obtain the average IPH for each time dimension based on all service IPHs for each time dimension, and obtain the IPH standard deviation based on the average IPH;

[0025] Determine the IPH standard range according to the average IPH and the IPH standard deviation;

[0026] If any service IPH is not within the IPH standard range, the service IPH is an abnormal service IPH;

[0027] Obtain the average listening time for each time dimension based on the listening time of all drivers in each time dimension, and obtain the standard deviation of the listening time based on the average listening time.

[0028] Determine a standard range of listening time according to the average listening time and the standard deviation of the listening time;

[0029] If the time taken by any driver to listen to an order is not within the standard deviation of the time taken to listen to an order, the time taken by the driver to listen to an order is considered abnormal.

[0030] The preset standard driver order revenue amount, the preset standard driver order service time and the preset standard driver order listening time are respectively the average driver order revenue amount, the average driver order service time and the average order listening time in each time dimension.

[0031] Preferably,

[0032] The method of obtaining the historical IPH under the corresponding time window dimension according to the total order revenue amount, total driver order service time and total order listening time in the three time dimensions of each time window includes:

[0033] The time window order revenue amount is obtained based on the total order revenue amount of the three time dimensions in each time window, the time window driver order service time is obtained based on the total order listening time of the three time dimensions in each time window, and the time window order listening time is obtained based on the total order listening time of the three time dimensions in each time window;

[0034] The IPH of each time window is obtained based on the order revenue amount of the time window, the order service time of the driver in the time window, and the order listening time in the time window, and the IPH of each time window is used as the historical IPH under the corresponding time window dimension.

[0035] Preferably,

[0036] The method of obtaining the real-time IPH in the current time dimension according to the current real-time online car-hailing order data and the driver's service trajectory data includes:

[0037] Get the Z time periods before the current time, and the length of each time period is Y;

[0038] According to the current real-time online car-hailing order data and the driver's service trajectory data, the total driver order revenue, total service time and total driver listening time of the previous Z time periods are obtained. Based on the total driver order revenue, total service time and total driver listening time of the previous Z time periods, the real-time IPH under the time dimension corresponding to the previous Z time periods is obtained.

[0039] Preferably,

[0040] The thermal rendering of the corresponding time window dimension of all honeycomb areas of the city based on the final fused IPH under the corresponding time window dimension of all honeycomb areas of the city and the maximum IPH of the city includes:

[0041] Select any hive area in the city as the target hive area;

[0042] Obtain F honeycomb areas adjacent to the target honeycomb area, and respectively obtain the final fused IPH of the corresponding time window dimensions of the F honeycomb areas adjacent to the target honeycomb area;

[0043] Based on the final fused IPH of the corresponding time window dimension of the adjacent F honeycomb areas and the maximum IPH of the city, the thermal levels of the F honeycomb areas are obtained respectively, and the smallest one of the F thermal levels is selected as the thermal level of the target honeycomb area in the corresponding time window dimension. All the honeycomb areas in the city are traversed to obtain the thermal levels of all honeycomb areas in the corresponding time window dimension. According to the thermal levels of all the honeycomb areas in the city in the corresponding time window dimension, thermal rendering of the corresponding time window dimension is performed on the city map.

[0044] Preferably,

[0045] The acquisition of online car-hailing order data and driver service trajectory data for the past N days and the current day in real time includes:

[0046] The business system obtains the online car-hailing order data of the past N days and the current day's real-time data as well as the driver's service trajectory data, and stores them in TableStore and MySQL.

[0047] The data in MySQL is synchronized to the data channel DataHub through Alibaba Cloud's data transmission tool DTS, and the data in TableStore is synchronized to the data channel DataHub through real-time computing Flink;

[0048] The data in the data channel DataHub is preprocessed through Flink jobs, and the preprocessed online car-hailing order data and driver service trajectory data are aggregated and written into the data warehouse MaxCompute for subsequent calculations.

[0049] Preferably, it also includes:

[0050] The thermal level data of all honeycomb areas in the city is periodically written to Redis and TableStore at the same time through the DataWorks data integration tool. Redis Geo and TableStore indexes are used to provide honeycomb area selection and honeycomb area thermal level data query APIs for downstream business use.

