CN109146134A - Part amount prediction technique, system, equipment and storage medium are pulled in a kind of peak - Google Patents
Part amount prediction technique, system, equipment and storage medium are pulled in a kind of peak Download PDFInfo
- Publication number
- CN109146134A CN109146134A CN201810788118.3A CN201810788118A CN109146134A CN 109146134 A CN109146134 A CN 109146134A CN 201810788118 A CN201810788118 A CN 201810788118A CN 109146134 A CN109146134 A CN 109146134A
- Authority
- CN
- China
- Prior art keywords
- part amount
- period
- pull
- time cycle
- predicted
- 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.)
- Pending
Links
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
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/083—Shipping
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)
- Supply And Distribution Of Alternating Current (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to a kind of peaks to pull part amount prediction technique, system, equipment and storage medium, it according to the historical time period and its accordingly pulls part amount and arbitrary period variable pulls part amount and standardized value in the time cycle preset relation obtains the first standardized value, part amount peak period is pulled using each history to pull part amount peak period corresponding time cycle to be predicted and its pull the similar feature of part measure feature, the second standardized value of time cycle to be predicted is predicted by the first standardized value, and part amount is pulled according to the preset relation backward prediction time cycle to be predicted that the second standardized value pulls part amount and standardized value according to arbitrary period variable in the time cycle, the method of the present invention dedicated for predict it is to be predicted pull the part amount peak period corresponding time cycle pull part amount, the characteristics of pulling part amount peak period is included in prediction completely, science, it is objective, part is accordingly pulled in high-precision prediction Part amount is pulled in amount peak period, has the function of early warning in advance, is reasonably dispatched convenient for logistics company, prevents wharf explosion.
Description
Technical field
The present invention relates to logistics more particularly to a kind of peak to pull part amount prediction technique, system, equipment and storage medium.
Background technique
It with e-commerce and the development of logistic industry, pulls part amount and increases rapidly, especially legal festivals and holidays, shopping section
Express delivery amount increases substantially during equal festivals or holidays, and " wharf explosion " happens occasionally, such as double 11 peak part amount is about daily
2-3 times.The Accurate Prediction on peak facilitates enterprise's rational allocation resource, prepares to respond in advance.The non-stationary of data, mutation
Property etc. features presence, although so that current prediction technique includes that recurrence, time series predicting model etc. are pre- to daily part amount
It is more accurate but unsatisfactory to the prediction effect on peak to survey.
(1) the peak part amount in each year is there are magnitude differences, and data to return there are non-stationary, mutability, when
Sequence prediction model etc. is difficult to carry out the prediction on peak well.
(2) each festivals or holidays feature is different, and after legal festivals and holidays peak appears in section prosthomere, and the peak for section class of doing shopping goes out
In saving now and part amount is influenced by promotion dynamics, active duration etc.；There is (example: 2015 618 and end simultaneously in festivals or holidays
Noon) cross influence is brought, current model is difficult to account for factors, and some models decomposed based on timing are to weigh
Weight form accounts for each factor, but there are subjectivities when determining influences option weight.
(3) when model carries out outside forecast at present, increase with extrapolation duration, decline by a big margin.
Summary of the invention
In order to solve the above-mentioned technical problem, the purpose of the present invention is to provide a kind of peak pull part amount prediction technique, system,
Equipment and storage medium.
According to an aspect of the invention, there is provided part amount prediction technique is pulled on a kind of peak, comprising the following steps:
Configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation the historical time period,
Acquire the historical time period pulls part amount；
By the historical time period and its pulls part amount part amount and standardized value are pulled according to arbitrary period variable in the time cycle
Preset relation is standardized to obtain the first standardized value；
The second standardized value of time cycle to be predicted is obtained according to the first standardized value；
Second standardized value is pulled to the default pass of part amount and standardized value according to arbitrary period variable in the time cycle
Be the time cycle to be predicted pull part amount to get it is to be predicted pull part amount peak period pull part amount.
Arbitrary period variable pulls the preset relation of part amount and standardized value in time cycle are as follows:
Wherein,
xiPart amount is pulled for arbitrary period variable in the time cycle,
Max (X) is the maximum value that arbitrary period variable pulls part amount in the time cycle,
Min (X) is the minimum value that arbitrary period variable pulls part amount in the time cycle.
Configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation the historical time period,
Acquire the historical time period pulls part amount, comprising:
Configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation the historical time period,
Acquire the historical time period pull part amount and period first time pull part amount, decomposite periodic component, and according to described
Periodic component, must reject the historical time period of Zhou Xiaoying pulls part amount, i.e., history pulls part amount, and when rejecting the first of Zhou Xiaoying
Between the period pull part amount, i.e., first pull part amount.
Acquire the historical time period pull part amount and period first time pull part amount, decomposite periodic component, wrap
It includes:
Acquire the historical time period pulls pulling part amount and configuring preset period of time for part amount and period first time, described
Historical time period, period first time include preset period of time；
According to the historical time period and its pull part amount, period first time and its pull part amount, preset period of time obtains week
Phase ingredient.
