CN112101671B - Region attribute prediction method and device, readable storage medium and electronic equipment - Google Patents
Region attribute prediction method and device, readable storage medium and electronic equipment Download PDFInfo
- Publication number
- CN112101671B CN112101671B CN202010998050.9A CN202010998050A CN112101671B CN 112101671 B CN112101671 B CN 112101671B CN 202010998050 A CN202010998050 A CN 202010998050A CN 112101671 B CN112101671 B CN 112101671B
- Authority
- CN
- China
- Prior art keywords
- demand
- supply
- vector
- sequence
- time slice
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 239000013598 vector Substances 0.000 claims abstract description 275
- 238000000605 extraction Methods 0.000 claims abstract description 35
- 230000000737 periodic effect Effects 0.000 claims abstract description 19
- 230000015654 memory Effects 0.000 claims description 48
- 230000006870 function Effects 0.000 claims description 24
- 238000004590 computer program Methods 0.000 claims description 12
- 238000010606 normalization Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 abstract description 22
- 238000004364 calculation method Methods 0.000 description 14
- 238000010586 diagram Methods 0.000 description 11
- 230000007613 environmental effect Effects 0.000 description 4
- 238000012804 iterative process Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
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
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- 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
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Theoretical Computer Science (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention discloses a region attribute prediction method, a region attribute prediction device, a readable storage medium and electronic equipment. The method comprises the steps of determining a target area comprising a plurality of sub-areas, and determining a supply vector sequence and a demand vector sequence corresponding to the target area according to a preset periodic time slice sequence. And determining a supply characteristic vector sequence and a demand characteristic vector sequence according to the supply vector sequence and the demand vector sequence, inputting each characteristic vector sequence and a preset external environment vector into a linked prediction model obtained by pre-training, determining a supply predicted value range and a demand predicted value range corresponding to each sub-region in a target time slice, and realizing vehicle scheduling of each sub-region. According to the embodiment of the invention, when the predicted value is determined, the characteristic extraction process and the prediction process are carried out separately, and the extracted supply characteristic and the extracted demand characteristic are input into the linkage prediction model together to obtain the corresponding supply predicted value range and demand predicted value range, so that the accuracy of the prediction result is improved.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a method and an apparatus for predicting a region attribute, a readable storage medium, and an electronic device.
Background
The current network car booking field is rapidly developed, and the life of people is greatly facilitated. The supply amount of online booking cars at different times and the demand amount of online booking cars of users for different areas in a city can be interfered by the positions, weather, environment and the like of the areas. In order to ensure that the vehicle supply quantity of different areas can meet the vehicle demand quantity and avoid that the vehicle booking requirement of a user cannot be met or the waiting time of a vehicle booking order is too long, the supply quantity and the demand quantity of the areas at the target time need to be predicted in advance so as to carry out vehicle dispatching on the different areas according to the prediction result.
Disclosure of Invention
In view of the above, the embodiments of the present invention provide a region attribute prediction method, a device, a readable storage medium and an electronic device, which aim to determine a supply predicted value range and a demand predicted value range of each sub-region in a target region of a target time slice for vehicle scheduling.
In a first aspect, an embodiment of the present invention provides a method for predicting a region attribute, where the method includes:
determining a target area, wherein the target area comprises a plurality of sub-areas;
determining a supply vector sequence corresponding to the target area according to a preset periodic time slice sequence, wherein the supply vector sequence comprises a plurality of supply vectors corresponding to each time slice in the time slice sequence and is used for representing the vehicle supply quantity of each sub-area;
determining a demand vector sequence corresponding to the target area according to the time slice sequence, wherein the demand vector sequence comprises a plurality of demand vectors corresponding to each time slice in the time slice sequence and is used for representing the vehicle demand of each sub-area;
respectively determining a corresponding supply characteristic vector sequence and a corresponding demand characteristic vector sequence according to the supply vector sequence and the demand vector sequence;
and inputting the supply characteristic vector sequence, the demand characteristic vector sequence and a preset external environment vector into a linkage prediction model obtained by pre-training so as to determine a supply predicted value range and a demand predicted value range of each sub-region corresponding to a target time slice, wherein the interval duration between the target time slice and the last time slice in the time slice sequence is one period of the time slice sequence.
In a second aspect, an embodiment of the present invention provides an area attribute prediction apparatus, where the apparatus includes:
the device comprises a region determining module, a judging module and a judging module, wherein the region determining module is used for determining a target region, and the target region comprises a plurality of sub-regions;
the first sequence determining module is used for determining a supply vector sequence corresponding to the target region according to a preset periodic time slice sequence, wherein the supply vector sequence comprises a plurality of supply vectors corresponding to time slices in the time slice sequence and is used for representing the vehicle supply quantity of each sub-region;
the second sequence determining module is used for determining a demand vector sequence corresponding to the target area according to the time slice sequence, wherein the demand vector sequence comprises a plurality of demand vectors corresponding to each time slice in the time slice sequence and is used for representing the vehicle demand of each sub-area;
the characteristic determining module is used for respectively determining a corresponding supply characteristic vector sequence and a corresponding demand characteristic vector sequence according to the supply vector sequence and the demand vector sequence;
and the prediction module is used for inputting the supply characteristic vector sequence, the demand characteristic vector sequence and a preset external environment vector into a linkage prediction model obtained by pre-training so as to determine a supply prediction value range and a demand prediction value range of each sub-region corresponding to a target time slice, wherein the interval duration between the target time slice and the last time slice in the time slice sequence is one period of the time slice sequence.
In a third aspect, an embodiment of the present invention provides a computer-readable storage medium for storing computer program instructions, which when executed by a processor implement the method according to the first aspect.
In a fourth aspect, an embodiment of the present invention provides an electronic device, including a memory and a processor, the memory being configured to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method according to the first aspect.