[0051] The driver side and other business systems query the thermal level of the honeycomb area through the data center API for display or functional application.

[0052] According to a second aspect of an embodiment of the present invention, there is provided a device for evaluating the thermal power of a geographical area based on a supply-demand relationship, the device comprising:

[0053] Data acquisition module: used to obtain historical N days and current real-time online car-hailing order data and driver service trajectory data;

[0054] Time dimension division module: used to divide each day of the historical N days into X time periods evenly, with the length of each time period being Y, and numbering the X time periods in chronological order; taking the time periods with the same number in the historical N days as a time dimension, for a total of X time dimensions, each of which includes N time periods with the same number;

[0055] Order classification module: used to obtain the order time of all passenger orders every day based on the online car-hailing order data of the past N days, and classify the passenger orders into time periods with different numbers every day according to the order time;

[0056] Dimension data acquisition module: used to obtain the driver's order income amount and the corresponding driver order service time of each passenger order in each time dimension based on the online car-hailing order data and driver operation service trajectory data of the past N days, and obtain the driver's order listening time in each time period;

[0057] An outlier replacement module: used to replace outliers for the driver's order income amount, driver's order service time, and driver's order listening time in each time period for each passenger order;

[0058] Dimension average data acquisition module: used to obtain the total order revenue amount of each time dimension based on the order revenue amount of the replaced driver, obtain the total driver order service time of each time dimension based on the order service time of the replaced driver; obtain the total order listening time of each time dimension based on the order listening time of the replaced driver;

[0059] Time window partitioning module: used to select any time dimension and two adjacent future time dimensions as a time window, traverse X time dimensions, and obtain X time windows;

[0060] Historical IPH acquisition module: used to obtain the historical IPH in the corresponding time window dimension based on the total order revenue amount, total driver order service time and total order listening time in the three time dimensions of each time window;

[0061] Real-time IPH acquisition module: used to obtain the real-time IPH in the current time dimension according to the current real-time online car-hailing order data and the driver's service trajectory data;

[0062] Fusion IPH acquisition module: used to set the historical data weight and real-time data weight respectively, and obtain the final fusion IPH under the corresponding time window dimension based on the historical IPH, real-time IPH, historical data weight and real-time data weight under the same time dimension;

[0063] Honeycomb thermal acquisition module: used to obtain the IPH of X time dimensions of any city in the history of P days, select the largest IPH as the maximum IPH of the city; based on the final fused IPH of the corresponding time window dimension of all honeycomb areas of the city and the maximum IPH of the city, perform thermal rendering of the corresponding time window dimension of all honeycomb areas of the city.

[0064] According to a third aspect of an embodiment of the present invention, there is provided a storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a main controller, each step in the above method is implemented.

[0065] The technical solution provided by the embodiments of the present invention may have the following beneficial effects:

[0066] This application obtains online car-hailing order data and driver service trajectory data for a period of history, and divides the time dimension at the same time. By extracting the driver's order income amount, driver order service time and driver listening time in each time dimension, the historical IPH in different time dimensions is obtained, and finally the historical IPH in the same time dimension is merged with the real-time IPH according to the set weight; this application combines data such as time, space, drivers, channels, prices and external objective factors to comprehensively evaluate the relationship between vehicle demand and driver capacity supply in the region, which is more in line with the hot and cold conditions of actual scenarios, and avoids the situation of many orders and many cars, few orders and no cars, and many orders generated by traditional thermal algorithms. It can effectively improve the fulfillment rate of passengers' car needs and reduce the situation of online car-hailing drivers driving idle, and at the same time, it combines historical and real-time honeycomb statistical data of fine-grained time intervals. According to the appropriate weight distribution, it can take into account the historical periodic laws of different spatial regions and different time periods, as well as the current real-time sudden changes, which solves the deficiency of the single dimension of traditional thermal calculation schemes. The thermal level calculated based on the IPH indicator is oriented towards driver income, which is more in line with the driver's wishes and service goals. Compared with other platforms based on order heat, it can more effectively improve the driver's experience and income, and avoid the scenario of many orders and low income.