By the historical time period and its pulls part amount part amount and standardized value are pulled according to arbitrary period variable in the time cycle
Preset relation is standardized to obtain the first standardized value, comprising:
Pull that part amount, arbitrary period variable pulls part amount and standardized value in the time cycle based on historical time period and its history
Preset relation be standardized the historical time period the first standardized value.
The second standardized value of time cycle to be predicted is obtained according to the first standardized value, comprising:
The second standardized value of time cycle to be predicted is obtained according to first standardized value in historical time period.
Second standardized value is pulled to the default pass of part amount and standardized value according to arbitrary period variable in the time cycle
Be the time cycle to be predicted pulls part amount, comprising:
Part amount, the second standardized value are pulled according to described first and arbitrary period variable pulls part amount initially most in the time cycle
Big value, the preset relation of initial minimum must pull part amount maximum value, minimum value, it is described pull part amount maximum value, minimum value as to
The predicted time period pulls part amount maximum value, minimum value；
Part amount maximum value, minimum value are pulled in conjunction with second standardized value and time cycle according to the time cycle to be predicted
What the preset relation that interior arbitrary period variable pulls part amount and standardized value must reject the time cycle to be predicted of Zhou Xiaoying pulls part amount,
I.e. part amount is pulled in prediction first；
Part amount is pulled into prediction first and pulls part amount based on what periodic component obtained the time cycle to be predicted.
Arbitrary period variable pulls part amount original maximum, the preset relation of initial minimum in time cycle are as follows:
Max (X) > min (X) > 0.
According to another aspect of the present invention, it provides a kind of peak and pulls part amount forecasting system, comprising:
Configuration and acquisition unit, for configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when
In the historical time period of foundation, acquire the historical time period pulls part amount；
First standardized value acquiring unit is configured to the historical time period and its pulls part amount according in the time cycle
Segment variable pulls part amount when meaning and the preset relation of standardized value is standardized to obtain the first standardized value；
Second standardized value predicting unit is configured to obtain the second mark of time cycle to be predicted according to the first standardized value
Quasi-ization value；
Part amount predicting unit is pulled, is configured to pull second standardized value according to arbitrary period variable in the time cycle
Part amount and the preset relation of standardized value obtain the time cycle to be predicted pull part amount to get it is to be predicted pull part amount peak period pull part
Amount.
Arbitrary period variable pulls the preset relation of part amount and standardized value in time cycle are as follows:
Wherein,
xiPart amount is pulled for arbitrary period variable in the time cycle,
Max (X) is the maximum value that arbitrary period variable pulls part amount in the time cycle,
Min (X) is the minimum value that arbitrary period variable pulls part amount in the time cycle.
Configuration and acquisition unit are also used to:
Configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation the historical time period,
Acquire the historical time period pull part amount and period first time pull part amount, decomposite periodic component, and according to described
Periodic component, must reject the historical time period of Zhou Xiaoying pulls part amount, i.e., history pulls part amount, and when rejecting the first of Zhou Xiaoying
Between the period pull part amount, i.e., first pull part amount.
Configuration and acquisition unit are also used to:
Acquire the historical time period pulls pulling part amount and configuring preset period of time for part amount and period first time, described
Historical time period, period first time include preset period of time；
According to the historical time period and its pull part amount, period first time and its pull part amount, preset period of time obtains week
Phase ingredient.
First standardized value acquiring unit is also configured to:
Pull that part amount, arbitrary period variable pulls part amount and standardized value in the time cycle based on historical time period and its history
Preset relation be standardized the historical time period the first standardized value.
Second standardized value predicting unit is also configured to:
The second standardized value of time cycle to be predicted is obtained according to first standardized value in historical time period.
Part amount predicting unit is pulled to be also configured to:
Part amount, the second standardized value are pulled according to described first and arbitrary period variable pulls part amount initially most in the time cycle
Big value, the preset relation of initial minimum must pull part amount maximum value, minimum value, it is described pull part amount maximum value, minimum value as to
The predicted time period pulls part amount maximum value, minimum value；
Part amount maximum value, minimum value are pulled in conjunction with second standardized value and time cycle according to the time cycle to be predicted
What the preset relation that interior arbitrary period variable pulls part amount and standardized value must reject the time cycle to be predicted of Zhou Xiaoying pulls part amount,
I.e. part amount is pulled in prediction first；
Part amount is pulled into prediction first and pulls part amount based on what periodic component obtained the time cycle to be predicted.
Arbitrary period variable pulls part amount original maximum, the preset relation of initial minimum in time cycle are as follows:
Max (X) > min (X) > 0.
According to another aspect of the present invention, a kind of equipment is provided, the equipment includes:
One or more processors；
Memory, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of places
It manages device and executes as above described in any item methods.
According to another aspect of the present invention, provide it is a kind of be stored with the computer-readable of computer program and deposit medium,
As above described in any item methods are realized when the program is executed by processor.
Compared with prior art, the invention has the following advantages:
1, part amount prediction technique is pulled on the exemplary peak of the present invention, configure it is to be predicted pull part amount peak period it is corresponding to be predicted when
Between the period and prediction when foundation the historical time period, acquire the historical time period pulls part amount；By the historical time period
And its it pulls part amount and is standardized according to the preset relation that arbitrary period variable pulls part amount and standardized value in the time cycle
Obtain the first standardized value；The second standardized value of time cycle to be predicted is obtained according to the first standardized value；By second standard
Change value obtains pulling for time cycle to be predicted according to the preset relation that arbitrary period variable in the time cycle pulls part amount and standardized value
Part amount to get it is to be predicted pull part amount peak period pull part amount.According to the historical time period and its accordingly pull part amount and time cycle
Interior arbitrary period variable pulls part amount and the preset relation of standardized value obtains the first standardized value, pulls part amount peak period using each history
It pulls part amount peak period corresponding time cycle to be predicted and its pulls the similar feature of part measure feature, pass through the first standardized value and predict
The second standardized value of time cycle to be predicted, and part is pulled according to arbitrary period variable in the time cycle according to the second standardized value
Amount and preset relation backward prediction time cycle to be predicted of standardized value pull part amount, the method for the present invention dedicated for prediction to
Being included in prediction completely the characteristics of pulling part amount, pull part amount peak period for part amount peak period corresponding time cycle, science, visitor are pulled in prediction
See, high-precision prediction accordingly pulls part amount peak period pulls part amount, determined without carrying out weight, caused by reducing subjective factor
Prediction is not accurate, has the function of early warning in advance, carries out reasonable scheduling of resource convenient for logistics company, prevents wharf explosion, promotes visitor
Family experience.
2, part amount forecasting system is pulled on the exemplary peak of the present invention, configuration and acquisition unit configuration is to be predicted pulls part amount peak period
In the historical time period of foundation when corresponding time cycle to be predicted and prediction, acquire the historical time period pulls part amount；
First standardized value acquiring unit is by the historical time period and its pulls part amount and pulls part amount according to arbitrary period variable in the time cycle
The first standardized value is standardized to obtain with the preset relation of standardized value；Second standardized value predicting unit is according to first
Standardized value obtains the second standardized value of time cycle to be predicted；Pull part amount predicting unit by second standardized value according to when
Between arbitrary period variable pulls part amount and standardized value in the period preset relation obtain the time cycle to be predicted pull part amount to get to
What part amount peak period was pulled in prediction pulls part amount.Each unit composition is simple, high using part amount is accordingly pulled by mutual cooperation
The peak phase corresponding time cycle and its similar feature of part measure feature is pulled, the part amount peak period corresponding time cycle is pulled by history
And its pull part amount, predict it is to be predicted pull the part amount peak period corresponding time cycle pull part amount, precision of prediction is high.
3, it the exemplary equipment of the present invention and is stored with the computer-readable of computer program and deposits medium, pulled using above-mentioned peak
Part amount Predicting Technique, by history pull the part amount peak period corresponding time cycle and its pull part amount the first standardized value prediction to
Second standardized value in predicted time period, and part amount is pulled according to the second standardized value backward prediction time cycle to be predicted,
Prediction accuracy is high, is worthy to be popularized.
Detailed description of the invention
Fig. 1 is flow chart of the present invention；
Fig. 2 pulls part amount tendency chart for the rejecting of embodiment one Zhou Xiaoying's；
Fig. 3 is to pull within embodiment 1 years the prediction of part amount and actual value comparative diagram；
Fig. 4 is 11 prediction effect figure of embodiment a pair of.
Specific embodiment
In order to be better understood by technical solution of the present invention, combined with specific embodiments below, Figure of description is to the present invention
It is described further.
Embodiment one:
It present embodiments provides a kind of peak and pulls part amount prediction technique, comprising the following steps:
S1, configuration it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation historical time week
Phase, acquire the historical time period pulls part amount.
It specifically includes:
Configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation the historical time period,
Acquire the historical time period pull part amount and period first time pull part amount, decomposite periodic component, and according to described
Periodic component, must reject the historical time period of Zhou Xiaoying pulls part amount, i.e., history pulls part amount, and when rejecting the first of Zhou Xiaoying
Between the period pull part amount, i.e., first pull part amount.
Acquire the historical time period pull part amount and period first time pull part amount, decomposite periodic component, wrap
It includes:
(1) acquire historical time period pulls pulling part amount and configuring preset period of time for part amount and period first time,
The historical time period, period first time include preset period of time；
(2) according to the historical time period and its pull part amount, period first time and its pull part amount, preset period of time obtains
Periodic component.
S2, it by the historical time period and its pulls part amount part amount and standardized value is pulled according to arbitrary period variable in the time cycle
Preset relation be standardized to obtain the first standardized value.
It specifically includes:
Pull that part amount, arbitrary period variable pulls part amount and standardized value in the time cycle based on historical time period and its history
Preset relation be standardized the historical time period the first standardized value.
S3, the second standardized value that the time cycle to be predicted is obtained according to the first standardized value.
The second standardized value of time cycle to be predicted is obtained with specific reference to first standardized value in historical time period.
S4, second standardized value is pulled into the default of part amount and standardized value according to arbitrary period variable in the time cycle
Relationship obtain the time cycle to be predicted pull part amount to get it is to be predicted pull part amount peak period pull part amount.
Arbitrary period variable pulls the preset relation of part amount and standardized value in time cycle are as follows:
Wherein,
xiPart amount is pulled for arbitrary period variable in the time cycle,
Max (X) is the maximum value that arbitrary period variable pulls part amount in the time cycle,
Min (X) is the minimum value that arbitrary period variable pulls part amount in the time cycle.
S4 is specifically included:
S41, part amount, the second standardized value are pulled according to described first and at the beginning of arbitrary period variable pulls part amount in the time cycle
Beginning maximum value, the preset relation of initial minimum must pull part amount maximum value, minimum value, described to pull part amount maximum value, minimum value work
Part amount maximum value, minimum value are pulled for the time cycle to be predicted,
Arbitrary period variable pulls part amount original maximum, the preset relation of initial minimum in time cycle are as follows:
Max (X) > min (X) > 0；
S42, part amount maximum value, minimum value are pulled in conjunction with second standardized value and time according to the time cycle to be predicted
The preset relation that arbitrary period variable pulls part amount and standardized value in period must reject pulling for the time cycle to be predicted of Zhou Xiaoying
Part amount predicts that first pulls part amount；
S43, it prediction first is pulled to part amount pulls part amount based on what periodic component obtained the time cycle to be predicted.
Part amount prediction technique is pulled on above-mentioned peak, is illustrated by taking the prediction of double 11 peaks as an example.
1, basic: double 11 data analysis found that pulling part amount rejecting Zhou Xiaoying for the whole network and being standardized (0-1
Standardization processing) after, related coefficient > 0.97 over the years, and nearby trend is almost the same on peak, as shown in Figure 2.
2, thought: by the 0-1 normalized number evidence of historical years, i.e. the first standardized value predicts following 0-1 standardization
Value, i.e. the second standardized value restore original and pull part amount data.
3, specific steps (by taking prediction 2017 as an example):
Input: 2013-2016 10.1-12.31,10.1-10.31 in 2017 pull part amount data, wherein 2013-
10.1-12.31 in 2016 is the historical time period to be predicted for pulling foundation when part amount peak period is predicted, 10.1-10.31 in 2017
For the period 1.
Output: 11.1-12.31's in 2017 pulls part amount predicted value, for it is to be predicted pull part amount peak period pull part amount.
(1) it calculates " week " periodic component S: 2013-2016 10.1-12.31,2017.10.1-10.31 being taken to pull part amount
Data Y extracts annual first Monday to the data between a last Sunday, decomposites to calculate global periodic component
Periodic component, and calculate 2013-2016 and reject Zhou Xiaoying and pull part amount X, i.e. history pulls part amount.
Period (preset period of time) is chosen when 1 periodic component of table calculates
| Start | 2013.10.7 | 2014.10.6 | 2015.10.5 | 2016.10.3 | 2017.10.2 |
| Terminate | 2013.12.29 | 2014.12.28 | 2015.12.27 | 2016.12.25 | 2017.10.29 |
(2) the 0-1 normative value R of part amount is pulled after calculating 2013-2016 rejecting Zhou Xiaoying:
Wherein,
xiPart amount is pulled for arbitrary period variable in the time cycle,
Max (X) is the maximum value that arbitrary period variable pulls part amount in the time cycle,
Min (X) is the minimum value that arbitrary period variable pulls part amount in the time cycle.
(3) the 0-1 normative value R2017 of predicted time 10.1-12.31 in 2017 is predicted:
R2017=(R2014+R201s+R2016)/3
(4) estimate maximum value max2017, the minimum value min2017 of second time period 10.8-11.30 in 2017: according to
(2) formula is known in, appoints and takes two o'clock (x0, r0)、(x1, r1) can obtain
It releases:
The rejecting Zhou Xiaoying of known 2017.10.1-10.31 pulls part amount and 0-1 normative value, and permutation and combination any two points are equal
It substitutes into formula and calculates one group of maximum, minimum value, retain all values for meeting max (X) > min (X) > 0, take median as most
Greatly, the estimation of minimum value.
(5) provide the rejecting Zhou Xiaoying of 11.1-12.31 in 2017 pulls part amount:
X2017=R2017*(max2017-min2017)+min2017
(6) provide 11.1-12.31 in 2017 pulls part amount predicted value.
Part amount predicted value is pulled obtained by the above-mentioned prediction technique of real example:
At present by the analysis of real example, it is found that the model predicts that upper accuracy rate is higher on festivals or holidays peak.
(1) 2017 year minimum, maximum value prediction (core) relative error only 0.35%, 0.19%.
Table 2:2017 pulls part amount maximum value, the relative error of minimum value predicted value and respective value in 2016
| Maximum value | Minimum value | |
| 2017 | 0.35% | 0.19% |
| 2016 | 0.33% | 0.17% |
(2) double 11 mountain portions predictions are the most accurate, are almost overlapped with actual value, as shown in Figure 3.