According to the embodiment of the invention, the target area comprising a plurality of sub-areas is determined, and the supply vector sequence and the demand vector sequence corresponding to the target area are determined according to the preset periodic time slice sequence. And respectively determining a corresponding supply characteristic vector sequence and a corresponding demand characteristic vector sequence according to the supply vector sequence and the demand vector sequence, and inputting the supply characteristic vector sequence, the demand characteristic vector sequence and a preset external environment vector into a linked prediction model obtained by pre-training so as to determine a supply predicted value range and a demand predicted value range corresponding to each sub-region in a target time slice, thereby realizing vehicle scheduling of each sub-region. According to the embodiment of the invention, when the predicted value is determined, the characteristic extraction process and the prediction process are carried out separately, and the extracted supply characteristic and the extracted demand characteristic are input into the linkage prediction model together to obtain the corresponding supply predicted value range and demand predicted value range, so that the accuracy of the prediction result is improved. And carrying out vehicle scheduling on each sub-area according to the corresponding supply predicted value range and demand predicted value range obtained by prediction, so that the vehicle using requirements of target time slice users are met, and the vehicle utilization rate is improved.
Drawings
The above and other objects, features and advantages of the present invention will become more apparent from the following description of the embodiments of the present invention with reference to the accompanying drawings, in which:
FIG. 1 is a flowchart of a region attribute prediction method according to an embodiment of the present invention;
FIG. 2 is a schematic illustration of a target area according to an embodiment of the present invention;
FIG. 3 is a schematic illustration of a sub-region in a target region according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a feature extraction model according to an embodiment of the invention;
FIG. 5 is a schematic diagram of a linkage prediction model according to an embodiment of the present invention;
FIG. 6 is a diagram illustrating an apparatus for predicting a region attribute according to an embodiment of the present invention;
fig. 7 is a schematic diagram of an electronic device according to an embodiment of the invention.
Detailed Description
The present invention will be described below based on examples, but the present invention is not limited to only these examples. In the following detailed description of the present invention, certain specific details are set forth. It will be apparent to one skilled in the art that the present invention may be practiced without these specific details. Well-known methods, procedures, components and circuits have not been described in detail so as not to obscure the present invention.
Further, those of ordinary skill in the art will appreciate that the drawings provided herein are for illustrative purposes and are not necessarily drawn to scale.
Unless the context clearly requires otherwise, throughout the description, the words "comprise", "comprising", and the like are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense; that is, what is meant is "including, but not limited to".
In the description of the present invention, it is to be understood that the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. In addition, in the description of the present invention, "a plurality" means two or more unless otherwise specified.
Fig. 1 is a flowchart of a region attribute prediction method according to an embodiment of the present invention. As shown in fig. 1, the region attribute prediction method includes the following steps:
and step S100, determining a target area.
Specifically, the target area may be directly selected from a plurality of areas pre-stored in the server, or may be temporarily determined according to information such as a received user instruction. Wherein the target region includes a plurality of sub-regions therein. For example, in the network car-booking platform, the target area may be an operation city of the network car-booking platform, and the sub-area may be a preset specific operation area in a business district, a cell, a playground, an attraction and the like in the network car-booking operation city. Optionally, after the target area is determined, the target area may be divided according to a preset rule to obtain a plurality of sub-areas.
Fig. 2 is a schematic diagram of a target area according to an embodiment of the present invention. As shown in fig. 2, the target region 20 is a preset region, and is divided according to a preset division rule to obtain a plurality of regular hexagonal sub-regions 21 having the same area. In the field of network car reservation, the target area 20 may be a network car reservation operating city area, each of the sub-areas 21 corresponds to a plurality of network car reservations, and the supply and demand of the network car reservations of different sub-areas 21 may change correspondingly with environmental changes such as weather, time, holidays, and the like.
Fig. 3 is a schematic diagram of a sub-region in a target region according to an embodiment of the invention. As shown in fig. 3, the sub-region 30 is a regular hexagonal region obtained by dividing the target region. In the field of net appointment vehicles, a plurality of net appointment vehicles 31 are included in the sub-area 30. The number of net jockey cars 31 in the sub-area 30 may be influenced by factors such as weather, time, environment, and surrounding area environment.
And S200, determining a supply vector sequence corresponding to the target area according to a preset periodic time slice sequence.
Specifically, the period, the time slice length, and the sequence length of the time slice sequence may be preset. Wherein the period may be, for example, a day, a week, a month, etc., and the time slice length may be, for example, a half hour, an hour, a day, etc. The supply vector sequence comprises a plurality of supply vectors corresponding to time slices in the time slice sequence, and each element in the supply vectors corresponds to each sub-area and is used for representing the vehicle supply quantity of the corresponding sub-area. In the field of network car reservation, the number of network car reservations corresponding to different operation areas changes along with changes of weather, holidays and time and changes of scheduling strategies of the local area and the nearby area. Thus, the supply vectors corresponding to different time slices in the supply vector sequence are different.
In this embodiment of the present invention, the process of determining the supply vector sequence corresponding to the target region includes the following steps:
step S210, acquiring a first number of vehicles corresponding to each sub-area at a starting time point of each time slice in a preset periodic time slice sequence.
Specifically, since the time slices are a time period, when determining the vehicle supply amount corresponding to each time slice in the periodic time slice sequence, it is necessary to determine the first number of vehicles in each sub-area in the target area at the starting time point of each time slice. The regional attribute prediction method is applied to a network appointment platform, and the time slices included in the time slice sequence are respectively 17:00-18:00 per day in the previous week. When determining historical vehicle supply conditions of vehicles in each subarea in the target area, acquiring the number of net appointment vehicles in each subarea at 17:00 pm of the previous week as a first number of vehicles.
Step S220, for each target object, determining the vehicle supply amount for the corresponding time slice according to the corresponding first vehicle number.
Specifically, the vehicle supply amount corresponding to each of the sub-areas at different time slices may be directly determined as the first number of vehicles acquired at the starting time point. For example, when the number of networked car reservation vehicles of a sub-area at the start time point of the corresponding time slice is 20, the vehicle supply amount of the sub-area at the time slice is determined to be 20.