[0067] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0069] Figure 1 is a flow chart of a method for evaluating heat in a geographical area based on supply and demand relationship according to an exemplary embodiment;

[0070] Figure 2 is a schematic diagram of a business data collection process according to another exemplary embodiment;

[0071] Figure 3 is a schematic diagram showing the thermal level division of target honeycomb areas according to another exemplary embodiment;

[0072] Figure 4 is a schematic diagram showing data integration according to another exemplary embodiment;

[0073] Figure 5 is a normal distribution diagram of the three sigma law according to another exemplary embodiment;

[0074] Figure 6is a system schematic diagram of an apparatus for evaluating thermal power in a geographical area based on supply and demand relationship according to another exemplary embodiment;

[0075] In the attached figure: 1-data acquisition module, 2-time dimension division module, 3-order classification module, 4-dimensional data acquisition module, 5-outlier replacement module, 6-dimensional average data acquisition module, 7-time window division module, 8-historical IPH acquisition module, 9-real-time IPH acquisition module, 10-fused IPH acquisition module, 11-honeycomb thermal acquisition module. DETAILED DESCRIPTION

[0076] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0077] Embodiment 1

[0078] Figure 1 is a flow chart of a method for evaluating the heat of a geographical area based on supply and demand relationship according to an exemplary embodiment. Figure 1 As shown, the method includes:

[0079] S1, obtain the historical N days and current real-time online car-hailing order data and driver service trajectory data;

[0080] S2, divide each day of the N historical days into X time periods, each time period is Y long, and number the X time periods in order; take the time periods with the same number in the N historical days as a time dimension, a total of X time dimensions, each time dimension includes N time periods with the same number;

[0081] S3, based on the online car-hailing order data of the past N days, obtain the order time of all passenger orders every day, and classify the passenger orders into time periods with different numbers every day according to the order time;

[0082] S4, based on the online car-hailing order data and driver service trajectory data of the past N days, obtain the driver's order income amount and the corresponding driver order service time of each passenger order in each time dimension, and obtain the driver's order listening time in each time period;

[0083] S5, replacing outliers on the driver's order income amount, driver's order service time, and driver's order listening time in each time period for each passenger order;

[0084] S6, based on the order income amount of the replaced driver, obtain the total order income amount of each time dimension, based on the order service time of the replaced driver, obtain the total driver order service time of each time dimension; based on the order listening time of the replaced driver, obtain the total order listening time of each time dimension;

[0085] S7, select any time dimension and two adjacent future time dimensions as a time window, traverse X time dimensions, and obtain X time windows;

[0086] S8, obtaining the historical IPH in the corresponding time window dimension according to the total order revenue amount, total driver order service time and total order listening time in the three time dimensions of each time window;

[0087] S9, obtaining the real-time IPH in the current time dimension according to the current real-time online car-hailing order data and the driver's service trajectory data;

[0088] S10, respectively set the historical data weight and the real-time data weight, and obtain the final fused IPH under the corresponding time window dimension based on the historical IPH, real-time IPH, historical data weight and real-time data weight under the same time dimension;

[0089] S11, obtaining the IPH of X time dimensions of any city in history P days, selecting the largest IPH as the maximum IPH of the city; performing thermal rendering of the corresponding time window dimension of all honeycomb areas of the city based on the final fused IPH of the corresponding time window dimension of all honeycomb areas of the city and the maximum IPH of the city;

[0090] It is understandable that if Figure 2 As shown in the figure, this application mainly relies on detailed data of passenger car order and data of urban driver service trajectory. At present, the business system stores order data and driver data through TableStore (table storage) and MySQL (relational database); then synchronize the data of the business database table to the data channel DataHub (metadata management platform) through Alibaba Cloud data transmission tool DTS and real-time computing FLink, and then perform cleaning, association, disassembly, aggregation and other operations through Flink jobs, and finally aggregate the order data and driver trajectory data into the data warehouse MaxCompute for downstream computing;

[0091] This application collects historical order and driver data for N days (usually 4 weeks of data is required) and real-time order and driver data for the current day. After being collected and entered into the data warehouse MaxCompute by the above process, the key fields of the detailed data (city, coordinates, time, date type, order to driver amount, driver order service start time, driver order service end time, driver dispatch order status, driver coordinate reporting time, etc.) are parsed and converted;

[0092] Each day is divided into X time periods on average. The length of each time period is Y, usually 10 minutes. Therefore, a day can be divided into 144 time periods, that is, X is 144, and the X time periods are numbered in order. The time periods with the same number in the history of N days (4 weeks, that is, 28 days) are used as a time dimension, with a total of X time dimensions. Each time dimension includes N time periods with the same number, that is, there are 144 time dimensions, and each time dimension has 28 time periods with the same number.

