CN112990526B - Method, apparatus and storage medium for predicting amount of material to be delivered - Google Patents
Method, apparatus and storage medium for predicting amount of material to be delivered Download PDFInfo
- Publication number
- CN112990526B CN112990526B CN201911293320.XA CN201911293320A CN112990526B CN 112990526 B CN112990526 B CN 112990526B CN 201911293320 A CN201911293320 A CN 201911293320A CN 112990526 B CN112990526 B CN 112990526B
- Authority
- CN
- China
- Prior art keywords
- time
- time period
- arrival
- predicting
- period
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; 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
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Quality & Reliability (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the application discloses a method for predicting logistics to a piece, equipment for predicting logistics to the piece and a storage medium. The method for predicting the flow-to-piece quantity comprises the following steps: dividing a fixed period of time for which prediction is desired into a plurality of time periods; acquiring historical data of the arrival quantity corresponding to each time period; calculating real-time on-road part quantity corresponding to each time period; predicting the arrival quantity of each time period based on the arrival quantity historical data corresponding to each time period through a time sequence prediction model, and generating a time sequence predicted value for each time period; and performing weighted average on the time series predicted value and the real-time on-road component quantity to obtain a final predicted result, wherein the final predicted result is equal to the sum of the time series predicted value multiplied by a first parameter and the real-time on-road component quantity multiplied by a second parameter.
Description
Technical Field
The application relates to prediction of a piece amount, in particular to a method for predicting a logistics piece amount, a device for predicting the logistics piece amount and a storage medium.
Background
With the rapid development of the logistics industry, the transfer field arrival amount is increased year by year, and in order to improve the operation quality of the transfer field, the resources are reasonably scheduled, and the arrival amount of each half hour period in the future 2 days of the transfer field is required to be predicted.
The transition field is predicted for half an hour, and the time granularity is extremely small. Currently, the time series model-based prediction of the arrival amount of the transition date is used for predicting the arrival amount of the transition day, but the predicted arrival amount cannot be thinned to each half hour period in one day.
In order to overcome the problems of the existing methods, there is a need for a method for predicting the amount of material to be fed, a device for predicting the amount of material to be fed, and a storage medium.
Disclosure of Invention
The embodiment of the application provides a method for predicting logistics to a piece, equipment for predicting logistics to the piece and a storage medium. According to the embodiment of the application, the flow to piece quantity prediction of each half-hour period, such as the transfer to the middle transfer, can be realized. The application can improve the operation efficiency and provide objective and scientific data basis for resource allocation.
In a first aspect, an embodiment of the present application provides a method for predicting a flow-to-piece amount, including:
dividing a fixed period of time for which prediction is desired into a plurality of time periods;
Acquiring historical data of the arrival quantity corresponding to each time period;
calculating real-time on-road part quantity corresponding to each time period;
Predicting the arrival quantity of each time period based on the arrival quantity historical data corresponding to each time period through a time sequence prediction model, and generating a time sequence predicted value for each time period; and
And carrying out weighted average on the time sequence predicted value and the real-time on-road component quantity to obtain a final predicted result, wherein the final predicted result is equal to the sum of the time sequence predicted value multiplied by a first parameter and the real-time on-road component quantity multiplied by a second parameter.
In some embodiments, the time series prediction model is an autoregressive integrated moving average model (ARIMA model); the step of dividing a fixed period of time for which prediction is desired into a plurality of time periods includes: at half hour intervals, a fixed 24 hour time period was divided into 48 time periods.
In some embodiments, the autoregressive integral moving average model is trained by:
Taking the first 80% of the historical data of the arrival quantity corresponding to the 48 time periods as a training set;
training with the first 80% of the training set as training input values and the last 20% of the training set as training true values; and
The first parameter and the second parameter are obtained by means of a mean square error loss function.
In some embodiments, the autoregressive integral moving average model is verified by:
In the history data of the arrival quantity corresponding to the 48 time periods, the last 20% is taken as a verification set; and
The verification is performed with the first 80% of the verification set as an input value for verification and the last 20% of the verification set as a true value for verification.