And step S230, determining a supply vector sequence according to the vehicle supply quantity of each sub-area in the target area corresponding to each time slice.
Specifically, after the vehicle supply amount of each sub-area in the target area in each time slice is determined, the corresponding supply vector is determined according to the vehicle supply amount of each sub-area in each time slice. For example, when a sub-area 1, a sub-area 2, and a sub-area 3 exist in one target area, and the supply amounts of vehicles corresponding to the sub-areas in one time slice are respectively 13,9, and 16, it is determined that the supply vector corresponding to the target area in the time slice is (13,9, 16). Further, after the supply vector corresponding to each time slice is determined, the supply vector sequence is determined according to the sequence in the time slice sequence. The case that the sub-area 1, the sub-area 2 and the sub-area 3 exist in the target area, and the time slices included in the time slice sequence are respectively 17:00-18:00 for three consecutive days is taken as an example for explanation. When the supply vector corresponding to the target area of 17:00-18:00 on the first day is (13,9,16), the supply vector corresponding to the target area of 17:00-18:00 on the second day is (6,7,20), and the supply vector corresponding to the target area of 17:00-18:00 on the third day is (16,15,3), the supply vector sequence corresponding to the target area is { (13,9,16), (6,7,20), (16,15,3) }.
And step S300, determining a demand vector sequence corresponding to the target area according to the time slice sequence.
In the embodiment of the present invention, the process of obtaining the demand vector sequence may be performed simultaneously with step S300. Specifically, the demand vector sequence includes a plurality of demand vectors corresponding to time slices in the time slice sequence, and each element in the demand vectors corresponds to each sub-area and is used for representing the vehicle demand of the corresponding sub-area. In the field of network car reservation, the number of network car reservations corresponding to different operation areas changes along with changes of weather, holidays and time and changes of scheduling strategies of the local area and the nearby area. Thus, the supply vectors corresponding to different time slices in the supply vector sequence are different.
In the embodiment of the present invention, the process of determining the demand vector sequence corresponding to the target area includes the following steps:
step S310, respectively acquiring a first vehicle number and a second vehicle number corresponding to each sub-area at a start time point and an end time point of each time slice in a preset periodic time slice sequence.
Specifically, since the time slice is a time period, when determining the demand vector corresponding to each time slice in the periodic time slice sequence, it is necessary to determine the first number of vehicles and the second number of vehicles in each sub-area in the target area at the starting time point and the ending time point of each time slice. The regional attribute prediction method is applied to a network appointment platform, and the time slices included in the time slice sequence are respectively 17:00-18:00 per day in the previous week. When the historical demand condition of the vehicles in all the sub-areas in the target area is determined, the number of the net appointment vehicles which are located in all the sub-areas at 17:00 pm of the day in the previous week is obtained as a first vehicle number, and the number of the net appointment vehicles which are located in all the sub-areas at 18:00 pm of the day is obtained as a second vehicle number.
Step S320, for each target object, determining the vehicle demand amount corresponding to the time slice according to the corresponding first vehicle number and the second vehicle number.
Specifically, the vehicle demand amounts corresponding to the sub-areas in different time slices can be determined according to a first vehicle number at the starting time point and a second vehicle number at the ending time point. Optionally, the vehicle demand is a difference between the first number of vehicles and the second number of vehicles. For example, when the number of networked car-ordering vehicles of a sub-area at the starting time point of a time slice is 20, and the number of networked car-ordering vehicles at the ending time point is 10, the vehicle demand of the sub-area at the time slice is determined to be 10.
And S330, determining a demand vector sequence according to the vehicle demand quantity of each sub-area in the target area corresponding to each time slice.
In the embodiment of the present invention, the process of determining the demand vector sequence according to the vehicle demand amount corresponding to each time slice of each sub-area is similar to the process of determining the supply vector sequence according to the corresponding vehicle supply amount in step S230, and is not described herein again.
And step S400, respectively determining a corresponding supply characteristic vector sequence and a corresponding demand characteristic vector sequence according to the supply vector sequence and the demand vector sequence.
Specifically, the process of determining the supply feature vector sequence and the demand feature vector sequence is to input the supply vector sequence into a feature extraction model obtained by pre-training to determine a corresponding supply feature vector sequence. And inputting the demand vector sequence into a feature extraction model obtained by pre-training to determine a corresponding demand feature vector sequence. In an embodiment of the present invention, the feature extraction model includes an encoding module and a decoding module, wherein the encoding module includes a Non-Local mean computation layer (Non-Local) and a first long short term memory Layer (LSTM), and the decoding module includes a second long short term memory layer. In the process of feature extraction, a vector sequence to be subjected to feature extraction, namely a supply feature vector sequence or a demand feature vector sequence, is input into the encoding module, so that the input vector sequence is encoded in an iterative manner through the non-local mean calculation layer and the first long-short term memory layer. And then inputting the output result coded by the coding module into a decoding module, and outputting a plurality of characteristic vectors by each second long-short term memory layer in the decoding module in an iterative mode to determine a corresponding characteristic vector sequence.
Fig. 4 is a schematic diagram of a feature extraction model according to an embodiment of the present invention. As shown in fig. 4, the feature extraction model includes an encoding module 40 and a decoding module 41. The encoding module 40 includes a plurality of non-local mean calculation layers 42 and a first long-short term memory layer 43, and the decoding module 41 includes a second long-short term memory layer 44.