[0093] Then get the order time of all passenger orders in these 28 days t 0, according to order time t 0 All passenger orders are assigned to the corresponding time period, and the expression formula is:

[0094] floor((HOUR(t 0 )*60+MINUTE(t 0 )) / 10+1)

[0095] Get the passenger orders in each time dimension, and for each passenger order, get the corresponding driver order income amount: a , driver order service start time: t 1 and the driver's order service end time: t 1. According to t 1 and t 2. Get the driver's order service time ST;

[0096] When the driver does not pick up passengers, that is, does not execute any passenger orders, a coordinate point is sent every 30 seconds. The driver's order waiting time LT is obtained based on all the coordinate points. The order waiting time LT can also be understood as the driver's waiting time, expressed as: COUNT(1) * 30s ;

[0097] Based on the driver's order income a And the driver's order service time ST obtains the service IPH of each order (the driver's order income amount a / driver order service time ST);

[0098] It is worth noting that IPH represents the hourly income based on the unit dimension, which is generally used for business evaluation of revenue.

[0099] For the corresponding service IPH of all passenger orders in a time dimension, the maximum IPHMax and the minimum IPHMin are calculated based on the three sigma law. The maximum IPHMax = m +3 s , minimum value IPHMin = m -3 s ,in, m Represents the average IPH of all services in a time dimension. s represents the standard deviation. It should be noted that, as shown in the attached Figure 5 As shown, in statistics, the 68–95–99.7 principle is the percentage of people in a normal distribution that are within one standard deviation, two standard deviations, and three standard deviations of the mean. More precise numbers are 68.27%, 95.45%, and 99.73%, which is 3 s Guidelines:

[0100] The values ​​are distributed in ( m - s , m+s ) has a probability of 0.6827;

[0101] The values ​​are distributed in ( m- 2 s , m+ 2 s ) has a probability of 0.9545;

[0102] The values ​​are distributed in ( m- 3 s , m+ 3 s ) has a probability of 0.9973;

[0103] That is, the final standard range is ( m- 3 s , m+ 3 s ), after determining the standard range, all service IPHs within a time dimension are judged. If they are not within the standard range, the service IPH is an outlier. a And the driver order service time ST is obtained, so for abnormal service IPH, the corresponding driver order income amount a The driver order service time ST is replaced with a preset standard value. The preset standard value can be the driver order income amount within a time dimension. a And the average value of driver order service time ST, expressed as:

[0104] IF IPH>IPH Max or IPH <IPH Min THEN a = AVG(a), ST = AVG(ST)

[0105] Similarly, for the driver's listening time LT within a time dimension, the maximum value LTMax and the minimum value LTMin are obtained through the three sigma law, and then the annotation range of the driver's listening time LT is determined. For the driver's listening time LT that is not within the standard range, it is replaced with the average value of the driver's listening time LT within a time dimension. The expression is:

[0106] IF LT>LTMax or LT <LTMin THEN LT = AVG(LT)

[0107] Based on the driver's order revenue amount after replacement a , the driver's order service time ST and the driver's order listening time LT, to obtain the total revenue amount of all orders within a time dimension, the total service time of all driver orders and the total order listening time;

[0108] Considering the ductility of time slices, the data of the displacement of 2 unit slices of the current time dimension slice is associated, that is, a 30-minute window is slid every 10 minutes for calculation. For example, the current time slice 1 (00:00-00:10) needs to be associated with the data of time slice 2 (00:10-00:20) and time slice 3 (00:20-00:30) to generate a time window. As mentioned above, there are a total of 144 time dimensions, numbered 1-144. For the time dimension with sequence number 1, its time window is: sequence number 1-sequence number 3, for the time dimension with sequence number 2, its time window is: sequence number 2-sequence number 4... For the time dimension with sequence number 143, its time window is: sequence number 143, sequence number 144 and sequence number 1, and for the last time dimension with sequence number 144, its time window is: sequence number 144, sequence number 1 and sequence number 2, so that 144 time windows can be obtained;