2016, the MAPE of 11-12 month prediction result in 2017 is both less than 10%, double 11 mountain portions (11.9-11.17)
MAPE < 3%, precision is higher, as shown in Figure 4；Daily precision of prediction is lower, but still MAPE < 15%.In " 0-1 normative value
Changed compared with historical years small " in the case of, precision of prediction is higher；It is predicted suitable for peak.
The present embodiment additionally provides a kind of peak and pulls part amount forecasting system, comprising:
Configuration and acquisition unit,
For configuring: it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation historical time
In the period, acquire the historical time period pulls part amount,
It is also used to:
Configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation the historical time period,
Acquire the historical time period pull part amount and period first time pull part amount, decomposite periodic component, and according to described
Periodic component, must reject the historical time period of Zhou Xiaoying pulls part amount, i.e., history pulls part amount, and when rejecting the first of Zhou Xiaoying
Between the period pull part amount, i.e., first pull part amount.
It is also used to:
Acquire the historical time period pulls pulling part amount and configuring preset period of time for part amount and period first time, described
Historical time period, period first time include preset period of time；
According to the historical time period and its pull part amount, period first time and its pull part amount, preset period of time obtains week
Phase ingredient；
First standardized value acquiring unit,
It is configured to:
By the historical time period and its pulls part amount part amount and standardized value are pulled according to arbitrary period variable in the time cycle
Preset relation is standardized to obtain the first standardized value,
It is also configured to:
Pull that part amount, arbitrary period variable pulls part amount and standardized value in the time cycle based on historical time period and its history
Preset relation be standardized the historical time period the first standardized value；
Second standardized value predicting unit is configured to obtain the second mark of time cycle to be predicted according to the first standardized value
Quasi-ization value obtains the second standardized value of time cycle to be predicted with specific reference to first standardized value in historical time period；
Part amount predicting unit is pulled, is configured to pull second standardized value according to arbitrary period variable in the time cycle
Part amount and the preset relation of standardized value obtain the time cycle to be predicted pull part amount to get it is to be predicted pull part amount peak period pull part
Amount.
Arbitrary period variable pulls the preset relation of part amount and standardized value in time cycle are as follows:
Wherein,
xiPart amount is pulled for arbitrary period variable in the time cycle,
Max (X) is the maximum value that arbitrary period variable pulls part amount in the time cycle,
Min (X) is the minimum value that arbitrary period variable pulls part amount in the time cycle,
riFor the standardized value of arbitrary period in the time cycle.
Part amount predicting unit is pulled to be also configured to:
Part amount, the second standardized value are pulled according to described first and arbitrary period variable pulls part amount initially most in the time cycle
Big value, the preset relation of initial minimum must pull part amount maximum value, minimum value, it is described pull part amount maximum value, minimum value as to
The predicted time period pulls part amount maximum value, minimum value；
Part amount maximum value, minimum value are pulled in conjunction with second standardized value and time cycle according to the time cycle to be predicted
What the preset relation that interior arbitrary period variable pulls part amount and standardized value must reject the time cycle to be predicted of Zhou Xiaoying pulls part amount,
I.e. part amount is pulled in prediction first；
Part amount is pulled into prediction first and pulls part amount based on what periodic component obtained the time cycle to be predicted.
Arbitrary period variable pulls part amount original maximum, the preset relation of initial minimum in time cycle are as follows:
Max (X) > min (X) > 0.
A kind of equipment is present embodiments provided, which includes:
One or more processors；
Memory, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of places
It manages device and executes method as described in any one of the above embodiments.
A kind of computer readable storage medium for being stored with computer program is present embodiments provided, the program is by processor
As above described in any item methods are realized when execution.
Embodiment two
The feature that the present embodiment is the same as example 1 repeats no more, and the present embodiment feature different from embodiment one exists
In:
The sequential value after 0-1 standardization is predicted by ARIMA method.
Embodiment three
The feature that the present embodiment is the same as example 1 repeats no more, and the present embodiment feature different from embodiment one exists
In:
The sequential value after 0-1 standardization, i.e. the second standardized value are predicted by homing method.
Example IV
The feature that the present embodiment is the same as example 1 repeats no more, and the present embodiment feature different from embodiment one exists
In:
The sequential value after 0-1 standardization, i.e. the second standardized value are predicted by LSTM method.
Embodiment five
The feature that the present embodiment is the same as example 1 repeats no more, and the present embodiment feature different from embodiment one exists
In:
Part amount prediction technique is pulled on above-mentioned peak, is illustrated by taking the prediction of double 12 peaks as an example.
1, basic: double 12 data analysis found that pulling part amount rejecting Zhou Xiaoying for the whole network and being standardized (0-1
Standardization processing).
2, thought: by the 0-1 normalized number evidence of historical years, i.e. the first standardized value predicts following 0-1 standardization
Value, i.e. the second standardized value restore original and pull part amount data.
3, specific steps (by taking prediction 2017 as an example):
Input: 2013-2016 11.1-12.31,11.1-11.30 in 2017 pull part amount data, wherein 2013-
11.1-12.31 in 2016 is the historical time period to be predicted for pulling foundation when part amount peak period is predicted, 11.1-11.30 in 2017
For the period 1.