In the embodiment of the present invention, the feature extraction module performs the same feature extraction process on the demand vector sequence as that of the supply vector sequence, and the input vector sequence of the feature extraction model is taken as an example of the supply vector sequence with a length T. The supply vector sequence is (x)1,x2,…,xT) Each of the supply vectors in the supply vector sequence is input into the non-local mean calculation layer 42 and the first long-short term memory layer 43 in sequence for encoding in an iterative manner. The iterative process is to input the supply vector to the non-local mean calculation layer 42 starting from the first supply vector in the supply vector sequence, and output the output result together with the result output by the first long-short term memory layer 43 after the last iterative process to the first long-short term memory layer 43. The encoding process is completed until all the supplied vector sequences are encoded by the non-local mean calculation layer 42 and the first long-short term memory layer 43. In the embodiment of the present invention, the non-local mean calculation layer 42 corresponds to a function of
Wherein x isiAnd xjEach is any one of the input supply vector sequences, f (-to) is a similarity calculation function, g (-to) is a linear embedding function, and C (-to) is a normalization function. Alternatively, the similarity calculation function may be a dot product function. Therefore, after the supply vector sequence is input into the feature extraction model, the weighted average of each supply vector is calculated according to the f (-to) function in the non-local average calculation layer 42, and then the g (-to) function is matched to realize the dimensional unification of input and output, and the output is obtained after normalization. When the time step or the spatial range of the input features is too large, the time complexity of the f (—) function is high. The original order can be replaced by statistical featuresThe way the columns are taken as input simplifies the calculation process, and the statistical features may be, for example, mean, variance, median, etc.
The supply vector sequence is encoded by the encoding module 40 and then input to the decoding module 41, and the second long-short term memory layer 44 in the decoding module 41 performs a plurality of decoding processes, and outputs a corresponding supply eigenvector after each decoding process, and further determines the supply eigenvector sequence according to the output order of each supply eigenvector. Alternatively, the number of decoding processes may be set in advance. After the feature extraction module carries out feature extraction, the main information of the input supply vector sequence is compressed and extracted, and the calculation pressure is relieved for subsequent prediction.
Step S500, inputting the supply characteristic vector sequence, the demand characteristic vector sequence and a preset external environment vector into a pre-trained linkage prediction model to determine a supply prediction value range and a demand prediction value range corresponding to each sub-region in a target time slice.
Specifically, the time duration of the interval between the target time slice and the last time slice in the time slice sequence is one period of the time slice sequence. For example, when the time slices included in the time slice sequence are respectively 17:00-18:00 per day of the previous week, the target time slice may be 17:00-18:00 per day. After feature extraction is respectively carried out on the supply vector sequence and the demand vector sequence according to the feature extraction model to obtain the corresponding supply feature vector sequence and demand feature vector sequence, the supply feature vector sequence, the demand feature vector sequence and a preset external environment vector are input into a linkage prediction model obtained by pre-training to predict the vehicle supply condition and the vehicle demand condition of each sub-area of the target time slice.
FIG. 5 is a schematic diagram of a linkage prediction model according to an embodiment of the invention. As shown in fig. 5, the linkage prediction model includes a first embedded layer 50, a second embedded layer 51, a third long-short term memory layer 52, a fourth long-short term memory layer 53, a normalization layer 54, a first fully connected layer 55, and a second fully connected layer 56. In the embodiment of the invention, the process of predicting the demand condition and the supply condition of the target time slice vehicle by the linkage prediction model comprises the following steps:
step S510, merging the requirement feature vector sequence and the external environment vector, and inputting the merged requirement feature vector sequence and the external environment vector into a first embedding layer to determine a requirement intermediate vector.
In particular, since the vehicle supply condition and the vehicle demand condition of each sub-area can be influenced by environmental factors, an external environment vector needs to be introduced in the prediction process. The external environment vector is used for representing the environment factors of the target moment. For example, weather forecasts, legal holidays, social events, and target time slices may be included. And combining the demand characteristic vector sequence and the environmental factors, inputting the combined demand characteristic vector sequence and the environmental factors into a first embedding layer, and outputting a demand intermediate vector.
Step S520, inputting the demand intermediate vector into the third long-short term memory layer to determine a demand prediction vector.
Specifically, after the demand intermediate vector is input into the third long-short term memory layer, a demand prediction vector for characterizing the vehicle demand condition of the target time slice is further determined based on the vehicle demand condition of each sub-area history.
Step S530, inputting the demand prediction vector into a first full-link layer, so as to determine a demand prediction value range corresponding to each sub-region in a target time slice.
Specifically, the function corresponding to the first fully-connected layer is a quantile regression function having a plurality of preset quantiles. After the demand prediction vector is input into the first full-link layer, calculation is performed based on the quantile regression function and a plurality of preset quantiles, so that a plurality of demand prediction results of each sub-region corresponding to each quantile are determined. For example, when the preset quantiles are 25%, 50% and 75%, respectively, after the demand prediction vector is input into the first fully-connected layer, three demand prediction results corresponding to 25%, 50% and 75% of each sub-region are output. And determining a demand prediction value range corresponding to each sub-area in the target time slice according to a preset output rule and a plurality of corresponding demand prediction results.
In the embodiment of the present invention, the output rule may be determined according to an attribute value determined as predicted from one of the plurality of demand prediction results, and then a reliable interval is determined according to the other demand prediction results to determine the attribute value range. For example, when the preset quantiles are 25%, 50% and 75%, respectively, the output rule may determine the reliable interval according to the demand prediction results corresponding to 25% and 75%, and determine the attribute value range by using the demand prediction result corresponding to 50% as the predicted attribute value. And determining the reliable interval according to the difference value of the demand prediction results corresponding to 25% and 75%. When the demand prediction results corresponding to the preset quantiles of 25%, 50% and 75% are respectively 10, 15 and 20, the attribute value range is 15 +/-10.
And step S540, merging the supply feature vector sequence and the external environment vector, and inputting the merged result into a second embedding layer to determine a supply intermediate vector.
In the embodiment of the present invention, the process of determining the supply intermediate vector is similar to the process of determining the demand intermediate vector in step S510, and is not described herein again.
And step S550, sequentially inputting the supply intermediate vector, the demand intermediate vector and the demand prediction vector into the fourth long-short term memory layer and the second full-connected layer to determine a supply prediction value range corresponding to each sub-region in the target time slice.