[0109] For each time window, based on the total order revenue amount, the average total driver order service time and the total order listening time obtained from the three time dimensions, the driver order revenue amount, driver order service time and driver dispatching order listening time of each time window are obtained. At the same time, the historical IPH under the corresponding time window dimension is obtained according to the driver order revenue amount, driver order service time and driver dispatching order listening time of each time window. The specific expression formula is:

[0110] Driver order revenue amount during time window: a ' = a 0+ a 1+ a 2

[0111] Time window driver order service time: ST' = ST0+ ST1+ ST2

[0112] The time window driver takes to dispatch the vehicle: LT' = LT0 + LT1 + LT2

[0113] Time Window IPH: a ' / (ST' + LT');

[0114] To obtain the current real-time IPH, specifically: starting from the current time, go back 3 time periods, each time period is 10 minutes, that is, go back 30 minutes, obtain the total driver order income amount, total driver order service time and total order listening time of all orders within these 30 minutes, and obtain the real-time IPH of these 30 minutes according to the above time window IPH calculation formula;

[0115] For the fusion of historical IPH and real-time IPH, first determine the time dimension to which the 30 minutes of real-time IPH belongs. Since 30 minutes includes the above three time dimensions, and the time window is divided into three consecutive time dimensions as one time window, there are a total of 144 time windows. In short, these 30 minutes must correspond to a time window, and the historical IPH of the historical time window corresponding to these 30 minutes is obtained;

[0116] Setting historical data weight x and real-time data weight y , x + y = 1;

[0117] According to historical data weight x and real-time data weight y The historical IPH and real-time IPH obtained above under the same time dimension are fused to obtain the final fused IPH under the corresponding time window dimension, which is expressed as:

[0118] Final Fusion IPH = Historical IPH* x + Real-time IPH * y

[0119] After obtaining the calculation method of the final fused IPH, the IPH of each time dimension of any city in the 7-day history is obtained. Similarly, according to the above method, each day of the 7 days is divided into 144 time periods, and each time period is numbered 1-144. The time periods with the same number in the 7 days are divided into the same time dimension, and 144 time dimensions are obtained. Finally, the IPH of these 144 time dimensions is obtained. The specific expression is: c_IPH = sum(a) / sum(ST + LT)

[0120] Get the maximum IPH value of the city's 7-day all-time slice history, that is, select the largest IPH from the 7-day IPH of each time dimension of the city's history as the city's maximum IPH. The expression is: maxIPH = max(c_ IPH);

[0121] Then, the final fused IPH of all honeycomb areas in any city is obtained, and the thermal level of each honeycomb area is determined by combining the maximum IPH value of the city. The thermal level determination formula is:

[0122] Level 1: 0 <= IPH <= max IPH * 0.25

[0123] Level 2: 0.25*max IPH <IPH<= max IPH * 0.50

[0124] Level 3: 0.50*max IPH <IPH<= max IPH * 0.75

[0125] Level 4: 0.75*max IPH <IPH;

[0126] Among them, level 1 and level 2 are represented as cold areas on the map rendering, and level 3 and level 4 are represented as hot areas on the map rendering;

[0127] It is worth noting that since the honeycomb strictly divides the geographical area into grids according to the H3 algorithm, there is a statistical possibility that there will be a large difference in hot and cold in adjacent areas. However, the actual online car-hailing business scenario is flexible and dynamic in terms of the distance between the passenger placing an order and the driver picking up the car. Therefore, from a business perspective, the hot and cold of adjacent areas should be gradual. Therefore, this embodiment also introduces the calculation logic of "hot and cold diffusion" to solve the problems of "hot covering cold" and "cold covering hot". Specifically:

[0128] For the determination of the thermal level of the target honeycomb area, as shown in the attached Figure 3 As shown, the six honeycomb areas adjacent to the target honeycomb area are obtained, and the thermal levels of the six honeycomb areas are obtained according to the above method respectively, and the smallest one is selected as the thermal level of the target honeycomb area. The specific expression is as follows:

[0129] Minimum surrounding honeycomb heat: hMin = min(h1, h2, h3, h4, h5, h6)

[0130] Final thermal value: h = if (h0 <hMin, hMin, h0);

[0131] It is worth emphasizing that, based on the above method, this application can display the thermal level of any honeycomb area in any city on weekdays or weekends in multiple dimensions;

[0132] As attached Figure 4As shown in the figure, the honeycomb thermal data of the above calculation results are periodically written to Redis and TableStore at the same time through the DataWorks data integration tool; the data center provides honeycomb selection and honeycomb thermal data query APIs through Redis Geo and TableStore indexes for use by downstream businesses; the driver-side heat map function and other business systems query the honeycomb thermal data through the data center API for display or functional application.