Output: 12.1-12.31's in 2017 pulls part amount predicted value, for it is to be predicted pull part amount peak period pull part amount.
(1) it calculates " week " periodic component S: 2013-2016 11.1-12.31,2017.11.1-11.30 being taken to pull part amount
Data Y, to calculate global periodic component, and calculate 2013-2016 rejecting Zhou Xiaoying pulls part amount X, i.e. history pulls part
Amount.
(2) the 0-1 normative value R of part amount is pulled after calculating 2013-2016 rejecting Zhou Xiaoying:
(3) the 0-1 normative value R2017 of predicted time 11.1-12.31 in 2017 is predicted:
R2017=(R2014+R2015+R2016)/3
(4) estimate maximum value max2017, the minimum value min2017 of second time period 11.8-12.30 in 2017: according to
(2) formula is known in, appoints and takes two o'clock (x0, r0)、(x1, r1) can obtain
It releases:
The rejecting Zhou Xiaoying of known 2017.11.1-11.30 pulls part amount and 0-1 normative value, and permutation and combination any two points are equal
It substitutes into formula and calculates one group of maximum, minimum value, retain all values for meeting max (X) > min (X) > 0, take median as most
Greatly, the estimation of minimum value.
(5) provide the rejecting Zhou Xiaoying of 12.1-12.31 in 2017 pulls part amount:
X2017=R2017*(max2017-min2017)+min2017
(6) provide 12.1-12.31 in 2017 pulls part amount predicted value:
Above description is only the preferred embodiment of the application and the explanation to institute's application technology principle.Those skilled in the art
Member is it should be appreciated that invention scope involved in the application, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic
Scheme, while should also cover in the case where not departing from the inventive concept, it is carried out by above-mentioned technical characteristic or its equivalent feature
Any combination and the other technical solutions formed.Such as features described above has similar function with (but being not limited to) disclosed herein
Energy.
Claims (16)
1. part amount prediction technique is pulled on a kind of peak, characterized in that the following steps are included:
Configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation the historical time period, acquisition
The historical time period pulls part amount；
By the historical time period and its pulls part amount the default of part amount and standardized value is pulled according to arbitrary period variable in the time cycle
Relationship is standardized to obtain the first standardized value；
The second standardized value of time cycle to be predicted is obtained according to the first standardized value；
Second standardized value is obtained according to the preset relation that arbitrary period variable pulls part amount and standardized value in the time cycle
Time cycle to be predicted pull part amount to get it is to be predicted pull part amount peak period pull part amount.
2. part amount prediction technique is pulled on peak according to claim 1, characterized in that arbitrary period variable is pulled in the time cycle
The preset relation of part amount and standardized value are as follows:
Wherein,
xiPart amount is pulled for arbitrary period variable in the time cycle,
Max (X) is the maximum value that arbitrary period variable pulls part amount in the time cycle,
Min (X) is the minimum value that arbitrary period variable pulls part amount in the time cycle.
3. part amount prediction technique is pulled on peak according to claim 1, characterized in that configuration is to be predicted to pull part amount peak period pair
In the historical time period of foundation when the time cycle to be predicted answered and prediction, acquire the historical time period pulls part amount, packet
It includes:
Configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation the historical time period, acquisition
The historical time period pull part amount and period first time pull part amount, decomposite periodic component, and according to the period
Ingredient, must reject the historical time period of Zhou Xiaoying pulls part amount, i.e. history pulls part amount, and rejects week first time of Zhou Xiaoying
Phase pulls part amount, i.e., first pulls part amount.
4. part amount prediction technique is pulled on peak according to claim 3, characterized in that acquire pulling for the historical time period
Part amount and period first time pull part amount, decomposite periodic component, comprising:
Acquire the historical time period pulls pulling part amount and configuring preset period of time for part amount and period first time, the history
Time cycle, period first time include preset period of time；
According to the historical time period and its pull part amount, period first time and its pull part amount, preset period of time obtain the period at
Point.
5. part amount prediction technique is pulled on peak according to claim 1 to 4, characterized in that by the historical time period and its
Pull part amount according to the preset relation that arbitrary period variable in the time cycle pulls part amount and standardized value be standardized
One standardized value, comprising:
Pull that part amount, arbitrary period variable pulls the pre- of part amount and standardized value in the time cycle based on historical time period and its history
If relationship be standardized the historical time period the first standardized value.
6. part amount prediction technique is pulled on peak according to claim 3, characterized in that by second standardized value according to when
Between arbitrary period variable pulls part amount and standardized value in the period preset relation obtain the time cycle to be predicted pull part amount, comprising:
According to described first pull part amount, the second standardized value and in the time cycle arbitrary period variable pull part amount original maximum,
The preset relation of initial minimum must pull part amount maximum value, minimum value, it is described pull part amount maximum value, minimum value as it is to be predicted when
Between the period pull part amount maximum value, minimum value；
Appointed according to pull part amount maximum value, the minimum value of time cycle to be predicted in conjunction with second standardized value and in the time cycle
What the preset relation that segment variable pulls part amount and standardized value when meaning must reject the time cycle to be predicted of Zhou Xiaoying pulls part amount, i.e., in advance
It surveys first and pulls part amount；
Part amount is pulled into prediction first and pulls part amount based on what periodic component obtained the time cycle to be predicted.
7. part amount prediction technique is pulled on peak according to claim 6, characterized in that arbitrary period variable is pulled in the time cycle
Part amount original maximum, the preset relation of initial minimum are as follows:
Max (X) > min (X) > 0.
8. part amount forecasting system is pulled on a kind of peak, characterized in that include:
Configuration and acquisition unit, for configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation
The historical time period, acquire the historical time period pulls part amount；
First standardized value acquiring unit, be configured to by the historical time period and its pull part amount according to it is any in the time cycle when
Segment variable pulls part amount and the preset relation of standardized value is standardized to obtain the first standardized value；
Second standardized value predicting unit is configured to obtain the second standardization of time cycle to be predicted according to the first standardized value
Value；
Part amount predicting unit is pulled, is configured to second standardized value pulling part amount according to arbitrary period variable in the time cycle
With the preset relation of standardized value obtain the time cycle to be predicted pull part amount to get it is to be predicted pull part amount peak period pull part amount.
9. part amount forecasting system is pulled on peak according to claim 8, characterized in that arbitrary period variable is pulled in the time cycle
The preset relation of part amount and standardized value are as follows:
Wherein,
xiPart amount is pulled for arbitrary period variable in the time cycle,
Max (X) is the maximum value that arbitrary period variable pulls part amount in the time cycle,
Min (X) is the minimum value that arbitrary period variable pulls part amount in the time cycle.
10. part amount forecasting system is pulled on peak according to claim 8, characterized in that configuration and acquisition unit are also used to:
Configure it is to be predicted pull the part amount peak period corresponding time cycle to be predicted and prediction when foundation the historical time period, acquisition
The historical time period pull part amount and period first time pull part amount, decomposite periodic component, and according to the period
Ingredient, must reject the historical time period of Zhou Xiaoying pulls part amount, i.e. history pulls part amount, and rejects week first time of Zhou Xiaoying
Phase pulls part amount, i.e., first pulls part amount.
11. part amount forecasting system is pulled on peak according to claim 10, characterized in that configuration and acquisition unit are also used to:
Acquire the historical time period pulls pulling part amount and configuring preset period of time for part amount and period first time, the history
Time cycle, period first time include preset period of time；
According to the historical time period and its pull part amount, period first time and its pull part amount, preset period of time obtain the period at
Point.
12. pulling part amount forecasting system according to any peak claim 8-11, characterized in that the first standardized value obtains
Unit is also configured to:
Pull that part amount, arbitrary period variable pulls the pre- of part amount and standardized value in the time cycle based on historical time period and its history
If relationship be standardized the historical time period the first standardized value.
13. part amount forecasting system is pulled on peak according to claim 10, characterized in that pull part amount predicting unit and also configure use
In:
According to described first pull part amount, the second standardized value and in the time cycle arbitrary period variable pull part amount original maximum,
The preset relation of initial minimum must pull part amount maximum value, minimum value, it is described pull part amount maximum value, minimum value as it is to be predicted when
Between the period pull part amount maximum value, minimum value；
Appointed according to pull part amount maximum value, the minimum value of time cycle to be predicted in conjunction with second standardized value and in the time cycle
What the preset relation that segment variable pulls part amount and standardized value when meaning must reject the time cycle to be predicted of Zhou Xiaoying pulls part amount, i.e., in advance
It surveys first and pulls part amount；
Part amount is pulled into prediction first and pulls part amount based on what periodic component obtained the time cycle to be predicted.
14. part amount forecasting system is pulled on peak according to claim 12, characterized in that arbitrary period variable in the time cycle
Pull part amount original maximum, the preset relation of initial minimum are as follows:
Max (X) > min (X) > 0.
15. a kind of equipment, characterized in that the equipment includes:
One or more processors；
Memory, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processors
Execute such as method of any of claims 1-7.