Specifically, after the supply intermediate vector, the demand prediction vector, and the demand intermediate vector are merged, the merged result is input to the fourth long-short term memory layer, and the corresponding supply prediction vector is output. And inputting the supplied prediction vector into the second fully-connected layer, wherein the function corresponding to the second fully-connected layer is also a quantile regression function with a plurality of preset quantiles. Therefore, after the supply prediction vector is input into the second fully-connected layer, calculation is performed based on the quantile regression function and a plurality of preset quantiles to determine a plurality of supply prediction results of each sub-region corresponding to each quantile. And determining the range of the supply predicted value corresponding to the target time slice of each sub-area according to a preset output rule and a plurality of corresponding supply predicted results. The process of determining the range of the supply predicted value is similar to the process of determining the range of the demand predicted value in step S530, and is not repeated herein.
Further, after the supply predicted value range and the demand predicted value range corresponding to each sub-region of the target time slice are determined, a corresponding vehicle scheduling strategy can be generated according to the corresponding supply predicted value range and the demand predicted value range. The vehicle dispatching strategy is used for dispatching vehicles for each sub-area, so that the vehicle supply quantity of each sub-area can meet the vehicle demand quantity, and meanwhile, the waste of vehicle resources is avoided. For example, when the vehicle demand range corresponding to the target time slice of one sub-area is predicted to be 15 +/-10, and the corresponding vehicle supply range is predicted to be 5 +/-10, a corresponding vehicle scheduling strategy is generated, and 10 network vehicles are scheduled to the sub-area before the target time slice arrives.
According to the regional attribute prediction method, when the predicted value is determined, the feature extraction process and the prediction process are carried out separately, after the supply feature and the demand feature are extracted through the feature extraction model, the supply feature and the demand feature and factors influencing the future vehicle demand and supply are input into the linkage prediction model together, the corresponding supply predicted value range and demand predicted value range are obtained, the accuracy of the prediction result is improved, and the interpretability is improved.
Fig. 6 is a schematic diagram of a region attribute prediction apparatus according to an embodiment of the present invention. As shown in fig. 6, the apparatus includes a region determining module 60, a first sequence determining module 61, a second sequence determining module 62, a feature determining module 63, and a predicting module 64.
In particular, the region determination module 60 is configured to determine a target region, which includes a plurality of sub-regions therein. The first sequence determining module 61 is configured to determine a supply vector sequence corresponding to the target region according to a preset periodic time slice sequence, where the supply vector sequence includes a plurality of supply vectors corresponding to each time slice in the time slice sequence, and is used to characterize the vehicle supply amount of each sub-region. The second sequence determining module 62 is configured to determine a demand vector sequence corresponding to the target region according to the time slice sequence, where the demand vector sequence includes a plurality of demand vectors corresponding to each time slice in the time slice sequence, and is used to characterize the vehicle demand of each sub-region. The characteristic determining module 63 is configured to determine a corresponding supply characteristic vector sequence and a corresponding demand characteristic vector sequence according to the supply vector sequence and the demand vector sequence. The prediction module 64 is configured to input the supply feature vector sequence, the demand feature vector sequence, and a preset external environment vector into a pre-trained linkage prediction model to determine a supply prediction value range and a demand prediction value range of each sub-region corresponding to a target time slice, where an interval duration between the target time slice and a last time slice in the time slice sequence is one period of the time slice sequence.
Further, the first sequence determination module comprises:
the first information acquisition unit is used for acquiring a first vehicle number corresponding to each subarea at the starting time point of each time slice in a preset periodic time slice sequence;
a first information determination unit configured to determine, for each of the target objects, a vehicle supply amount for a corresponding time slot based on the corresponding number of each of the first vehicles;
and the first sequence determining unit is used for determining a supply vector sequence according to the vehicle supply quantity of each sub-region in the target region corresponding to each time slice.
Further, the second sequence determination module comprises:
the second information acquisition unit is used for respectively acquiring a first vehicle number and a second vehicle number corresponding to each subarea at the starting time point and the ending time point of each time slice in a preset periodic time slice sequence;
the second information determining unit is used for determining the vehicle demand amount corresponding to the time slice according to the corresponding first vehicle number and the second vehicle number for each target object;
and the second sequence determining unit is used for determining a demand vector sequence according to the vehicle demand corresponding to each time slice of each sub-region in the target region.
Further, the feature determination module includes:
the first feature extraction unit is used for inputting the supply vector sequence into a feature extraction model obtained by pre-training so as to determine a corresponding supply feature vector sequence; and the number of the first and second groups,
and the second feature extraction unit is used for inputting the demand vector sequence into a feature extraction model obtained by pre-training so as to determine a corresponding demand feature vector sequence.
Further, the feature extraction model comprises an encoding module and a decoding module, wherein the encoding module comprises a non-local mean calculation layer and a first long-short term memory layer, and the decoding module comprises a second long-short term memory layer.
Further, the linkage prediction model comprises a first embedding layer, a second embedding layer, a third long-short term memory layer, a fourth long-short term memory layer, a normalization layer, a first full-connection layer and a second full-connection layer;
the prediction module comprises:
the demand vector determining unit is used for merging the demand characteristic vector sequence and the external environment vector and inputting the merged demand characteristic vector sequence and the external environment vector into a first embedding layer so as to determine a demand intermediate vector;
a first demand forecast value determining unit, configured to input the demand intermediate vector into the third long-short term memory layer to determine a demand forecast vector;
a second demand prediction value determination unit, configured to input the demand prediction vector into a first full-link layer, so as to determine a demand prediction value range corresponding to each sub-region in a target time slice;
a demand vector determining unit, configured to combine the supply feature vector sequence and the external environment vector and input the combined result to a second embedding layer, so as to determine a supply intermediate vector;
and the demand forecast value determining unit is used for sequentially inputting the supply intermediate vector, the demand intermediate vector and the demand forecast vector into the fourth long-short term memory layer and the second full connection layer so as to determine a supply forecast value range corresponding to each sub-region in a target time slice.