In some embodiments, the step of calculating the real-time in-transit piece quantity corresponding to each time period includes:
Locating each in-transit parcel for each time period;
calculating the navigation time of each in-transit package arrival in each time period;
Calculating the residence time of each in-transit package in each time period;
calculating an arrival time for each in-transit package within each time period, wherein the arrival time is equal to a current time plus the navigation time plus the hold-up time; and
An on-transit item quantity is calculated for each time period, wherein the on-transit item quantity is equal to a total item quantity of the on-transit package having an arrival time that falls between a start and end time of each time period.
In some embodiments, one or more of the plurality of time periods are selected on demand, and the selected one or more time periods are predicted based on the arrival amount history data corresponding to each time period.
In a second aspect, an embodiment of the present application further provides a device for predicting an amount of a material flow to a piece, including a processor and a storage, where the processor invokes a computer program in the storage to execute any one of the methods for predicting an amount of a material flow to a piece provided by the embodiments of the present application.
In some embodiments, the computer program comprises:
The feature selection module is used for dividing a fixed period of time expected to be predicted into a plurality of time periods and acquiring the arrival quantity historical data corresponding to each time period;
the real-time on-road part quantity calculating module is used for calculating the real-time on-road part quantity corresponding to each time period; and
The prediction module is used for predicting the arrival quantity of each time period based on the arrival quantity historical data corresponding to each time period through a time sequence prediction model, generating a time sequence prediction value for each time period, and carrying out weighted average on the time sequence prediction value and the real-time arrival quantity to obtain a final prediction result, wherein the final prediction result is equal to the sum of the time sequence prediction value multiplied by a first parameter and the real-time arrival quantity multiplied by a second parameter.
In a third aspect, the present application further provides a storage medium, where the storage medium is used to store a computer program, where the computer program is loaded by a processor, so as to execute the steps in any one of the method for predicting a flow-to-piece amount provided by the embodiment of the present application.
The embodiment of the application realizes the prediction of the logistics arrival amount of each half-hour period, such as the prediction of the arrival amount of a medium-transition. The application can improve the operation efficiency and provide objective and scientific data basis for resource allocation.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of steps of a method for predicting a flow-to-part quantity according to an embodiment of the present application;
FIG. 2 is a flowchart of the steps for calculating the real-time in-transit volume corresponding to each time period according to an embodiment of the present application;
FIG. 3 is a flow-to-part quantity prediction device provided by an embodiment of the present application;
FIG. 4 is a schematic diagram illustrating the operation of the prediction module according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to fall within the scope of the application.
Referring to fig. 1 to 4, fig. 1 is a flowchart illustrating steps of a method for predicting an amount of a current component according to an embodiment of the present application, fig. 2 is a flowchart illustrating steps of calculating an amount of a current component in real time corresponding to each time period according to an embodiment of the present application, fig. 3 is an apparatus for predicting an amount of a current component according to an embodiment of the present application, and fig. 4 is an operation schematic diagram of a prediction module according to an embodiment of the present application.
As shown in fig. 1, the method for predicting the arrival amount of a logistics object provided by the embodiment of the present application may be used for predicting the arrival amount of a destination (for example, a transition) where a logistics object is often transported, and may include steps S110 to S150, where:
Step S110, dividing a fixed period of time for which prediction is desired into a plurality of time periods. In step S110, dividing the fixed time period for which prediction is desired into a plurality of time periods may include: at half hour intervals, a fixed 24 hour time period was divided into 48 time periods. For example, a day is divided into 48 half-hour periods, where 0:00-0:30 is the 1 st half-hour period of the day, 00:30-01:00 is the 2 nd half-hour period of the day, and so on.
Step S120, obtaining the history data of the arrival quantity corresponding to each time period. Such as the arrival volume history of the stream for each half hour of the day, and arranged in a time series. It should be understood that the arrival volume history data corresponding to each time period may be arranged according to requirements, for example, the arrival volume history data of the time period of 0:00-0:30 per day in the past year is formed into a time sequence.