16. a kind of computer readable storage medium for being stored with computer program, characterized in that when the program is executed by processor
Realize such as method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810788118.3A CN109146134A (en) | 2018-07-18 | 2018-07-18 | Part amount prediction technique, system, equipment and storage medium are pulled in a kind of peak |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810788118.3A CN109146134A (en) | 2018-07-18 | 2018-07-18 | Part amount prediction technique, system, equipment and storage medium are pulled in a kind of peak |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109146134A true CN109146134A (en) | 2019-01-04 |
Family
ID=64800988
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810788118.3A Pending CN109146134A (en) | 2018-07-18 | 2018-07-18 | Part amount prediction technique, system, equipment and storage medium are pulled in a kind of peak |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109146134A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109816442A (en) * | 2019-01-16 | 2019-05-28 | 四川驹马科技有限公司 | A kind of various dimensions freight charges prediction technique and its system based on feature tag |
| CN112183832A (en) * | 2020-09-17 | 2021-01-05 | 上海东普信息科技有限公司 | Express pickup quantity prediction method, device, equipment and storage medium |
| CN112862137A (en) * | 2019-11-27 | 2021-05-28 | 顺丰科技有限公司 | Method and device for predicting quantity, computer equipment and computer readable storage medium |
| CN112949888A (en) * | 2019-12-11 | 2021-06-11 | 顺丰科技有限公司 | Component prediction model training and component prediction method, device and computer equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140188748A1 (en) * | 2012-12-28 | 2014-07-03 | Exapaq Sas | Methods and systems for determining estimated package delivery/pick-up times |
| CN105976153A (en) * | 2016-06-20 | 2016-09-28 | 中联物流（中国）有限公司 | Vehicle departure data predication method of logistics management system |
| CN106909990A (en) * | 2017-03-01 | 2017-06-30 | 腾讯科技（深圳）有限公司 | A kind of Forecasting Methodology and device based on historical data |
| CN108288115A (en) * | 2018-03-15 | 2018-07-17 | 安徽大学 | A kind of daily short-term express delivery amount prediction technique of loglstics enterprise |
-
2018
- 2018-07-18 CN CN201810788118.3A patent/CN109146134A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140188748A1 (en) * | 2012-12-28 | 2014-07-03 | Exapaq Sas | Methods and systems for determining estimated package delivery/pick-up times |
| CN105976153A (en) * | 2016-06-20 | 2016-09-28 | 中联物流（中国）有限公司 | Vehicle departure data predication method of logistics management system |
| CN106909990A (en) * | 2017-03-01 | 2017-06-30 | 腾讯科技（深圳）有限公司 | A kind of Forecasting Methodology and device based on historical data |
| CN108288115A (en) * | 2018-03-15 | 2018-07-17 | 安徽大学 | A kind of daily short-term express delivery amount prediction technique of loglstics enterprise |
Non-Patent Citations (1)
| Title |
|---|
| 林国华: "时间序列分析法在移动通信数据分析中的研究与应用", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109816442A (en) * | 2019-01-16 | 2019-05-28 | 四川驹马科技有限公司 | A kind of various dimensions freight charges prediction technique and its system based on feature tag |
| CN112862137A (en) * | 2019-11-27 | 2021-05-28 | 顺丰科技有限公司 | Method and device for predicting quantity, computer equipment and computer readable storage medium |
| CN112949888A (en) * | 2019-12-11 | 2021-06-11 | 顺丰科技有限公司 | Component prediction model training and component prediction method, device and computer equipment |
| CN112183832A (en) * | 2020-09-17 | 2021-01-05 | 上海东普信息科技有限公司 | Express pickup quantity prediction method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109146134A (en) | Part amount prediction technique, system, equipment and storage medium are pulled in a kind of peak | |
| Khatib | Integrating project risk management and value engineering in tendering processes | |
| Anggraeni et al. | Performance comparisons between arima and arimax method in moslem kids clothes demand forecasting: Case study | |
| Dück et al. | Increasing stability of crew and aircraft schedules | |
| US8108235B2 (en) | System and method for applying predictive metric analysis for a business monitoring subsystem | |
| Alfieri et al. | A project scheduling approach to production planning with feeding precedence relations | |
| CN109284881A (en) | Order allocation method, device, computer readable storage medium and electronic equipment | |
| US20080183553A1 (en) | Method and apparatus for workforce demand forecasting | |
| Giesecke et al. | Monte Carlo algorithms for default timing problems | |
| CN107993088A (en) | A kind of Buying Cycle Forecasting Methodology and device, electronic equipment | |
| CN115689451A (en) | Method, device, terminal and medium for determining replenishment quantity of off-line retail store | |
| CN110322169A (en) | A kind of method and device of task publication | |
| Alfieri et al. | Dynamic inventory rationing: How to allocate stock according to managerial priorities. An empirical study | |
| Kolková et al. | Demand forecasting: an alternative approach based on technical indicator Pbands | |
| US8401944B2 (en) | Marketing investment optimizer with dynamic hierarchies | |
| CN111222668A (en) | Warehouse list prediction method and device, electronic equipment and readable storage medium | |
| Rausch et al. | Beyond the beaten paths of forecasting call center arrivals: on the use of dynamic harmonic regression with predictor variables | |
| CN111353797B (en) | Resource allocation method and device and electronic equipment | |
| CN102103714A (en) | Real-time processing platform for predicting service data and predicting method | |
| WO2013061324A2 (en) | A method for estimating the total cost of ownership (tcp) for a requirement | |
| CN107844496A (en) | Statistical information output intent and device | |
| CN110689177A (en) | Method and device for predicting order preparation time, electronic equipment and storage medium | |
| Halkos et al. | Evaluating alternative frequentist inferential approaches for optimal order quantities in the newsvendor model under exponential demand | |
| CN114219247A (en) | Processing method and device of equipment consumables, computer equipment and storage medium | |
| US20170132644A1 (en) | Forecasting demand for groups of items, such as mobile phones |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190104 |