Further, the demand forecast value determination unit includes:
the first prediction subunit is used for inputting the demand prediction vector into the first full-connection layer so as to calculate and determine a plurality of demand prediction results corresponding to the sub-regions according to a quantile regression function and a plurality of preset quantiles;
and the second prediction subunit is used for determining a demand prediction value range corresponding to each sub-area in the target time slice according to a preset output rule and a plurality of corresponding demand prediction results.
Further, the supply prediction value determination unit includes:
a third predictor unit for inputting the supply intermediate vector, demand prediction vector and demand intermediate vector into the fourth long-short term memory layer to output corresponding supply prediction vectors;
the fourth prediction subunit is used for inputting the supply prediction vector into the second full-connection layer so as to calculate and determine a plurality of supply prediction results corresponding to the sub-regions according to a quantile regression function and a plurality of preset quantiles;
and the fifth prediction subunit is used for determining a supply prediction value range corresponding to each sub-area in the target time slice according to a preset output rule and a plurality of corresponding supply prediction results.
The regional attribute prediction device of the embodiment of the invention separately performs the feature extraction process and the prediction process when determining the predicted value, and after the supply feature and the demand feature are extracted by the feature extraction model, the supply feature and the demand feature and factors influencing the future vehicle demand and supply are input into the linkage prediction model together to obtain the corresponding supply predicted value range and demand predicted value range, thereby improving the accuracy of the prediction result and increasing the interpretability.
Fig. 7 is a schematic diagram of an electronic device according to an embodiment of the invention. As shown in fig. 7, the electronic device shown in fig. 7 is a general address query device, which includes a general computer hardware structure, which includes at least a processor 70 and a memory 71. The processor 70 and the memory 71 are connected by a bus 72. The memory 71 is adapted to store instructions or programs executable by the processor 70. Processor 70 may be a stand-alone microprocessor or may be a collection of one or more microprocessors. Thus, the processor 70 implements the processing of data and the control of other devices by executing instructions stored by the memory 71 to perform the method flows of embodiments of the present invention as described above. The bus 72 connects the above components together, as well as to a display controller 73 and a display device and an input/output (I/O) device 74. Input/output (I/O) devices 74 may be a mouse, keyboard, modem, network interface, touch input device, motion sensing input device, printer, and other devices known in the art. Typically, the input/output devices 74 are connected to the system through input/output (I/O) controllers 75.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, apparatus (device) or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may employ a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations of methods, apparatus (devices) and computer program products according to embodiments of the application. It will be understood that each flow in the flow diagrams can be implemented by computer program instructions.
These computer program instructions may be stored in a computer-readable memory that can direct a computer or other programmable vehicle dispatch device to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows.
These computer program instructions may also be provided to a processor of a general purpose computer, special purpose computer, embedded processor or other programmable vehicle scheduling apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable vehicle scheduling apparatus, create means for implementing the functions specified in the flowchart flow or flows.
Another embodiment of the invention is directed to a non-transitory storage medium storing a computer-readable program for causing a computer to perform some or all of the above-described method embodiments.
That is, as can be understood by those skilled in the art, all or part of the steps in the method for implementing the embodiments described above may be implemented by a program instructing related hardware, where the program is stored in a storage medium and includes several instructions to enable a device (which may be a single chip, a chip, or the like) or a processor (processor) to execute all or part of the steps of the method described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (18)
1. A method for predicting regional attributes, the method comprising:
determining a target area, wherein the target area comprises a plurality of sub-areas;
determining a supply vector sequence corresponding to the target area according to a preset periodic time slice sequence, wherein the supply vector sequence comprises a plurality of supply vectors corresponding to each time slice in the time slice sequence and is used for representing the vehicle supply quantity of each sub-area;
determining a demand vector sequence corresponding to the target area according to the time slice sequence, wherein the demand vector sequence comprises a plurality of demand vectors corresponding to each time slice in the time slice sequence and is used for representing the vehicle demand of each sub-area;
respectively determining a corresponding supply characteristic vector sequence and a corresponding demand characteristic vector sequence according to the supply vector sequence and the demand vector sequence;
inputting the supply characteristic vector sequence, the demand characteristic vector sequence and a preset external environment vector into a linkage prediction model obtained by pre-training so as to simultaneously determine a supply prediction value range and a demand prediction value range of each sub-region corresponding to a target time slice, wherein the interval duration between the target time slice and the last time slice in the time slice sequence is one period of the time slice sequence;
the linkage prediction model comprises a first embedded layer, a second embedded layer, a third long-short term memory layer, a fourth long-short term memory layer, a normalization layer, a first full-connection layer and a second full-connection layer.
2. The method according to claim 1, wherein the determining the sequence of supply vectors corresponding to the target region according to the preset periodic time slice sequence comprises:
acquiring a first vehicle number corresponding to each subregion at the starting time point of each time slice in a preset periodic time slice sequence;
for each subregion, determining the vehicle supply quantity of the corresponding time slice according to the corresponding first vehicle number;
and determining a supply vector sequence according to the vehicle supply quantity of each sub-area in the target area corresponding to each time slice.
3. The method of claim 1, wherein the determining the sequence of demand vectors corresponding to the target region according to the sequence of time slices comprises:
respectively acquiring a first vehicle number and a second vehicle number corresponding to each subarea at the starting time point and the ending time point of each time slice in a preset periodic time slice sequence;
for each target object, determining the vehicle demand amount corresponding to the time slice according to the corresponding first vehicle number and second vehicle number;
and determining a demand vector sequence according to the vehicle demand corresponding to each time slice of each sub-area in the target area.
4. The method of claim 1, wherein determining the corresponding supply and demand eigenvector sequences from the supply and demand vector sequences, respectively, comprises:
inputting the supply vector sequence into a feature extraction model obtained by pre-training to determine a corresponding supply feature vector sequence; and the number of the first and second groups,
and inputting the demand vector sequence into a feature extraction model obtained by pre-training to determine a corresponding demand feature vector sequence.