Step S130, calculating the real-time on-road part quantity corresponding to each time period. Referring to fig. 2, step S130 may include steps S1301 to S1305, where:
Step S1301, locating each in-transit parcel in each time period; wherein packages in transit within each time period can be located by a geographic information system (GIS, geographic information system);
step S1302, calculating the navigation time of each in-transit package arrival in each time period;
step S1303, calculating the retention time of each in-transit package in each time period;
step S1304, calculating an arrival time of each in-transit package within each time period, wherein the arrival time is equal to a current time plus the navigation time plus the residence time; and
Step S1305 calculates an in-transit item quantity within each time period, wherein the in-transit item quantity is equal to a total item quantity of the in-transit package whose arrival time falls between a start and end time of each time period.
In step S140, the arrival amount of each time zone is predicted based on the arrival amount history data corresponding to each time zone by an autoregressive integral moving average model (ARIMA model), and a time-series predicted value (ARIMA predicted value) is generated for each time zone.
Step S150, performing weighted average on the ARIMA prediction value and the real-time on-road component quantity to obtain a final prediction result, where the final prediction result is equal to the sum of the ARIMA prediction value multiplied by the first parameter and the real-time on-road component quantity multiplied by the second parameter. The first parameter and the second parameter may be obtained by training.
The ARIMA model may be trained by: taking the first 80% of the historical data of the arrival quantity corresponding to the 48 time periods as a training set; training with the first 80% of the training set as training input values and the last 20% of the training set as training true values; and obtaining the first parameter and the second parameter through a mean square error loss function. Specifically, the first 80% of the training set is used as an input time sequence for training, x is used as x, and the last 20% of the training set is used as a true value for training, y is used as y. The MSE (mean square error) loss function is used, and the formula is as follows:
Where N represents the number of time series,
Representing the predicted value of the ith half hour period,
Y i represents the true value of the ith half hour period.
And calculating the gradient of the loss function on the ARIMA prediction result and the weight of the in-transit piece quantity, and modifying the weights, namely modifying the first parameter and the second parameter. The parameter modification formula is:
wherein W' represents the parameters of the modified ARIMA predicted value and the real-time in-transit volume,
W represents parameters of the pre-modification ARIMA forecast and the real-time in-transit volume,
Representing the gradient of the parameter w.
Training is continuously performed through the training input value until the training input value is used up. Thus, accurate first parameters and second parameters can be obtained through a large amount of training.
Furthermore, the ARIMA model is validated by the following steps: in the history data of the arrival quantity corresponding to the 48 time periods, the last 20% is taken as a verification set; and performing verification with the first 80% of the verification set as an input value for verification and the second 20% of the verification set as a true value for verification. Specifically, the first 80% of the verification set is used as the verification input time sequence, and the last 20% of the training set is used as the verification true value. The ARIMA model after training is input with an input time series for verification (input values for verification) and a corresponding final predicted value for verification is obtained, and an error is calculated using the loss function as verification.
It should be understood that the present application may also select one or more of the plurality of time periods as needed, and predict the selected one or more time periods based on the arrival amount history data corresponding to each time period. Specifically, a day may be divided into 48 half-hour periods, and the same half-hour period of history is used every day, and the time series is formed by arranging the history according to the date to predict. For example, the arrival volume historical data of the time period of 0:00-0:30 per day is formed into a time series to predict the arrival volume historical data in one year.
In order to better implement the method for predicting the amount of the material flow provided by the embodiment of the present application, the embodiment of the present application further provides a device 200 for predicting the amount of the material flow, where the meaning of the term is the same as that of the foregoing method for predicting the amount of the material flow, and specific implementation details may be referred to in the following description of embodiments. As shown in fig. 3, the flow-to-volume prediction apparatus 200 may include one or more processors 210 of a processing core, one or more storage 220 of a computer-readable storage medium, and a power supply 230. It will be appreciated by those skilled in the art that the schematic diagram shown in fig. 3 is not limiting of the flow-to-part quantity prediction apparatus 200, and may include more or fewer components than the figures, or may combine certain components, or a different arrangement of components.
The processor 210 connects the various parts of the overall logistics-to-volume prediction apparatus 200 using various interfaces and lines and runs or executes a computer program 221 (e.g., a software program, module, or algorithm) stored in the memory 220. Optionally, processor 201 may include one or more processing cores; preferably, the processor 201 may integrate an application processor and a modem processor, wherein the application processor primarily processes an operating system, a user interface, an application program, etc., and the modem processor primarily processes wireless communication. It will be appreciated that the modem processor described above may not be integrated into the processor 201.
The memory 220 may be used to store a computer program 221 (e.g., a software program, a module, or an algorithm), and the processor 210 performs various functional applications and data processing by executing the computer program 221 stored in the memory 220. The memory 220 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program (such as a sound playing function, an image playing function, etc.) required for at least one function, etc.; the storage data area may store data created according to the use of the server, etc. In addition, the memory 220 may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 220 may also include a memory controller to provide the processor 210 with access to the memory 220.
The power supply 230 is configured to supply power to each component of the predicting apparatus 200 for logistics to a piece, and preferably, the power supply 230 may be logically connected to the processor 210 through a power management system, so as to implement functions of managing charging, discharging, and power consumption management through the power management system. The power supply 230 may also include one or more of any components, such as a DC or AC power supply, a recharging system, a power failure detection circuit, a power converter or inverter, a power status indicator, etc.
Although not shown, the apparatus 200 for predicting the amount of the delivered-goods may further include a display unit, etc., and will not be described herein. In particular, the processor 210 in the flow-to-volume prediction apparatus 200 runs the computer program 221 stored in the memory 220, thereby implementing various functions. The computer program 221 may comprise: the feature selection module 221a, the real-time in-transit piece calculation module 221b, the prediction module 221c, the training and verification module 221d, and the prediction effect evaluation module 221e.
The feature selection module 221a is configured to divide a fixed period of time expected to be predicted into a plurality of time periods, and obtain the arrival amount history data corresponding to each time period.
The real-time on-road component calculation module 221b is configured to calculate a real-time on-road component amount corresponding to each time period.
The prediction module 221c is configured to predict, by using an ARIMA model, an arrival amount of each time slot based on arrival amount historical data corresponding to each time slot, generate an ARIMA prediction value for each time slot, and perform weighted average on the ARIMA prediction value and the real-time in-transit amount to obtain a final prediction result, where the final prediction result is equal to a sum of the ARIMA prediction value multiplied by a first parameter and the real-time in-transit amount multiplied by a second parameter. As shown in fig. 4, the prediction module 221c performs at least the following operations, inputs the time series into the ARIMA model, and obtains the ARIMA prediction value. The ARIMA prediction value is then weighted averaged with the real-time in-transit item quantity to obtain a final prediction result.
The training and validation module 221d is configured to train the ARIMA model and validate the ARIMA model. Wherein the details of training and validating the ARIMA model are described in detail in the above embodiments.
The prediction effect evaluation module 221e is configured to evaluate a prediction error. For example, a mean square error is used to calculate the prediction error. That is, the errors between the final predicted result and the actual amount of work are compared to determine if the current ARIMA model requires retraining and verification.
The following is an example of the transition amount in prediction, and is described with reference to the accompanying drawings. Based on business experience and historical data analysis, the half hour to piece quantity factor of the transfer field is affected by the following two points: (1) Historical half-hour time period to the amount of the part and (2) real-time on-the-way amount, that is, the on-the-way amount sent to the intermediate transfer. The effect of the historical half-hour time period to the part quantity is that the predicted part quantity of a specific half-hour time period is often related to the part quantity of the historical half-hour time period, for example, the half-hour time period from 1 day of 2019 month 9 to 5 days of 2019 month 5:00-5:30 is greatly increased in a relatively normal state, and the predicted part quantity of the half-hour time period from 5:00-5:30 of 2019 month 6 is also greatly increased in a relatively normal state.
Based on the above factors, the feature selection module 221a divides a day into 48 half-hour time periods, such as: 0:00-0:30 is the 1 st half hour period of the day, and 00:30-01:00 is the 2 nd half hour period of the day. And simultaneously acquiring the arrival quantity historical data corresponding to each half-hour time period.
Subsequently, the feature selection module 221a arranges the same half-hour time period as each day in the volume history data into a time series by date. For example, historical data of the arrival volume of the transition field for a period of 0:00-0:30 per day in the past year is formed into a time series.
And the real-time on-road parts calculation module 221b calculates the real-time on-road parts corresponding to the time period. The calculation method of the real-time on-the-way quantity sequentially comprises the following steps:
(a) Positioning packages through a GIS system to position packages in the transportation process in each time period;
(b) Calculating the navigation time of each in-transit parcel reaching the transition field in each time period on the transition path diagram by adopting a Di-Jie-Style algorithm;
(c) Calculating the residence time of each in-transit package in the transit field in each time period, for example, calculating the residence average time of each in-transit package in each time period of each transit field through historical data;
(d) Calculating an arrival time for each in-transit package within each time period, wherein the arrival time is equal to a current time plus the navigation time plus the hold-up time;
(e) An on-transit item quantity is calculated for each time period, wherein the on-transit item quantity is equal to a total item quantity of the on-transit package having an arrival time that falls between a start and end time of each time period.
The prediction module 221c inputs the time series into an ARIMA model and predicts based on the time series (historical data) to generate an ARIMA prediction value. The prediction module 221c then performs a weighted average of the ARIMA prediction value and the real-time in-transit volume to obtain a final prediction result (e.g., a predicted volume for a time period of 0:00-0:30). That is, the final prediction result is equal to the sum of the ARIMA prediction value multiplied by a first parameter and the real-time in-transit volume multiplied by a second parameter.
In addition, the first parameter and the second parameter may be adjusted/modified by the training and verifying module 221d, so that the final prediction result may be closer to the actual amount of the piece. The training and verification module 221d uses the first 80% of the historical data of the quantity as a training set and the last 20% as a verification set. Wherein the first 80% of the training set is trained as training input values (i.e., time series of inputs) and the last 20% of the training set is trained as training true values. Specifically, the first 80% of the training set is used as an input time sequence for training, x is used as x, and the last 20% of the training set is used as a true value for training, y is used as y. The MSE (mean square error) loss function is used, and the formula is as follows:
Where N represents the number of time series,
Representing the predicted value of the ith half hour period,
Y i represents the true value of the ith half hour period.
And calculating the gradient of the loss function on the ARIMA prediction result and the weight of the in-transit piece quantity, and modifying the weights, namely modifying the first parameter and the second parameter. The parameter modification formula is:
wherein W' represents the parameters of the modified ARIMA predicted value and the real-time in-transit volume,
W represents parameters of the pre-modification ARIMA forecast and the real-time in-transit volume,
Representing the gradient of the parameter w.
Training is continuously performed through the training input value until the training input value is used up. Therefore, the first parameter and the second parameter after training can be more close to the actual condition through carrying out a large amount of training on the historical data of the quantity, and the final prediction result is more accurate.
Further, the training and verification module 221d performs verification using the first 80% of the verification set as an input value for verification and the second 20% of the verification set as a true value for verification. Specifically, the first 80% of the verification set is used as the verification input time sequence, and the last 20% of the training set is used as the verification true value. The ARIMA model after training is input with an input time series for verification (input values for verification) and a corresponding final predicted value for verification is obtained, and an error is calculated using the loss function as verification. In this way, the first parameter and the second parameter after training can be further determined, and actual prediction can be performed.
It should be appreciated that each half hour period may have a respective first parameter and second parameter after a respective training. In this way, the arrival volume historical data and the real-time in-transit volume of other half-hour time periods do not affect the final predicted value of the half-hour time period. For example, two half-hour time periods of 00:00-00:30 and 12:00-12:30, the first and second parameters respectively, which are generated by training due to the difference of the respective arrival volume historical data and the real-time in-transit volume.
In practical application, if the arrival amount of the half-hour time period of 00:00-00:30 on two days of 9.8.2019 and 9.9.9 needs to be predicted, the prediction module 221c performs weighted average with the first parameter and the second parameter corresponding to the half-hour time period by using the ARIMA predicted value of 00:00-00:30 in the half-hour time period and the real-time in-transit amount to obtain the final prediction result corresponding to the half-hour time period of 00:00-00:30. Thus, the accuracy of predicting the piece quantity can be greatly improved.
Finally, the difference between the predicted amount of work (final predicted result) and the actual amount of work can be further confirmed by the predicted effect evaluation module 221 e. The following table is an example, and is a mean square error arrangement of the predicted and actual arrival quantities of the transfer field after 3 months of actual use of the embodiment of the present application.
| 6 Months of | 7 Months of | 8 Months of | |
| Mean square error | 98.2 | 89.1 | 92.3 |
It can be found that the error between the predicted piece quantity and the actual piece quantity is limited, and basically the prediction accuracy is nearly 9, which indicates that the predicted piece quantity of the intermediate transfer can be accurately provided by the embodiment of the application.
It will be appreciated by those of ordinary skill in the art that all or part of the steps of the various methods of the above embodiments may be performed by the computer program or by controlling related hardware by the computer program, which may be stored in a computer readable storage medium and loaded and executed by the processor.
In addition, the embodiment of the application also provides a storage medium, which is used for storing a computer program, and the computer program is loaded by a processor to execute the steps in any of the logistics to piece quantity prediction methods provided by the embodiment of the application. The storage medium may include: read Only Memory (ROM), random access Memory (RAM, random Access Memory), magnetic or optical disk, and the like. The computer program stored in the storage medium may perform the following steps when loaded by a processor:
dividing a fixed period of time for which prediction is desired into a plurality of time periods;
Acquiring historical data of the arrival quantity corresponding to each time period;
calculating real-time on-road part quantity corresponding to each time period;
predicting the arrival quantity of each time period based on the arrival quantity historical data corresponding to each time period through an ARIMA model, and generating an ARIMA predicted value for each time period; and
And carrying out weighted average on the ARIMA predicted value and the real-time on-road part quantity to obtain a final predicted result, wherein the final predicted result is equal to the sum of the ARIMA predicted value multiplied by a first parameter and the real-time on-road part quantity multiplied by a second parameter.
In the foregoing embodiments, the descriptions of the embodiments are focused on, and the portions of one embodiment that are not described in detail in the foregoing embodiments may be referred to in the foregoing detailed description of other embodiments, which are not described herein again.
In the implementation, each unit or structure may be implemented as an independent entity, or may be implemented as the same entity or several entities in any combination, and the implementation of each unit or structure may be referred to the foregoing method embodiments and will not be repeated herein.
The specific implementation of each operation above may be referred to the previous embodiments, and will not be described herein.
The method, the equipment or the storage medium for predicting the logistics to the stock quantity can be used for predicting the logistics to the stock quantity in each half hour period, such as the stock quantity prediction of the transfer in the middle. Therefore, the resource allocation can be adjusted according to the predicted piece quantity of each half hour period, and the operation efficiency is further improved.
The above description is made in detail of a method for predicting the amount of the material flow to the part, a device for predicting the amount of the material flow to the part or a storage medium provided by the embodiment of the present application, and specific examples are applied to illustrate the principle and implementation of the present application, and the above description of the embodiment is only used to help understand the method and core idea of the present application; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in light of the ideas of the present application, the present description should not be construed as limiting the present application.
Claims (9)
1. A method of predicting a quantity of a material to be fed, comprising:
dividing a fixed period of time for which prediction is desired into a plurality of time periods;
Acquiring historical data of the arrival quantity corresponding to each time period;
calculating real-time on-road part quantity corresponding to each time period;
Predicting the arrival quantity of each time period based on the arrival quantity historical data corresponding to each time period through a time sequence prediction model, and generating a time sequence predicted value for each time period; and
Performing weighted average on the time sequence predicted value and the real-time on-road component quantity to obtain a final predicted result, wherein the final predicted result is equal to the sum of the time sequence predicted value multiplied by a first parameter and the real-time on-road component quantity multiplied by a second parameter;
The step of calculating the real-time on-the-way quantity corresponding to each time period comprises the following steps:
Locating each in-transit parcel for each time period;
calculating the navigation time of each in-transit package arrival in each time period;
Calculating the residence time of each in-transit package in each time period;
calculating an arrival time for each in-transit package within each time period, wherein the arrival time is equal to a current time plus the navigation time plus the hold-up time; and
An on-transit item quantity is calculated for each time period, wherein the on-transit item quantity is equal to a total item quantity of the on-transit package having an arrival time that falls between a start and end time of each time period.
2. The method for predicting the amount of material flowing to a piece according to claim 1, wherein the time-series prediction model is an autoregressive integral moving average model; the step of dividing a fixed period of time for which prediction is desired into a plurality of time periods includes:
At half hour intervals, a fixed 24 hour time period was divided into 48 time periods.
3. The method of claim 2, wherein the autoregressive integrated moving average model is trained by:
Taking the first 80% of the historical data of the arrival quantity corresponding to the 48 time periods as a training set;
training with the first 80% of the training set as training input values and the last 20% of the training set as training true values; and
The first parameter and the second parameter are obtained by means of a mean square error loss function.
4. A method of predicting the amount of material to be fed according to claim 3, wherein the autoregressive integral moving average model is verified by:
In the history data of the arrival quantity corresponding to the 48 time periods, the last 20% is taken as a verification set; and
The verification is performed with the first 80% of the verification set as an input value for verification and the last 20% of the verification set as a true value for verification.
5. The method for predicting the amount of material to be fed according to claim 1, further comprising:
One or more time periods are selected according to requirements, and the selected one or more time periods are predicted based on the arrival amount historical data corresponding to each time period.
6. A device for predicting a quantity of a stream to be processed, comprising a processor and a memory, wherein the processor invokes a computer program in the memory to perform the method for predicting a quantity of a stream to be processed according to any one of claims 1 to 5.
7. The apparatus for predicting volume of material to be fed as set forth in claim 6, wherein said computer program comprises:
The feature selection module is used for dividing a fixed period of time expected to be predicted into a plurality of time periods and acquiring the arrival quantity historical data corresponding to each time period;
the real-time on-road part quantity calculating module is used for calculating the real-time on-road part quantity corresponding to each time period; and
The prediction module is used for predicting the arrival quantity of each time period based on the arrival quantity historical data corresponding to each time period through a time sequence prediction model, generating a time sequence prediction value for each time period, and carrying out weighted average on the time sequence prediction value and the real-time on-way component quantity to obtain a final prediction result, wherein the final prediction result is equal to the sum of the time sequence prediction value multiplied by a first parameter and the real-time on-way component quantity multiplied by a second parameter;
The step of calculating the real-time on-the-way quantity corresponding to each time period comprises the following steps:
Locating each in-transit parcel for each time period;
calculating the navigation time of each in-transit package arrival in each time period;
Calculating the residence time of each in-transit package in each time period;
calculating an arrival time for each in-transit package within each time period, wherein the arrival time is equal to a current time plus the navigation time plus the hold-up time; and
An on-transit item quantity is calculated for each time period, wherein the on-transit item quantity is equal to a total item quantity of the on-transit package having an arrival time that falls between a start and end time of each time period.
8. The logistics in quantity prediction apparatus of claim 7, wherein the computer program further comprises:
the training and verifying module is used for training the time sequence prediction model and verifying the time sequence prediction model.
9. A storage medium having stored thereon a computer program to be loaded by a processor for performing the steps of the method of predicting a commodity circulation to commodity circulation volume according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911293320.XA CN112990526B (en) | 2019-12-16 | 2019-12-16 | Method, apparatus and storage medium for predicting amount of material to be delivered |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911293320.XA CN112990526B (en) | 2019-12-16 | 2019-12-16 | Method, apparatus and storage medium for predicting amount of material to be delivered |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112990526A CN112990526A (en) | 2021-06-18 |
| CN112990526B true CN112990526B (en) | 2024-08-02 |
Family
ID=76343229
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911293320.XA Active CN112990526B (en) | 2019-12-16 | 2019-12-16 | Method, apparatus and storage medium for predicting amount of material to be delivered |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112990526B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109583625A (en) * | 2018-10-19 | 2019-04-05 | 顺丰科技有限公司 | One kind pulling part amount prediction technique, system, equipment and storage medium |
| CN109767024A (en) * | 2017-11-09 | 2019-05-17 | 顺丰科技有限公司 | Part amount prediction technique and device, equipment and storage medium |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5068797A (en) * | 1989-10-03 | 1991-11-26 | Pitney Bowes Inc. | Optimizing mail delivery systems by routing |
| US6611726B1 (en) * | 1999-09-17 | 2003-08-26 | Carl E. Crosswhite | Method for determining optimal time series forecasting parameters |
| US20060106664A1 (en) * | 2004-11-17 | 2006-05-18 | International Business Machines Corporation | Method, system, and storage medium for developing a forecast of goods and services |
| CN109961165B (en) * | 2017-12-25 | 2023-11-28 | 顺丰科技有限公司 | Method, device, equipment and storage medium for predicting part quantity |
| CN110019401B (en) * | 2017-12-25 | 2024-04-05 | 顺丰科技有限公司 | Method, device, equipment and storage medium for predicting part quantity |
| CN109978208A (en) * | 2017-12-28 | 2019-07-05 | 顺丰科技有限公司 | A kind of time series data prediction technique, device and computer storage medium |
| CN109784820A (en) * | 2019-03-28 | 2019-05-21 | 上海中通吉网络技术有限公司 | A kind of express delivery routing intelligence solution, device, terminal device and storage medium |
| CN110428206B (en) * | 2019-08-08 | 2023-07-04 | 上海中通吉网络技术有限公司 | Intelligent solutions, devices, storage media and electronic equipment for express terminal delivery |
-
2019
- 2019-12-16 CN CN201911293320.XA patent/CN112990526B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109767024A (en) * | 2017-11-09 | 2019-05-17 | 顺丰科技有限公司 | Part amount prediction technique and device, equipment and storage medium |
| CN109583625A (en) * | 2018-10-19 | 2019-04-05 | 顺丰科技有限公司 | One kind pulling part amount prediction technique, system, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112990526A (en) | 2021-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110414731B (en) | Order distribution method and device, computer readable storage medium and electronic equipment | |
| CN107123004A (en) | Commodity dynamic pricing data processing method and system | |
| CN112529301A (en) | Power consumption prediction method, equipment and storage medium | |
| US8494823B2 (en) | Method and system for determining service area of supply chain by simulating service cycle time | |
| CN115409439B (en) | Multi-model supply chain scheduling method and electronic equipment based on improved ant colony algorithm | |
| US8560359B2 (en) | System and methods for modeling consequences of events | |
| KR20220115357A (en) | A method and apparatus for generating future demand forecast data based on attention mechanism | |
| CN108416619A (en) | A kind of consumption interval time prediction technique, device and readable storage medium storing program for executing | |
| CN116562588A (en) | Enterprise supply chain analysis system, method and equipment based on ERP | |
| CN114462298A (en) | Power metering asset management method, device, equipment and storage medium | |
| US20200285990A1 (en) | Sequence, frequency, and time interval based journey recommendation | |
| US8352407B2 (en) | Systems and methods for modeling consequences of events | |
| Jung et al. | A workflow scheduling technique using genetic algorithm in spot instance-based cloud | |
| CN108197825A (en) | System scheduling method and device | |
| CN112990526B (en) | Method, apparatus and storage medium for predicting amount of material to be delivered | |
| CN116862574A (en) | Product demand prediction method, device, electronic equipment and readable storage medium | |
| Volpe et al. | A deep reinforcement learning approach for competitive task assignment in enterprise blockchain | |
| US11036516B2 (en) | Parallel distributed processing control system, program, and parallel distributed processing control method | |
| CN109858787B (en) | Energy scheduling management method and device, readable medium and electronic equipment | |
| Kapur et al. | Software reliability growth and innovation diffusion models: An interface | |
| CN110070247A (en) | Logistics information processing method and device | |
| Lutsenko et al. | A METHOD TO FORM CONTROL OVER QUEUING SYSTEMS TAKING INTO CONSIDERATION THE PROBABILISTIC CHARACTER OF DEMAND. | |
| CN113962618A (en) | Information display method and device, storage medium and electronic equipment | |
| US8688488B2 (en) | Method and apparatus for the prediction of order turnaround time in an information verification system | |
| CN112070561A (en) | Electronic ticket recommendation method and device |
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 |