[0133] Embodiment 2

[0134] Figure 6 is a system schematic diagram of a device for evaluating the thermal power of a geographical area based on a supply-demand relationship according to another exemplary embodiment, the device comprising:

[0135] Data acquisition module 1: used to obtain historical N days and current real-time online car-hailing order data and driver service trajectory data;

[0136] Time dimension division module 2: used to divide each day of the historical N days into X time periods evenly, with the length of each time period being Y, and numbering the X time periods in chronological order; taking the time periods with the same number in the historical N days as a time dimension, a total of X time dimensions, each of which includes N time periods with the same number;

[0137] Order classification module 3: used to obtain the order time of all passenger orders every day based on the online car-hailing order data of the past N days, and classify the passenger orders into time periods with different numbers every day according to the order time;

[0138] Dimension data acquisition module 4: used to obtain the driver order income amount and the corresponding driver order service time of each passenger order in each time dimension according to the online car-hailing order data and driver operation service trajectory data of the past N days, and obtain the driver's order listening time in each time period;

[0139] Outlier replacement module 5: used to replace outliers for the driver's order income amount, driver's order service time, and driver's order listening time in each time period for each passenger order;

[0140] Dimension average data acquisition module 6: used to obtain the total order revenue amount of each time dimension based on the order revenue amount of the replaced driver, obtain the total driver order service time of each time dimension based on the order service time of the replaced driver; obtain the total order listening time of each time dimension based on the order listening time of the replaced driver;

[0141] Time window division module 7: used to select any time dimension and two adjacent future time dimensions as a time window, traverse X time dimensions, and obtain X time windows;

[0142] Historical IPH acquisition module 8: used to obtain the historical IPH under the corresponding time window dimension according to the total order revenue amount, total driver order service time and total order listening time in the three time dimensions of each time window;

[0143] Real-time IPH acquisition module 9: used to acquire the real-time IPH in the current time dimension according to the current real-time online car-hailing order data and the driver's service trajectory data;

[0144] Fusion IPH acquisition module 10: used to set the historical data weight and the real-time data weight respectively, and obtain the final fusion IPH under the corresponding time window dimension based on the historical IPH, real-time IPH, historical data weight and real-time data weight under the same time dimension;

[0145] Honeycomb thermal acquisition module 11: used to obtain the IPH of X time dimensions of any city in the history of P days, select the largest IPH as the maximum IPH of the city; based on the final fused IPH of the corresponding time window dimension of all honeycomb areas of the city and the maximum IPH of the city, perform thermal rendering of the corresponding time window dimension of all honeycomb areas of the city.

[0146] Embodiment three:

[0147] This embodiment provides a storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a main controller, each step in the above method is implemented;

[0148] It is understandable that the storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0149] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0150] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0151] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0152] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0153] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0154] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0155] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0156] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0157] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for evaluating the thermal power of a geographical area based on the relationship between supply and demand, characterized in that: The method comprises: Obtain historical N days and current real-time online car-hailing order data and driver service trajectory data; Divide each day of the N historical days into X time periods, each of which has a length of Y. Number the X time periods in chronological order. Use the time periods with the same number in the N historical days as a time dimension, for a total of X time dimensions, each of which includes N time periods with the same number. Based on the online ride-hailing order data of N days in history, the order time of all passenger orders is obtained every day, and the passenger orders are classified into time periods with different numbers every day according to the order time; Based on the online car-hailing order data and driver service trajectory data for the past N days, obtain the driver's order income amount and the corresponding driver order service time for each passenger order in each time dimension, and obtain the driver's order listening time in each time period; Replace the outliers on the driver's order income amount, driver's order service time, and driver's order listening time in each time period for each passenger order; Based on the order income amount of the replaced driver, obtain the total order income amount in each time dimension; based on the order service time of the replaced driver, obtain the total driver order service time in each time dimension; based on the order listening time of the replaced driver, obtain the total order listening time in each time dimension; Select any time dimension and two adjacent future time dimensions as a time window, traverse X time dimensions, and obtain X time windows; Obtain the historical IPH in the corresponding time window dimension based on the total order revenue amount, total driver order service time, and total order listening time in the three time dimensions of each time window; Obtaining the real-time IPH in the current time dimension according to the current real-time online car-hailing order data and the driver's service trajectory data; Set the historical data weight and real-time data weight respectively, and obtain the final fused IPH under the corresponding time window dimension based on the historical IPH, real-time IPH, historical data weight and real-time data weight under the same time dimension; Get the IPH of X time dimensions of any city in history P days, and select the largest IPH as the maximum IPH of the city; perform thermal rendering of all honeycomb areas in the city in the corresponding time window dimension based on the final fused IPH of all honeycomb areas in the city and the maximum IPH of the city; The IPH represents the hourly income based on the unit dimension.

2. The method according to claim 1, characterized in that The replacing of outliers for the driver's order income amount, the driver's order service time, and the driver's order listening time in each time period for each passenger order includes: Obtain the service IPH corresponding to each passenger order according to the driver order income amount and the driver order service time of each passenger order; Based on the Three Sigma Principle, abnormal service IPHs and abnormal driver listening time are screened out, and the driver order income amount and driver order service time corresponding to the abnormal service IPH are replaced with the preset standard driver order income amount and the preset standard driver order service time, and the abnormal driver listening time is replaced with the preset standard driver listening time.

3. The method according to claim 2, characterized in that The three sigma-based screening of abnormal service IPHs and abnormal driver listening time includes: Obtain the average IPH for each time dimension based on all service IPHs for each time dimension, and obtain the IPH standard deviation based on the average IPH; Determine the IPH standard range according to the average IPH and the IPH standard deviation; If any service IPH is not within the IPH standard range, the service IPH is an abnormal service IPH; Obtain the average listening time for each time dimension based on the listening time of all drivers in each time dimension, and obtain the standard deviation of the listening time based on the average listening time. Determine a standard range of listening time according to the average listening time and the standard deviation of the listening time; If the time taken by any driver to listen to an order is not within the standard deviation of the time taken to listen to an order, the time taken by the driver to listen to an order is considered abnormal. The preset standard driver order revenue amount, the preset standard driver order service time and the preset standard driver order listening time are respectively the average driver order revenue amount, the average driver order service time and the average order listening time in each time dimension.

4. The method according to claim 3, characterized in that The method of obtaining the historical IPH under the corresponding time window dimension according to the total order revenue amount, total driver order service time and total order listening time in the three time dimensions of each time window includes: The time window order revenue amount is obtained based on the total order revenue amount of the three time dimensions in each time window, the time window driver order service time is obtained based on the total order listening time of the three time dimensions in each time window, and the time window order listening time is obtained based on the total order listening time of the three time dimensions in each time window; The IPH of each time window is obtained based on the order revenue amount of the time window, the order service time of the driver in the time window, and the order listening time in the time window, and the IPH of each time window is used as the historical IPH under the corresponding time window dimension.

5. The method according to claim 4, characterized in that The method of obtaining the real-time IPH in the current time dimension according to the current real-time online car-hailing order data and the driver's service trajectory data includes: Get the Z time periods before the current time, and the length of each time period is Y; According to the current real-time online car-hailing order data and the driver's service trajectory data, the total driver order revenue, total service time and total driver listening time of the previous Z time periods are obtained. Based on the total driver order revenue, total service time and total driver listening time of the previous Z time periods, the real-time IPH under the time dimension corresponding to the previous Z time periods is obtained.

6. The method according to claim 2, characterized in that The thermal rendering of all honeycomb areas of the city in the corresponding time window dimension based on the final fused IPH in the corresponding time window dimension of all honeycomb areas of the city and the maximum IPH of the city includes: Select any hive area in the city as the target hive area; Obtain F honeycomb areas adjacent to the target honeycomb area, and respectively obtain the final fused IPH of the corresponding time window dimensions of the F honeycomb areas adjacent to the target honeycomb area; Based on the final fused IPH of the corresponding time window dimension of the adjacent F honeycomb areas and the maximum IPH of the city, the thermal levels of the F honeycomb areas are obtained respectively, and the smallest one of the F thermal levels is selected as the thermal level of the target honeycomb area in the corresponding time window dimension. All the honeycomb areas in the city are traversed to obtain the thermal levels of all honeycomb areas in the corresponding time window dimension. According to the thermal levels of all the honeycomb areas in the city in the corresponding time window dimension, thermal rendering of the corresponding time window dimension is performed on the city map.

7. The method according to claim 1, characterized in that The acquisition of online car-hailing order data and driver service trajectory data for the past N days and the current day in real time includes: The business system obtains the online car-hailing order data of the past N days and the current day's real-time data as well as the driver's service trajectory data, and stores them in TableStore and MySQL. The data in MySQL is synchronized to the data channel DataHub through Alibaba Cloud's data transmission tool DTS, and the data in TableStore is synchronized to the data channel DataHub through real-time computing Flink; The data in the data channel DataHub is preprocessed through Flink jobs, and the preprocessed online car-hailing order data and driver service trajectory data are aggregated and written into the data warehouse MaxCompute for subsequent calculations.

8. The method according to claim 6, characterized in that Also includes: The thermal level data of all honeycomb areas in the city is periodically written to Redis and TableStore at the same time through the DataWorks data integration tool. RedisGeo and TableStore indexes are used to provide honeycomb area selection and honeycomb area thermal level data query APIs for downstream business use. The driver side and other business systems query the thermal level of the honeycomb area through the data center API for display or functional application.

9. A device for evaluating the thermal capacity of a geographical area based on supply and demand, characterized in that: The device comprises: Data acquisition module: used to obtain historical N days and current real-time online car-hailing order data and driver service trajectory data; Time dimension division module: used to divide each day of the historical N days into X time periods evenly, with the length of each time period being Y, and numbering the X time periods in chronological order; taking the time periods with the same number in the historical N days as a time dimension, for a total of X time dimensions, each of which includes N time periods with the same number; Order classification module: used to obtain the order time of all passenger orders every day based on the online car-hailing order data of the past N days, and classify the passenger orders into time periods with different numbers every day according to the order time; Dimension data acquisition module: used to obtain the driver's order income amount and the corresponding driver order service time of each passenger order in each time dimension based on the online car-hailing order data and driver operation service trajectory data of the past N days, and obtain the driver's order listening time in each time period; An outlier replacement module: used to replace outliers for the driver's order income amount, driver's order service time, and driver's order listening time in each time period for each passenger order; Dimension average data acquisition module: used to obtain the total order revenue amount of each time dimension based on the order revenue amount of the replaced driver, obtain the total driver order service time of each time dimension based on the order service time of the replaced driver; obtain the total order listening time of each time dimension based on the order listening time of the replaced driver; Time window partitioning module: used to select any time dimension and two adjacent future time dimensions as a time window, traverse X time dimensions, and obtain X time windows; Historical IPH acquisition module: used to obtain the historical IPH in the corresponding time window dimension based on the total order revenue amount, total driver order service time and total order listening time in the three time dimensions of each time window; Real-time IPH acquisition module: used to obtain the real-time IPH in the current time dimension according to the current real-time online car-hailing order data and the driver's service trajectory data; Fusion IPH acquisition module: used to set the historical data weight and real-time data weight respectively, and obtain the final fusion IPH under the corresponding time window dimension based on the historical IPH, real-time IPH, historical data weight and real-time data weight under the same time dimension; Honeycomb thermal acquisition module: used to obtain the IPH of X time dimensions of any city in history P days, select the largest IPH as the city's maximum IPH; perform thermal rendering of all honeycomb areas in the city in the corresponding time window dimension based on the final fused IPH in the corresponding time window dimension of all honeycomb areas in the city and the city's maximum IPH; The IPH represents the hourly income based on the unit dimension.

10. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the main controller, the various steps in the method for evaluating the thermal power of a geographical area based on the supply and demand relationship as described in any one of claims 1 to 8 are implemented.

Images