5. The method of claim 4, wherein the feature extraction model comprises an encoding module and a decoding module, wherein the encoding module comprises a non-local mean computation layer and a first long-short term memory layer, and wherein the decoding module comprises a second long-short term memory layer.
6. The method of claim 1, wherein the inputting the supply eigenvector sequence, the demand eigenvector sequence, and the preset external environment vector into a pre-trained linkage prediction model to determine a supply predicted value range and a demand predicted value range corresponding to each sub-region in a target time slice comprises:
merging the demand characteristic vector sequence and the external environment vector and inputting the merged demand characteristic vector sequence and the external environment vector into a first embedding layer to determine a demand intermediate vector;
inputting the demand intermediate vector into the third long-short term memory layer to determine a demand prediction vector;
inputting the demand forecast vector into a first full-connection layer to determine a demand forecast value range corresponding to each sub-area in a target time slice;
merging the supply characteristic vector sequence and the external environment vector and inputting the merged supply characteristic vector sequence and the external environment vector into a second embedding layer to determine a supply intermediate vector;
and sequentially inputting the supply intermediate vector, the demand intermediate vector and the demand prediction vector into the fourth long-short term memory layer and the second full-connection layer to determine a supply prediction value range corresponding to each sub-region in a target time slice.
7. The method of claim 6, wherein inputting the demand prediction vector into a first fully-connected layer to determine a range of demand prediction values for each sub-region in a target time slice comprises:
inputting the demand prediction vector into the first full-connection layer to calculate and determine a plurality of demand prediction results corresponding to each sub-region according to a quantile regression function and a plurality of preset quantiles;
and determining the range of the demand predicted value corresponding to the target time slice of each sub-area according to a preset output rule and a plurality of corresponding demand predicted results.
8. The method of claim 6, wherein inputting the supply intermediate vector, demand forecast vector and demand intermediate vector into the fourth long-short term memory layer and second fully-connected layer to determine a supply forecast range for each sub-region in a target time slice comprises:
inputting the supply intermediate vector, demand forecast vector and demand intermediate vector into the fourth long-short term memory layer to output corresponding supply forecast vector;
inputting the supply prediction vector into the second full-connection layer to calculate and determine a plurality of supply prediction results corresponding to each sub-region according to a quantile regression function and a plurality of preset quantiles;
and determining the supply predicted value range corresponding to each sub-area in the target time slice according to a preset output rule and a plurality of corresponding supply predicted results.
9. An area attribute prediction apparatus, the apparatus comprising:
the device comprises a region determining module, a judging module and a judging module, wherein the region determining module is used for determining a target region, and the target region comprises a plurality of sub-regions;
the first sequence determining module is used for determining a supply vector sequence corresponding to the target region according to a preset periodic time slice sequence, wherein the supply vector sequence comprises a plurality of supply vectors corresponding to time slices in the time slice sequence and is used for representing the vehicle supply quantity of each sub-region;
the second sequence determining module is used for determining a demand vector sequence corresponding to the target area according to the time slice sequence, wherein the demand vector sequence comprises a plurality of demand vectors corresponding to each time slice in the time slice sequence and is used for representing the vehicle demand of each sub-area;
the characteristic determining module is used for respectively determining a corresponding supply characteristic vector sequence and a corresponding demand characteristic vector sequence according to the supply vector sequence and the demand vector sequence;
the prediction module is used for inputting the supply characteristic vector sequence, the demand characteristic vector sequence and a preset external environment vector into a linkage prediction model obtained through pre-training so as to simultaneously determine a supply prediction value range and a demand prediction value range of each sub-region corresponding to a target time slice, wherein the interval duration between the target time slice and the last time slice in the time slice sequence is one period of the time slice sequence;
the linkage prediction model comprises a first embedded layer, a second embedded layer, a third long-short term memory layer, a fourth long-short term memory layer, a normalization layer, a first full-connection layer and a second full-connection layer.
10. The apparatus of claim 9, wherein the first sequence determining module comprises:
the first information acquisition unit is used for acquiring a first vehicle number corresponding to each subarea at the starting time point of each time slice in a preset periodic time slice sequence;
a first information determination unit for determining, for each of the sub-areas, a vehicle supply amount for a corresponding time slot from the corresponding first number of vehicles;
and the first sequence determining unit is used for determining a supply vector sequence according to the vehicle supply quantity of each sub-region in the target region corresponding to each time slice.
11. The apparatus of claim 9, wherein the second sequence determining module comprises:
the second information acquisition unit is used for respectively acquiring a first vehicle number and a second vehicle number corresponding to each subarea at the starting time point and the ending time point of each time slice in a preset periodic time slice sequence;
the second information determining unit is used for determining the vehicle demand amount corresponding to the time slice according to the corresponding first vehicle number and the second vehicle number for each target object;
and the second sequence determining unit is used for determining a demand vector sequence according to the vehicle demand corresponding to each time slice of each sub-region in the target region.
12. The apparatus of claim 9, wherein the feature determination module comprises:
the first feature extraction unit is used for inputting the supply vector sequence into a feature extraction model obtained by pre-training so as to determine a corresponding supply feature vector sequence; and the number of the first and second groups,
and the second feature extraction unit is used for inputting the demand vector sequence into a feature extraction model obtained by pre-training so as to determine a corresponding demand feature vector sequence.
13. The apparatus of claim 12, wherein the feature extraction model comprises an encoding module and a decoding module, wherein the encoding module comprises a non-local mean computation layer and a first long-short term memory layer, and wherein the decoding module comprises a second long-short term memory layer.
14. The apparatus of claim 9, wherein the prediction module comprises:
the demand vector determining unit is used for merging the demand characteristic vector sequence and the external environment vector and inputting the merged demand characteristic vector sequence and the external environment vector into a first embedding layer so as to determine a demand intermediate vector;
a first demand forecast value determining unit, configured to input the demand intermediate vector into the third long-short term memory layer to determine a demand forecast vector;
a second demand prediction value determination unit, configured to input the demand prediction vector into a first full-link layer, so as to determine a demand prediction value range corresponding to each sub-region in a target time slice;
a demand vector determining unit, configured to combine the supply feature vector sequence and the external environment vector and input the combined result to a second embedding layer, so as to determine a supply intermediate vector;
and the supply predicted value determining unit is used for sequentially inputting the supply intermediate vector, the demand intermediate vector and the demand predicted vector into the fourth long-short term memory layer and the second full-connection layer so as to determine the supply predicted value range corresponding to each sub-region in the target time slice.
15. The apparatus of claim 14, wherein the second demand prediction value determination unit comprises:
the first prediction subunit is used for inputting the demand prediction vector into the first full-connection layer so as to calculate and determine a plurality of demand prediction results corresponding to the sub-regions according to a quantile regression function and a plurality of preset quantiles;
and the second prediction subunit is used for determining a demand prediction value range corresponding to each sub-area in the target time slice according to a preset output rule and a plurality of corresponding demand prediction results.
16. The apparatus according to claim 14, wherein the supply prediction value determination unit comprises:
a third predictor unit for inputting the supply intermediate vector, demand prediction vector and demand intermediate vector into the fourth long-short term memory layer to output corresponding supply prediction vectors;
the fourth prediction subunit is used for inputting the supply prediction vector into the second full-connection layer so as to calculate and determine a plurality of supply prediction results corresponding to the sub-regions according to a quantile regression function and a plurality of preset quantiles;
and the fifth prediction subunit is used for determining a supply prediction value range corresponding to each sub-area in the target time slice according to a preset output rule and a plurality of corresponding supply prediction results.
17. A computer readable storage medium storing computer program instructions, which when executed by a processor implement the method of any one of claims 1-8.
18. An electronic device comprising a memory and a processor, wherein the memory is configured to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method of any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010998050.9A CN112101671B (en) | 2020-09-21 | 2020-09-21 | Region attribute prediction method and device, readable storage medium and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010998050.9A CN112101671B (en) | 2020-09-21 | 2020-09-21 | Region attribute prediction method and device, readable storage medium and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112101671A CN112101671A (en) | 2020-12-18 |
| CN112101671B true CN112101671B (en) | 2022-03-25 |
Family
ID=73756410
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010998050.9A Active CN112101671B (en) | 2020-09-21 | 2020-09-21 | Region attribute prediction method and device, readable storage medium and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112101671B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113222287A (en) * | 2021-06-01 | 2021-08-06 | 平安科技（深圳）有限公司 | Network appointment demand prediction method, device, equipment and storage medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111476588A (en) * | 2019-01-24 | 2020-07-31 | 北京嘀嘀无限科技发展有限公司 | Order demand prediction method and device, electronic equipment and readable storage medium |
-
2020
- 2020-09-21 CN CN202010998050.9A patent/CN112101671B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111476588A (en) * | 2019-01-24 | 2020-07-31 | 北京嘀嘀无限科技发展有限公司 | Order demand prediction method and device, electronic equipment and readable storage medium |
Non-Patent Citations (2)
| Title |
|---|
| Hexagon-Based Convolutional Neural Network for Supply-Demand Forecasting of Ride-Sourcing Services;Jintao ke 等;《IEEE Transactions on Intelligent Transportation Systems》;20181231;第1-15页 * |
| 基于时间序列分析的网约车需求短时预测研究;况东钰;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20200131;第19-73页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112101671A (en) | 2020-12-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108010378B (en) | Parking lot berth occupancy rate prediction method, device, equipment and storage medium | |
| CN107463701B (en) | Method and device for pushing information stream based on artificial intelligence | |
| CN105808784A (en) | Recommendation method and device | |
| US20210174270A1 (en) | Rideshare vehicle demand forecasting device, method for forecasting rideshare vehicle demand, and storage medium | |
| CN115730790A (en) | Charging configuration method, device and equipment based on edge calculation and storage medium | |
| CN112101671B (en) | Region attribute prediction method and device, readable storage medium and electronic equipment | |
| CN109243619A (en) | Generation method, device and the computer readable storage medium of prediction model | |
| CN111832876A (en) | Vehicle scheduling method, readable storage medium and electronic device | |
| CN108694182B (en) | Activity pushing method, activity pushing device and server | |
| CN114816690A (en) | Task allocation method, device, equipment and storage medium | |
| CN112101804B (en) | Vehicle scheduling method and device, readable storage medium and electronic equipment | |
| CN112988992B (en) | Information interaction method and device and electronic equipment | |
| CN112529602A (en) | Data processing method and device, readable storage medium and electronic equipment | |
| CN114625340A (en) | Commercial software research and development method, device, equipment and medium based on demand analysis | |
| TW202215313A (en) | Communications apparatus and method for estimating potential demand for improved forecasting and resource management | |
| CN112116167A (en) | Data processing method, data processing apparatus, storage medium, and electronic device | |
| WO2021230808A1 (en) | Server and method of determining an advanced booking fee for an advance booking | |
| CN114723469A (en) | Method, system and electronic device for user satisfaction degree prediction and attribution | |
| CN113837782A (en) | Method and device for optimizing periodic item parameters of time series model and computer equipment | |
| US20150134394A1 (en) | System for planning a building project | |
| CN112687096B (en) | Information processing apparatus and information processing method | |
| CN110956294A (en) | Queuing time estimation method and device | |
| US20240112790A1 (en) | System and method for optimizing resource allocation | |
| CN117078147A (en) | Recommended path confirmation method, recommended path confirmation device, electronic equipment and storage medium | |
| CN118334631A (en) | Driving behavior prediction model construction method, prediction device, driving behavior prediction model construction equipment and driving behavior prediction model construction medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |