WO2023155425A1 - 调货方法、装置、电子设备和计算机可读介质 - Google Patents

调货方法、装置、电子设备和计算机可读介质 Download PDF

Info

Publication number
WO2023155425A1
WO2023155425A1 PCT/CN2022/118574 CN2022118574W WO2023155425A1 WO 2023155425 A1 WO2023155425 A1 WO 2023155425A1 CN 2022118574 W CN2022118574 W CN 2022118574W WO 2023155425 A1 WO2023155425 A1 WO 2023155425A1
Authority
WO
WIPO (PCT)
Prior art keywords
item
circulation volume
quantile
item data
determining
Prior art date
Application number
PCT/CN2022/118574
Other languages
English (en)
French (fr)
Inventor
于莹
庄晓天
Original Assignee
北京京东振世信息技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京京东振世信息技术有限公司 filed Critical 北京京东振世信息技术有限公司
Publication of WO2023155425A1 publication Critical patent/WO2023155425A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
    • 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
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • 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
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0607—Regulated
    • 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

Definitions

  • the embodiments of the present disclosure relate to the field of computer technology, and specifically relate to a method, device, electronic equipment, and computer-readable media for goods transfer.
  • interval prediction is an important branch of uncertainty prediction. That is, under the specified confidence level, the narrowest interval estimate of the future circulation of goods is given.
  • the usual method is: often specify that the historical item circulation obeys a certain distribution (for example, normal distribution), and obtain items under different confidence levels by calculating the mean and standard deviation Turnover forecast interval.
  • Some embodiments of the present disclosure provide a method, device, electronic device, and computer-readable medium to solve one of the technical problems mentioned in the background section above.
  • some embodiments of the present disclosure provide a goods transfer method, including: obtaining a pre-trained model for determining the quantile of the first item’s circulation volume and a pre-trained model for determining the quantile of the second item’s circulation volume, wherein , the above-mentioned first item circulation volume quantile point determination model and the above-mentioned second item circulation volume quantile point determination model are trained based on the first item data subset in the target item data set and a preset confidence level; according to the above-mentioned pre-training
  • the quantile point determination model of the first item circulation volume and the above-mentioned pre-trained second item circulation volume quantile point determination model generate the item circulation volume prediction interval corresponding to each item data in the second item data subset in the target item data set , to obtain the item circulation forecast interval set; according to the above item circulation forecast interval set, determine the residual information corresponding to each item data in the second item data subset, and obtain the residual information set; according to the above residual information set, generate
  • some embodiments of the present disclosure provide a goods transfer device, including: an acquisition unit configured to acquire a pre-trained model for determining the quantile point of the first item circulation volume and a pre-trained second item circulation volume quantile point determination model.
  • the location determination model wherein, the above-mentioned first item circulation volume quantile point determination model and the above-mentioned second item circulation volume quantile point determination model are trained based on the first item data subset in the target item data set and a preset confidence level
  • the first generation unit is configured to generate the second item in the target item data set according to the above-mentioned pre-trained model for determining the quantile of the circulation volume of the first item and the above-mentioned pre-trained model for determining the quantile of the circulation volume of the second item
  • the item circulation volume prediction interval corresponding to each item data in the data subset is obtained to obtain an item circulation volume prediction interval set; the determination unit is configured to determine each item data in the second item data subset according to the above item circulation volume prediction interval set
  • the corresponding residual information is obtained by obtaining a residual information set; the second generating unit is configured to generate item circulation quantity conversion information according to the above residual information set.
  • the goods transfer processing unit is configured to perform goods transfer processing on the items
  • some embodiments of the present disclosure provide an electronic device, including: at least one processor; and a storage device, on which at least one program is stored.
  • the implementer implements the method described in any implementation manner in the first aspect.
  • some embodiments of the present disclosure provide a computer-readable medium on which a computer program is stored, wherein, when the program is executed by a processor, the method described in any implementation manner in the first aspect is implemented.
  • Fig. 1 is a schematic diagram of an application scenario of a goods transfer method according to some embodiments of the present disclosure
  • FIG. 2 is a flow chart of some embodiments of a method for transferring goods according to the present disclosure
  • Fig. 3 is a flow chart of other embodiments of the method for transferring goods according to the present disclosure.
  • Fig. 4 is a flow chart of some other embodiments of the goods transfer method according to the present disclosure.
  • Fig. 5 is a structural schematic diagram of some embodiments of a cargo transfer device according to the present disclosure.
  • FIG. 6 is a schematic structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure.
  • Relevant goods transfer methods for example, specifying that the circulation of historical items obey a certain distribution (for example, normal distribution), and obtaining the forecast interval of the circulation of items under different confidence levels by calculating the mean and standard deviation, etc. often exist as follows Technical problem: For many situations, the circulation of historical items often cannot obey a certain distribution. Therefore, the accuracy rate of determining the forecast interval of the item circulation volume by using the assumed distribution is low. When the accuracy of the prediction interval is low, it is very easy to cause a backlog of items in the warehouse or insufficient supply of items in the warehouse, often resulting in a large amount of waste of inventory resources, resulting in low utilization of inventory resources.
  • some embodiments of the present disclosure propose methods and devices for stock transfer, which can greatly reduce the waste of inventory resources and greatly improve the utilization rate of inventory resources. Without considering the distribution of item circulation, Efficiently and accurately generate item turnover conversion information.
  • Fig. 1 is a schematic diagram of an application scenario of a goods transfer method according to some embodiments of the present disclosure.
  • the electronic device 101 may first obtain a pre-trained model 104 for determining a quantile of the first item circulation volume and a pre-trained model 105 for determining a quantile point of the second item circulation volume.
  • the above-mentioned first item circulation volume quantile point determination model 104 and the above-mentioned second item circulation volume quantile point determination model 105 are trained based on the first item data subset 102 in the target item data set and the preset confidence level 103 .
  • the first item data subset 102 may include: data 1021 , data 1022 , and data 1023 .
  • the aforementioned confidence level 103 may be 80%.
  • the electronic device 101 can generate the second item data set in the target item data set based on the above-mentioned pre-trained model 104 for determining the quantile of the circulation volume of the first item and the above-mentioned pre-trained model 105 for determining the quantile point of the second item circulation volume.
  • the item circulation volume prediction interval corresponding to each item data in the set 106 is obtained to obtain the item circulation volume prediction interval set 107 .
  • the second item data subset 106 includes: data 1061 , data 1062 , and data 1063 .
  • the item circulation volume prediction interval set 107 includes: item circulation volume prediction interval 1071 corresponding to the above data 1061 , item circulation volume prediction interval 1072 corresponding to the above data 1062 , and item circulation volume prediction interval 1073 corresponding to the above data 1063 .
  • the forecast interval 1071 of the above-mentioned item circulation volume may be: [50, 145].
  • the forecast interval 1072 of the above-mentioned item circulation volume may be: [30, 125].
  • the forecast interval 1073 of the above-mentioned item circulation volume may be: [90, 225].
  • the electronic device 101 may determine the residual information corresponding to each item data in the second item data subset 107 according to the item circulation volume prediction interval set 107 to obtain the residual information set 108 .
  • the residual information 108 may include: residual information 1081 corresponding to item data 1071 , residual information 1082 corresponding to item data 1072 , and residual information 1083 corresponding to item data 1073 .
  • the above residual information 1081 may be: -14.
  • the above residual information 1082 may be: -12.
  • the above residual information 1083 may be: 24.
  • the electronic device 101 may generate item circulation volume transformation information 109 according to the above residual information set 108 .
  • the above item flow conversion information 109 may be: 10.
  • the items corresponding to the target item data set are transferred.
  • the above-mentioned electronic device 101 may be hardware or software.
  • the electronic device When the electronic device is hardware, it can be realized as a distributed cluster composed of multiple servers or terminal devices, or as a single server or a single terminal device.
  • an electronic device When an electronic device is embodied as software, it can be installed in the hardware devices listed above. It can be implemented, for example, as a plurality of software or software modules for providing distributed services, or as a single software or software module. No specific limitation is made here.
  • FIG. 1 the number of electronic devices in FIG. 1 is merely illustrative. There may be any number of electronic devices depending on implementation needs.
  • the transfer method includes the following steps:
  • Step 201 acquiring a pre-trained model for determining the quantile point of the first item circulation volume and a pre-trained model for determining the second item circulation volume quantile point.
  • the executive body of the above-mentioned goods transfer method can obtain the pre-trained model for determining the quantile point of the first item circulation volume and the pre-trained The trained model for determining the quantile of the second item turnover.
  • the above-mentioned first item circulation volume quantile determination model and the second item circulation volume quantile determination model may be regression models used to determine the item circulation volume of the item (for example, the item circulation volume may be the sales volume of the item).
  • the above-mentioned first item circulation volume quantile determination model and the second item circulation volume quantile determination model may be one of the following: Gradient Boosting Decision Tree (GBDT, Gradient Boosting Decision Tree) model, LightGBM model.
  • the above-mentioned first item circulation volume quantile determination model and the above-mentioned second item circulation volume quantile determination model are trained through the following steps:
  • the first step is to preprocess the above target item data set to obtain the preprocessed item data set.
  • the item features corresponding to the target item data in the target item data set may include but not limited to one of the following: basic time features, event features, time lag features, time aggregation features, and item turnover trend features.
  • the above-mentioned basic time features may include: year information, month information, information on whether it is a weekend, and season information.
  • the above-mentioned event characteristics may include: information on whether it is a statutory holiday, information on whether it is a promotional holiday.
  • the above-mentioned time-lag characteristics may include: the characteristics of the circulation volume of items after a lag of 7-11 days.
  • the above-mentioned time aggregation feature may include: the average value, maximum value, minimum value, and skewness peak value of the item circulation within the sliding window (2 days) after a lag of 7 days.
  • the above-mentioned article circulation trend characteristics include: after a lag of 7 days, the percentage change of the article turnover relative to the previous few days, and whether the article turnover is high.
  • the above execution subject may perform normalization processing on the target item data in the target item data set to obtain a preprocessed item data set.
  • the second step is to divide the above-mentioned preprocessed item data set into the above-mentioned first item data subset and the above-mentioned second item data subset.
  • the execution subject may equally divide the preprocessed item data set into the first item data subset and the second item data subset.
  • the third step is to determine the model structure of the model for determining the quantile point of the first item circulation volume and the model structure of the model for determining the second item circulation volume quantile point according to the above confidence level.
  • the aforementioned executive body may first determine the confidence level. Then, the above-mentioned executive body can determine each quantile point through the target formula set. Finally, the above-mentioned executive body can determine the loss functions of the first item circulation volume quantile point determination model and the second item circulation volume quantile point determination model through each quantile point.
  • the confidence level is 80%.
  • the above-mentioned executive body may determine that the loss function of the quantile determination model of the first item circulation volume is related to the 10% quantile.
  • the aforementioned executive body may determine that the loss function of the model for determining the quantile of the second item circulation volume is related to the 90% quantile.
  • Step 4 According to the first item data subset, the model structure of the first item circulation quantile determination model and the model structure of the second item circulation quantile determination model, train the first item circulation Quantile point determination model and the quantile point determination model of the second article circulation volume, the above-mentioned trained first article circulation volume quantile point determination model and the above-mentioned trained second article circulation volume quantile point determination model are obtained.
  • the model structure of the above-mentioned first item circulation volume quantile determination model and the model structure of the above-mentioned second item circulation volume quantile determination model can be trained by deep learning
  • the method is to train the determination model of the quantile point of the circulation volume of the first article and the determination model of the quantile point of the circulation volume of the second article, and obtain the determination model of the quantile point of the circulation volume of the first article after the training and the second article after the training.
  • Quantile determination model for turnover is to train the determination model of the quantile point of the circulation volume of the first article and the determination model of the quantile point of the circulation volume of the second article, and obtain the determination model of the quantile point of the circulation volume of the first article after the training and the second article after the training.
  • Step 202 Generate data for each item in the second item data subset in the target item data set based on the above-mentioned pre-trained model for determining the quantile point of the first item circulation volume and the above-mentioned pre-trained model for determining the quantile point of the second item circulation volume The corresponding item circulation forecast intervals are used to obtain the item circulation forecast interval set.
  • the executive body may generate the target item in various ways according to the above-mentioned pre-trained model for determining the quantile of the circulation volume of the first item and the above-mentioned pre-trained model for determining the quantile point of the circulation volume of the second item.
  • the item circulation volume prediction interval corresponding to each item data in the second item data subset in the data set is obtained to obtain an item circulation volume prediction interval set.
  • the item circulation volume prediction interval corresponding to the item data may be a range of times of possible item value operations performed on the item corresponding to the item data.
  • the above-mentioned target item data is generated according to the above-mentioned pre-trained model for determining the quantile point of the first item circulation volume and the above-mentioned pre-trained model for determining the quantile point of the second item circulation volume
  • Concentrating the item turnover forecast interval corresponding to each item data in the second item data subset may include the following steps:
  • the above-mentioned item data is respectively input into the above-mentioned pre-trained model for determining the quantile point of the first item circulation volume and the above-mentioned pre-trained model for determining the quantile point of the second item circulation volume to obtain the first value and the second value.
  • both the above-mentioned first value and the above-mentioned second value can represent the item circulation volume of the item corresponding to the item data.
  • an item circulation volume prediction interval corresponding to the above-mentioned item data is generated.
  • the above execution subject may first determine the size between the first value and the second value. Then, the above-mentioned executive body may use the smaller value of the first value and the second value as the minimum item circulation amount in the item circulation amount prediction interval, and use the larger value among the first value and the second value as the highest item amount in the item flow amount prediction interval turnover.
  • the generated item circulation forecast interval may be: [23, 45].
  • Step 203 Determine the residual information corresponding to each item data in the second item data subset according to the item circulation volume prediction interval set, and obtain a residual information set.
  • the execution subject may determine the residual information corresponding to each item data in the second item data subset according to the item circulation volume prediction interval set, and obtain a residual information set.
  • the residual information can represent the gap between the actual sales value of the item data and the predicted value of the item circulation.
  • the first sub-step is to determine the real item circulation value of the above-mentioned second item data.
  • the second sub-step is to determine the item circulation volume prediction interval of the above-mentioned second item data.
  • the third sub-step is to determine the first value and the second value corresponding to the forecast interval of the above-mentioned item circulation volume.
  • the fourth sub-step is to make a difference between the first value corresponding to the item circulation forecast interval and the second value corresponding to the item circulation forecast interval and the real item circulation value corresponding to the item data to obtain the third The differential value and the fourth differential value.
  • the average between the above-mentioned third differential value and the fourth differential value is used as the residual information.
  • step 204 according to the above-mentioned residual information set, the conversion information of the item circulation amount is generated.
  • the above execution subject may determine the conversion information of the item circulation amount according to the above residual information set.
  • the above item circulation volume transformation information may represent the transformation information of the item data corresponding to the item circulation volume prediction interval.
  • the above-mentioned executive body may use the average value corresponding to the residual information set as the item circulation quantity conversion information.
  • the item prediction interval corresponding to the item data is [30,70].
  • the conversion information of item circulation is 10, which indicates that the interval of item circulation can be between [20,80].
  • Step 205 according to the conversion information of the above-mentioned item circulation, perform goods transfer processing on the items corresponding to the above-mentioned target item data set.
  • the above-mentioned execution subject may perform goods transfer processing on the items corresponding to the above-mentioned target item data set according to the above-mentioned item turnover conversion information.
  • the above-mentioned executive body may use the allocating device to perform replenishment processing or purchase processing on the items corresponding to the above-mentioned target item data set.
  • the above execution subject may first determine the number of items in the item set stored in the target warehouse. Then, the execution subject dynamically performs replenishment processing or purchase processing on the item set in the target warehouse according to the item turnover conversion information.
  • the execution subject may determine the item circulation volume range information corresponding to each item data in the second item data subset.
  • the above-mentioned execution subject may first determine the interval information of the item circulation volume of the item data. Then, the difference between the first value corresponding to the item circulation volume interval information and the item circulation volume conversion information, and the sum of the second value and the item circulation volume conversion information to obtain the item circulation volume interval information corresponding to the item data.
  • the goods transfer method in some embodiments of the present disclosure can efficiently and accurately generate item turnover conversion information without considering the distribution of item turnover.
  • the circulation of historical items often cannot obey a certain distribution.
  • the accuracy of using the assumed distribution to determine the forecast range of item circulation is low.
  • the forecasted quantity is greater than the actual demand, it is easy to cause a backlog of goods in the warehouse.
  • the forecasted quantity is smaller than the actual demand, it is often Cause a lot of waste of inventory resources, the above two situations will result in low utilization of inventory resources.
  • the goods transfer method in some embodiments of the present disclosure may first obtain a pre-trained model for determining the quantile of the circulation volume of the first item and a pre-trained model for determining the quantile of the circulation volume of the second item, wherein the above-mentioned first item
  • the quantile point determination model of circulation volume and the above-mentioned second quantile point determination model of item circulation volume are trained based on the first item data subset in the target item data set and a preset confidence level.
  • the determination model of the first item circulation volume quantile point and the second item circulation volume quantile point determination model are obtained for subsequent more efficient and accurate generation of item circulation volume prediction intervals.
  • the pre-trained model for determining the quantile point of the first item circulation volume and the above-mentioned pre-trained model for determining the quantile point of the second item circulation volume generate the data corresponding to each item in the second item data subset in the target item data set
  • the prediction interval of the item circulation volume is obtained, and the set of item circulation volume prediction intervals is obtained.
  • a more accurate item turnover forecast interval can be generated.
  • the interval length of the item circulation forecast interval generated by the existing method is often a fixed length, which leads to the inability of the item circulation forecast interval to better adapt to heteroscedastic data, and the length of the item circulation forecast interval cannot be better adaptively adjusted .
  • the item circulation volume prediction interval generated based on the first item circulation volume quantile point determination model and the above-mentioned second item circulation volume quantile point determination model can adaptively adjust the length of the interval, so that the accuracy of the item circulation volume prediction interval higher.
  • the residual information corresponding to each item data in the second item data subset is determined to obtain a residual information set.
  • the residual information is determined according to the forecast interval of the circulation volume of the goods, so as to be used for determining the conversion information of the circulation volume of the subsequent goods. Then, according to the above residual information set, more accurate item circulation conversion information can be generated. Finally, according to the conversion information of the above-mentioned item circulation volume, the items corresponding to the above-mentioned target item data set are transferred.
  • the goods transfer process for the items corresponding to the target item data set can greatly reduce the waste of inventory resources and greatly improve the utilization rate of inventory resources.
  • the transfer method includes the following steps:
  • Step 301 acquiring a pre-trained model for determining the quantile point of the first item circulation volume and a pre-trained model for determining the second item circulation volume quantile point.
  • Step 302 Generate data for each item in the second item data subset in the target item data set based on the above-mentioned pre-trained model for determining the quantile point of the first item circulation volume and the above-mentioned pre-trained model for determining the quantile point of the second item circulation volume The corresponding item circulation forecast intervals are used to obtain the item circulation forecast interval set.
  • Step 303 for each item data in the second item data subset, perform the following residual information determination step.
  • Step 3031 determine the item circulation volume prediction interval corresponding to the above item data.
  • the execution subject may determine the item circulation volume prediction interval corresponding to the item data from the target database.
  • the above-mentioned target database may be a database storing an association relationship between item data and item circulation volume prediction intervals.
  • Step 3032 determining the first value and the second value corresponding to the forecast interval of the above-mentioned item circulation volume.
  • the execution subject may determine the first value and the second value corresponding to the item circulation volume prediction interval.
  • the above execution subject may determine the maximum value and the minimum value in the item circulation volume prediction interval as the first value and the second value.
  • the forecast interval of the above item circulation volume may be: [30,70]. Then the first value is 70. The second value is 30.
  • Step 3033 making a difference between the first value corresponding to the item circulation forecast interval and the second value corresponding to the item circulation forecast interval and the real item circulation value corresponding to the item data to obtain the first difference value and the second difference value.
  • the execution subject may operate the first value corresponding to the item circulation forecast interval and the second value corresponding to the item circulation forecast interval with the real item circulation value corresponding to the item data respectively. difference, get the first difference value and the second difference value.
  • the first differential value is 10.
  • the second differential value is -30.
  • Step 3034 Determine the maximum value between the first difference value and the second difference value as the residual information.
  • the execution subject may determine the maximum value between the first difference value and the second difference value as the residual information.
  • the above execution subject may determine 10 as the residual information.
  • step 304 according to the above-mentioned residual information set, the conversion information of the item circulation amount is generated.
  • step 305 according to the conversion information of the circulation volume of the above-mentioned items, the item corresponding to the above-mentioned target item data set is transferred.
  • steps 301-302 and 304-305 for the specific implementation of steps 301-302 and 304-305 and the technical effects brought about by them, reference may be made to steps 201-202 and 204-205 in the embodiment corresponding to FIG. 2 , which will not be repeated here. .
  • the process 300 of the goods transfer method in some embodiments corresponding to FIG. 3 highlights the specific steps of determining the residual information corresponding to the item data.
  • the prediction interval may not be well adapted to heteroscedastic data
  • the length of the prediction interval cannot be well adaptively adjusted.
  • ordinary machine learning predicts the quantile points based on the training set to obtain the upper and lower bounds of the interval, but does not make further error adjustments for the upper and lower bounds, which leads to the prediction interval being too large , thus losing the significance of interval forecasting, and there may also be certain errors.
  • the solutions described in these embodiments can avoid the problems of not being able to adapt well to heteroscedastic data, the length of the prediction interval cannot be well adaptively adjusted, and no further error adjustment is made to the upper and lower bounds, which greatly improves the generation of residual information. Accuracy, so as to make the generation of subsequent item circulation conversion information more accurate.
  • the transfer method includes the following steps:
  • Step 401 acquiring a pre-trained model for determining the quantile point of the first item circulation volume and a pre-trained model for determining the second item circulation volume quantile point.
  • Step 402 Generate data for each item in the second item data subset in the target item data set based on the above-mentioned pre-trained model for determining the quantile point of the first item circulation volume and the above-mentioned pre-trained model for determining the quantile point of the second item circulation volume The corresponding item circulation forecast intervals are used to obtain the item circulation forecast interval set.
  • Step 403 Determine residual information corresponding to each item data in the second item data subset according to the item circulation volume prediction interval set, and obtain a residual information set.
  • Step 404 filter out the residual information satisfying the predetermined condition from the above-mentioned residual information set, and use it as the above-mentioned commodity circulation conversion information.
  • the execution subject may filter residual information satisfying a predetermined condition from the residual information set according to the aforementioned confidence level, as the item circulation conversion information.
  • the above execution subject may sort the residual information set to obtain a residual information sequence. Then, the execution subject can find the target residual information whose numerical value is greater than the confidence level residual information from the residual information sequence. For example, a confidence level of 80%.
  • the residual information sequence is: [2,4,5,6,7]. Then the target residual information is 6.
  • the above-mentioned executive body determines the target residual information as the item circulation quantity transformation information.
  • Step 405 according to the conversion information of the above-mentioned item circulation volume, perform goods transfer processing on the items corresponding to the above-mentioned target item data set.
  • steps 401-403 and 405 for the specific implementation of steps 401-403 and 405 and the technical effects brought about by them, reference may be made to steps 201-203 and 205 in the embodiment corresponding to FIG. 2 , which will not be repeated here.
  • the process 400 of the goods transfer method in some embodiments corresponding to FIG. 4 highlights the specific steps of determining the conversion information of the item circulation.
  • the protocols described in these examples pass the predetermined condition.
  • the transformation information of the item circulation can be determined more accurately.
  • the present disclosure provides some embodiments of a cargo transfer device. These device embodiments correspond to those method embodiments shown in FIG. 2 , and the device can specifically Used in various electronic equipment.
  • a goods transfer device 500 includes: an acquisition unit 501 , a first generation unit 502 , a determination unit 503 , a second generation unit 504 and a goods transfer processing unit 505 .
  • the acquiring unit 501 is configured to acquire a pre-trained model for determining the quantile point of the first item circulation volume and a pre-trained model for determining the second item circulation volume quantile point, wherein the above-mentioned first item circulation volume quantile point determination model
  • the model and the above-mentioned second item turnover quantile point determination model are trained based on the first item data subset in the target item data set and a preset confidence level
  • the first generation unit 502 is configured to be based on the above-mentioned pre-trained first A quantile point determination model of the item circulation volume and the above-mentioned pre-trained second item circulation volume quantile point determination model generate the item circulation volume prediction interval corresponding to each item data in the second item data subset in the target item data set
  • the foregoing apparatus 500 further includes: a third determining unit (not shown in the figure).
  • the above-mentioned third determination unit may be configured to: according to the above-mentioned item circulation volume transformation information, determine the item circulation volume interval information corresponding to each item data in the above-mentioned second item data subset.
  • the first generation unit 502 in the above-mentioned apparatus 500 may be configured to: respectively input the above-mentioned item data into the above-mentioned pre-trained first item flow quantile determination model and The above-mentioned pre-trained model for determining quantiles of the second item circulation volume obtains the first value and the second value; according to the above-mentioned first value and the above-mentioned second value, the item circulation volume prediction interval corresponding to the above-mentioned item data is generated.
  • the determination unit 503 in the above-mentioned apparatus 500 may be configured to: for each item data in the second item data subset, perform the following residual information determination step: determine the above-mentioned item The item circulation forecast interval corresponding to the data; determining the first value and the second value corresponding to the item circulation forecast interval; The difference between the second value and the real item circulation value corresponding to the above item data is respectively obtained to obtain the first difference value and the second difference value; the maximum difference between the above first difference value and the above second difference value The value is determined as the above residual information.
  • the second generating unit 504 in the above-mentioned apparatus 500 may be configured to: filter out residual information that meets a predetermined condition from the above-mentioned residual information set according to the above-mentioned confidence level, as The above-mentioned item circulation conversion information.
  • the model for determining the quantile point of the first item circulation volume and the above-mentioned second item circulation volume quantile point determination model are trained through the following steps: performing the following steps on the target item data set Preprocessing, obtaining the preprocessed item data set; dividing the above preprocessed item data set into the above-mentioned first item data subset and the above-mentioned second item data subset; determining the circulation of the above-mentioned first item according to the above-mentioned confidence level
  • the model structure of the model for determining the quantile point of the second item circulation volume train the above-mentioned first item circulation volume quantile point determination model and the above-mentioned second item circulation volume
  • the units recorded in the device 500 correspond to the steps in the method described with reference to FIG. 2 . Therefore, the operations, features and beneficial effects described above for the method are also applicable to the device 500 and the units contained therein, and will not be repeated here.
  • FIG. 6 it shows a schematic structural diagram of an electronic device (such as the electronic device 101 in FIG. 1 ) 600 suitable for implementing some embodiments of the present disclosure.
  • the electronic device shown in FIG. 6 is only an example, and should not limit the functions and scope of use of the embodiments of the present disclosure.
  • an electronic device 600 may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 601, which may be randomly accessed according to a program stored in a read-only memory (ROM) 602 or loaded from a storage device 608.
  • a processing device such as a central processing unit, a graphics processing unit, etc.
  • RAM memory
  • various programs and data necessary for the operation of the electronic device 600 are also stored.
  • the processing device 601, ROM 602, and RAM 603 are connected to each other through a bus 604.
  • An input/output (I/O) interface 605 is also connected to the bus 604 .
  • the following devices can be connected to the I/O interface 605: input devices 606 including, for example, a touch screen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; including, for example, a liquid crystal display (LCD), speaker, vibration an output device 607 such as a computer; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609.
  • the communication means 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. While FIG. 6 shows electronic device 600 having various means, it should be understood that implementing or having all of the means shown is not a requirement. More or fewer means may alternatively be implemented or provided. Each block shown in FIG. 6 may represent one device, or may represent multiple devices as required.
  • the processes described above with reference to the flowcharts may be implemented as computer software programs.
  • some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, where the computer program includes program codes for executing the methods shown in the flowcharts.
  • the computer program may be downloaded and installed from a network via communication means 609, or from storage means 608, or from ROM 602.
  • the processing device 601 When the computer program is executed by the processing device 601, the above functions defined in the methods of some embodiments of the present disclosure are performed.
  • the above-mentioned computer-readable medium in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two.
  • a computer readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination thereof. More specific examples of computer readable storage media may include, but are not limited to: electrical connections having at least one lead, portable computer diskettes, hard disks, random access memory (RAM), read only memory (ROM), erasable programmable Read memory (EPROM or flash memory), fiber optics, portable compact disk read only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above.
  • a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.
  • a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code therein. Such propagated data signals may take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • a computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can transmit, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device .
  • Program code embodied on a computer readable medium may be transmitted by any appropriate medium, including but not limited to wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.
  • the client and the server can communicate using any currently known or future network protocols such as HTTP (HyperText Transfer Protocol, Hypertext Transfer Protocol), and can communicate with digital data in any form or medium
  • HTTP HyperText Transfer Protocol
  • the communication eg, communication network
  • Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), internetworks (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future developed network of.
  • the above-mentioned computer-readable medium may be included in the above-mentioned electronic device, or may exist independently without being incorporated into the electronic device.
  • the above-mentioned computer-readable medium carries one or more programs, and when the above-mentioned one or more programs are executed by the electronic device, the electronic device: obtains the pre-trained model for determining the quantile point of the first article turnover and the pre-trained The model for determining the quantile point of the second item circulation volume, wherein, the above-mentioned first item circulation volume quantile point determination model and the above-mentioned second item circulation volume quantile point determination model are based on the first item data subset in the target item data set and Trained with a pre-set confidence level; according to the above-mentioned pre-trained model for determining the quantile point of the first item's circulation volume and the above-mentioned pre-trained model for determining the quantile point of the second item's circulation volume, generate the second item data subclass in
  • Computer program code for carrying out operations of some embodiments of the present disclosure may be written in one or more programming languages, or combinations thereof, including object-oriented programming languages—such as Java, Smalltalk, C++, Also included are conventional procedural programming languages - such as the "C" language or similar programming languages.
  • the program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server.
  • the remote computer may be connected to the user computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (for example, using an Internet service provider to connected via the Internet).
  • LAN local area network
  • WAN wide area network
  • Internet service provider for example, using an Internet service provider to connected via the Internet.
  • each block in the flowchart or block diagram may represent a module, program segment, or part of code that contains at least one programmable logic function for implementing the specified logical function.
  • Execute instructions may also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or they may sometimes be executed in the reverse order, depending upon the functionality involved.
  • each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations can be implemented by a dedicated hardware-based system that performs the specified functions or operations , or may be implemented by a combination of dedicated hardware and computer instructions.
  • the units described in some embodiments of the present disclosure may be realized by software or by hardware.
  • the described units may also be set in a processor.
  • a processor includes an acquisition unit, a first generation unit, a determination unit, a second generation unit, and a transfer processing unit.
  • the names of these units do not constitute a limitation to the unit itself under certain circumstances.
  • the acquisition unit can also be described as "obtaining the pre-trained model for determining the quantile of the first item circulation volume and the pre-trained first The unit of the model for determining the quantile of the second commodity circulation volume".
  • exemplary types of hardware logic components include: Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), System on Chips (SOCs), Complex Programmable Logical device (CPLD) and so on.
  • FPGAs Field Programmable Gate Arrays
  • ASICs Application Specific Integrated Circuits
  • ASSPs Application Specific Standard Products
  • SOCs System on Chips
  • CPLD Complex Programmable Logical device

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • General Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • Marketing (AREA)
  • General Physics & Mathematics (AREA)
  • Development Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Human Resources & Organizations (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

调货方法、装置、电子设备和计算机可读介质，包括：获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型；生成该目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集；确定该第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集；生成物品流转量变换信息；根据该物品流转量变换信息，对该目标物品数据集对应的物品进行调货处理。

Description

调货方法、装置、电子设备和计算机可读介质
相关申请的交叉引用
本申请要求于申请日为2022年02月17日提交的，申请号为202210148324.4、发明名称为“调货方法、装置、电子设备和计算机可读介质”的中国专利申请的优先权，其全部内容作为整体并入本申请中。
技术领域
本公开的实施例涉及计算机技术领域，具体涉及调货方法、装置、电子设备和计算机可读介质。
背景技术
目前，现实中的物品流转量预测场景中常常存在较多的不可控因素，给出一个确切的预测值几乎是不可能的。基于风险预防的需要，不确定性的预测在库存处理等规划活动中有着重要的应用。其中，区间预测是不确定性预测的一个重要分支。即，在指定的置信水平下，给出未来物品流转量的最窄区间估计。对于物品流转量最窄区间的生成，通常采用的方式为：常常指定历史物品流转量服从某一确定的分布(例如，正态分布)，通过计算均值和标准差来得到不同置信度下的物品流转量预测区间。
发明内容
本公开的内容部分用于以简要的形式介绍构思，这些构思将在后面的具体实施方式部分被详细描述。本公开的内容部分并不旨在标识要求保护的技术方案的关键特征或必要特征，也不旨在用于限制所要求的保护的技术方案的范围。
本公开的一些实施例提出了调货方法、装置、电子设备和计算机 可读介质，来解决以上背景技术部分提到的技术问题中的一项。
第一方面，本公开的一些实施例提供了一种调货方法，包括：获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型，其中，上述第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型是基于目标物品数据集中的第一物品数据子集和预先设置的置信水平训练的；根据上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，生成上述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集；根据上述物品流转量预测区间集，确定上述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集；根据上述残差信息集，生成物品流转量变换信息。根据上述物品流转量变换信息，对上述目标物品数据集对应的物品进行调货处理。
第二方面，本公开的一些实施例提供了一种调货装置，包括：获取单元，被配置成获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型，其中，上述第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型是基于目标物品数据集中的第一物品数据子集和预先设置的置信水平训练的；第一生成单元，被配置成根据上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，生成上述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集；确定单元，被配置成根据上述物品流转量预测区间集，确定上述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集；第二生成单元，被配置成根据上述残差信息集，生成物品流转量变换信息。调货处理单元，被配置成根据上述物品流转量变换信息，对上述目标物品数据集对应的物品进行调货处理。
第三方面，本公开的一些实施例提供了一种电子设备，包括：至少一个处理器；存储装置，其上存储有至少一个程序，当至少一个程序被至少一个处理器执行，使得至少一个处理器实现如第一方面中任 一实现方式描述的方法。
第四方面，本公开的一些实施例提供了一种计算机可读介质，其上存储有计算机程序，其中，程序被处理器执行时实现如第一方面中任一实现方式描述的方法。
附图说明
结合附图并参考以下具体实施方式，本公开各实施例的上述和其他特征、优点及方面将变得更加明显。贯穿附图中，相同或相似的附图标记表示相同或相似的元素。应当理解附图是示意性的，元件和元素不一定按照比例绘制。
图1是根据本公开的一些实施例的调货方法的一个应用场景的示意图；
图2是根据本公开的调货方法的一些实施例的流程图；
图3是根据本公开的调货方法的另一些实施例的流程图；
图4是根据本公开的调货方法的又一些实施例的流程图；
图5是根据本公开的调货装置的一些实施例的结构示意图；
图6是适于用来实现本公开的一些实施例的电子设备的结构示意图。
具体实施方式
下面将参照附图更详细地描述本公开的实施例。虽然附图中显示了本公开的某些实施例，然而应当理解的是，本公开可以通过各种形式来实现，而且不应该被解释为限于这里阐述的实施例。相反，提供这些实施例是为了更加透彻和完整地理解本公开。应当理解的是，本公开的附图及实施例仅用于示例性作用，并非用于限制本公开的保护范围。
另外还需要说明的是，为了便于描述，附图中仅示出了与有关发明相关的部分。在不冲突的情况下，本公开中的实施例及实施例中的特征可以相互组合。
需要注意，本公开中提及的“第一”、“第二”等概念仅用于对不 同的装置、模块或单元进行区分，并非用于限定这些装置、模块或单元所执行的功能的顺序或者相互依存关系。
需要注意，本公开中提及的“一个”、“多个”的修饰是示意性而非限制性的，本领域技术人员应当理解，除非在上下文另有明确指出，否则应该理解为“至少一个”。
本公开实施方式中的多个装置之间所交互的消息或者信息的名称仅用于说明性的目的，而并不是用于对这些消息或信息的范围进行限制。
相关的调货方法，例如，指定历史物品流转量服从某一确定的分布(例如，正态分布)，通过计算均值和标准差来得到不同置信度下的物品流转量预测区间等经常会存在如下技术问题：针对很多情况，历史物品流转量往往不能够服从某一确定的分布。从而，导致通过使用假定的分布来确定物品流转量预测区间的准确率较低。当预测区间的准确率较低时，极易造成仓库内的物品积压或仓库内的物品供货不足，往往会造成大量的库存资源的浪费，造成库存资源利用率低下。
为了解决以上所阐述的问题，本公开的一些实施例提出了调货方法及装置，可以极大较少库存资源的浪费，大大提高库存资源利用率，在不考虑物品流转量分布的情况下，高效、精准地生成物品流转量变换信息。
下面将参考附图并结合实施例来详细说明本公开。
图1是根据本公开一些实施例的调货方法的一个应用场景的示意图。
在图1的应用场景中，电子设备101可以首先获取预先训练的第一物品流转量分位点确定模型104和预先训练的第二物品流转量分位点确定模型105。其中，上述第一物品流转量分位点确定模型104和上述第二物品流转量分位点确定模型105是基于目标物品数据集中的第一物品数据子集102和预先设置的置信水平103训练的。在本应用场景中，上述第一物品数据子集102可以包括：数据1021、数据1022、数据1023。上述置信水平103可以是80％。然后，电子设备101可以根据上述预先训练的第一物品流转量分位点确定模型104和上述预先 训练的第二物品流转量分位点确定模型105，生成上述目标物品数据集中第二物品数据子集106中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集107。在本应用场景中，上述第二物品数据子集106包括：数据1061、数据1062、数据1063。物品流转量预测区间集107包括：与上述数据1061对应的物品流转量预测区间1071、与上述数据1062对应的物品流转量预测区间1072、与上述数据1063对应的物品流转量预测区间1073。上述物品流转量预测区间1071可以是：[50,145]。上述物品流转量预测区间1072可以是：[30,125]。上述物品流转量预测区间1073可以是：[90,225]。进而，电子设备101可以根据上述物品流转量预测区间集107，确定上述第二物品数据子集107中每个物品数据对应的残差信息，得到残差信息集108。在本应用场景中，上述残差信息108可以包括：与物品数据1071对应的残差信息1081、与物品数据1072对应的残差信息1082、与物品数据1073对应的残差信息1083。上述残差信息1081可以是：-14。上述残差信息1082可以是：-12。上述残差信息1083可以是：24。接着，电子设备101可以根据上述残差信息集108，生成物品流转量变换信息109。在本应用场景中，上述物品流转量变换信息109可以是：10。最后，根据上述物品流转量变换信息109，对上述目标物品数据集对应的物品进行调货处理。
需要说明的是，上述电子设备101可以是硬件，也可以是软件。当电子设备为硬件时，可以实现成多个服务器或终端设备组成的分布式集群，也可以实现成单个服务器或单个终端设备。当电子设备体现为软件时，可以安装在上述所列举的硬件设备中。其可以实现成例如用来提供分布式服务的多个软件或软件模块，也可以实现成单个软件或软件模块。在此不做具体限定。
应该理解，图1中的电子设备的数目仅仅是示意性的。根据实现需要，可以具有任意数目的电子设备。
继续参考图2，示出了根据本公开的调货方法的一些实施例的流程200。该调货方法，包括以下步骤：
步骤201，获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型。
在一些实施例中，上述调货方法的执行主体(例如图1所示的电子设备101)可以通过有线连接方式或者无线连接方式来获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型。其中，上述第一物品流转量分位点确定模型和第二物品流转量分位点确定模型可以是用于确定物品的物品流转量(例如，物品流转量可以是物品销量)的回归模型。作为示例，上述第一物品流转量分位点确定模型和第二物品流转量分位点确定模型可以是以下之一：梯度提升决策树(GBDT,Gradient Boosting Decision Tree)模型、LightGBM模型。
在一些实施例的一些可选的实现方式中，上述第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型是通过以下步骤训练的：
第一步、对上述目标物品数据集进行预处理，得到预处理后的物品数据集。
例如，目标物品数据集中目标物品数据对应的物品特征可以包括但不限于以下之一：基础时间特征、事件特征、时间滞后特征、时间聚合特征、物品流转量趋势特征。上述基础时间特征可以包括：年份信息，月份信息，是否为周末的信息，季节信息。上述事件特征可以包括：是否为法定假期的信息、是否为促销节日的信息。上述时间滞后特征可以包括：滞后7-11天后的物品流转量特征。上述时间聚合特征可以包括：滞后7天后滑动窗口(2天)内物品流转量的均值、最大值、最小值、偏度峰值。上述物品流转量趋势特征包括：滞后7天后，相对前几天物品流转量的百分比变化、是否为高物品流转量。
作为示例，上述执行主体可以对目标物品数据集中的目标物品数据进行归一化处理，得到预处理后的物品数据集。
第二步、将上述预处理后的物品数据集划分为上述第一物品数据子集和上述第二物品数据子集。
作为示例，上述执行主体可以将上述预处理后的物品数据集平均 划分为上述第一物品数据子集和上述第二物品数据子集。
第三步、根据上述置信水平，确定上述第一物品流转量分位点确定模型的模型结构和上述第二物品流转量分位点确定模型的模型结构。
作为示例，上述执行主体可以首先确定置信水平。然后，上述执行主体可以通过目标公式集来确定各个分位点。最后，上述执行主体可以通过各个分位点，来确定第一物品流转量分位点确定模型和第二物品流转量分位点确定模型的损失函数。其中，目标公式集可以包括：C 1＝(1-a)/2、C 2＝(1+a)/2。其中，C 1可以是下分位点。C 2可以是上分位点。
作为示例，置信水平为80％。上述执行主体可以确定第一物品流转量分位点确定模型的损失函数是与10％分位数相关的。上述执行主体可以确定第二物品流转量分位点确定模型的损失函数是与90％分位数相关的。
第四步、根据上述第一物品数据子集、上述第一物品流转量分位点确定模型的模型结构和上述第二物品流转量分位点确定模型的模型结构，训练上述第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型，得到上述训练后的第一物品流转量分位点确定模型和上述训练后的第二物品流转量分位点确定模型。
作为示例，根据上述第一物品数据子集、上述第一物品流转量分位点确定模型的模型结构和上述第二物品流转量分位点确定模型的模型结构，上述执行主体可以通过深度学习训练方法，训练上述第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型，得到上述训练后的第一物品流转量分位点确定模型和上述训练后的第二物品流转量分位点确定模型。
步骤202，根据上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，生成上述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集。
在一些实施例中，上述执行主体可以根据上述预先训练的第一物 品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，通过各种方式来生成上述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集。其中，物品数据对应的物品流转量预测区间可以是可能对物品数据对应物品执行物品价值操作的次数范围。
在一些实施例的一些可选的实现方式中，上述根据上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，生成上述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，可以包括以下步骤：
第一步，将上述物品数据分别输入至上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，得到第一数值和第二数值。其中，上述第一数值和上述第二数值都可以表征物品数据对应物品的物品流转量。
第二步，根据上述第一数值和上述第二数值，生成上述物品数据对应的物品流转量预测区间。
作为示例，上述执行主体可以首先确定第一数值和第二数值之间的大小。然后，上述执行主体可以将第一数值和第二数值中小的数值作为物品流转量预测区间的最低物品流转量，将第一数值和第二数值中大的数值作为物品流转量预测区间的最高物品流转量。
作为示例，响应于确定第一数值为23，第二数值为45，生成的物品流转量预测区间可以是：[23,45]。
步骤203，根据上述物品流转量预测区间集，确定上述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集。
在一些实施例中，上述执行主体可以根据上述物品流转量预测区间集，确定上述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集。其中，残差信息可以表征物品数据真实销售数值与物品流转量预测数值之间的差距。
作为示例，针对第二物品数据子集中每个第二物品数据，执行以下残差信息确定步骤：
第一子步骤，确定上述第二物品数据的真实物品流转量数值。
第二子步骤，确定上述第二物品数据的物品流转量预测区间。
第三子步骤，确定上述物品流转量预测区间对应的第一数值和第二数值。
第四子步骤，将上述物品流转量预测区间相对应的第一数值和上述物品流转量预测区间相对应的第二数值分别与上述物品数据对应的真实物品流转量数值进行作差，得到第三作差数值和第四作差数值。
第五子步骤，将上述第三作差数值与第四作差数值之间的平均作为残差信息。
步骤204，根据上述残差信息集，生成物品流转量变换信息。
在一些实施例中，上述执行主体可以根据上述残差信息集，确定物品流转量变换信息。上述物品流转量变换信息可以表征物品数据对应物品流转量预测区间的变换信息。
作为示例，上述执行主体可以将残差信息集对应的平均值作为物品流转量变换信息。
作为又一个示例，物品数据对应的物品预测区间为[30,70]。物品流转量变换信息为10，则表征物品流转量区间可以在[20,80]之间。
步骤205，根据上述物品流转量变换信息，对上述目标物品数据集对应的物品进行调货处理。
在一些实施例中，上述执行主体可以根据上述物品流转量变换信息，对上述目标物品数据集对应的物品进行调货处理。
作为示例，根据上述物品流转量变换信息，上述执行主体可以使用调拨装置，对上述目标物品数据集对应的物品进行补货处理或进货处理。
作为示例，上述执行主体可以首先确定目标仓库中所存储的物品集的物品数目。然后，上述执行主体根据物品流转量变换信息动态对目标仓库中物品集进行补货处理或进货处理。
在一些实施例的一些可选的实现方式中，根据上述物品流转量变换信息，上述执行主体可以确定上述第二物品数据子集中各个物品数据对应的物品流转量区间信息。
作为示例，上述执行主体可以首先确定物品数据的物品流转量区间信息。然后，通过物品流转量区间信息对应的第一数值与物品流转量变换信息作差，以及第二数值与物品流转量变换信息作和，得到物品数据对应的物品流转量区间信息。
本公开的上述各个实施例中具有如下有益效果：本公开的一些实施例的调货方法可以在不考虑物品流转量分布的情况下，高效、精准地生成物品流转量变换信息。具体来说，针对很多情况，历史物品流转量往往不能够服从某一确定的分布。从而，导致通过使用假定的分布来确定物品流转量预测区间的准确率较低，当预测量大于实际需求量时，极易造成仓库内的货物积压，当预测量小于实际需求量时，往往会造成大量的库存资源的浪费，上述两种情况都会造成库存资源利用率低下。基于此，本公开的一些实施例的调货方法可以首先获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型，其中，上述第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型是基于目标物品数据集中的第一物品数据子集和预先设置的置信水平训练的。在这里，通过获取第一物品流转量分位点确定模型和第二物品流转量分位点确定模型以用于后续更为高效、准确地生成物品流转量预测区间。然后，根据上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，生成上述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集。在这里，通过第一物品流转量分位点确定模型和第二物品流转量分位点确定模型，可以生成更为精准的物品流转量预测区间。可选地，现有方式生成的物品流转量预测区间的区间长度常常是固定长度，导致物品流转量预测区间无法较好地适应异方差数据，物品流转量预测区间长度无法较好地自适应调整。然而，基于第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型生成的物品流转量预测区间可以自适应的调整区间长度，以使得物品流转量预测区间的准确率较高。进而，根据上述物品流转量预测区间集，确定上述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集。在 这里，通过根据物品流转量预测区间确定残差信息，以用于后续物品流转量变换信息的确定。接着，根据上述残差信息集，可以生成更为精准地物品流转量变换信息。最后，根据上述物品流转量变换信息，对上述目标物品数据集对应的物品进行调货处理。在这里，对目标物品数据集对应的物品进行调货处理可以极大较少了库存资源的浪费，大大提高了库存资源利用率。
参考图3，示出了根据本公开的调货方法的另一些实施例的流程300。该调货方法，包括以下步骤：
步骤301，获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型。
步骤302，根据上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，生成上述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集。
步骤303，对于上述第二物品数据子集中每个物品数据，执行以下残差信息确定步骤。
步骤3031，确定上述物品数据对应的物品流转量预测区间。
在一些实施例中，执行主体(例如图1所示的电子设备101)可以从目标数据库中确定物品数据对应的物品流转量预测区间。其中，上述目标数据库可以是存储物品数据与物品流转量预测区间之间关联关系的数据库。
步骤3032，确定上述物品流转量预测区间相对应的第一数值和第二数值。
在一些实施例中，上述执行主体可以确定上述物品流转量预测区间相对应的第一数值和第二数值。作为示例，上述执行主体可以将物品流转量预测区间中的最大值和最小值确定为第一数值和第二数值。
作为示例，上述物品流转量预测区间可以是：[30,70]。则第一数值为70。第二数值为30。
步骤3033，将上述物品流转量预测区间相对应的第一数值和上述物品流转量预测区间相对应的第二数值分别与上述物品数据对应的真实物品流转量数值进行作差，得到第一作差数值和第二作差数值。
在一些实施例中，上述执行主体可以将上述物品流转量预测区间相对应的第一数值和上述物品流转量预测区间相对应的第二数值分别与上述物品数据对应的真实物品流转量数值进行作差，得到第一作差数值和第二作差数值。
作为示例，上述物品数据对应的真实物品流转量数值为60，则第一作差数值为10。第二作差数值为-30。
步骤3034，将上述第一作差数值与上述第二作差数值之间的最大值确定为上述残差信息。
在一些实施例中，上述执行主体可以将上述第一作差数值与上述第二作差数值之间的最大值确定为上述残差信息。
作为示例，上述执行主体可以将10确定为残差信息。
步骤304，根据上述残差信息集，生成物品流转量变换信息。
步骤305，根据上述物品流转量变换信息，对上述目标物品数据集对应的物品进行调货处理。
在一些实施例中，步骤301-302和304-305的具体实现及其所带来的技术效果，可以参考图2对应的实施例中的步骤201-202和204-205，在此不再赘述。
从图3中可以看出，与图2对应的一些实施例的描述相比，图3对应的一些实施例中的调货方法的流程300更加突出了确定物品数据对应残差信息的具体步骤。针对预测区间可能无法很好的适应异方差数据，预测区间长度无法很好的自适应调整。除此之外，在一定的置信度下，普通的机器学习基于训练集进行分位点预测从而得到区间的上下界，但是并没有对上下界做进一步的误差调整，这就导致预测区间过大，从而丧失了区间预测的意义，同时也可能存在一定的误差。由此，这些实施例描述的方案可以避免无法很好的适应异方差数据，预测区间长度无法很好的自适应调整以及没有对上下界做进一步的误差调整的问题，大大提高了生成残差信息的精准性，以使得后续物品 流转量变换信息的生成更为精准。
参考图4，示出了根据本公开的调货方法的又一些实施例的流程400。该调货方法，包括以下步骤：
步骤401，获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型。
步骤402，根据上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，生成上述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集。
步骤403，根据上述物品流转量预测区间集，确定上述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集。
步骤404，根据上述置信水平，从上述残差信息集中筛选出满足预定条件的残差信息，作为上述物品流转量变换信息。
在一些实施例中，执行主体(例如图1中电子设备101)可以根据上述置信水平，从上述残差信息集中筛选出满足预定条件的残差信息，作为上述物品流转量变换信息。
作为示例，上述执行主体可以对残差信息集进行排序，得到残差信息序列。然后，上述执行主体可以从残差信息序列中找到数值大于置信水平残差信息的目标残差信息。例如，置信水平为80％。残差信息序列为：[2,4,5,6,7]。则目标残差信息为6。最后，上述执行主体将目标残差信息确定为物品流转量变换信息。
步骤405，根据上述物品流转量变换信息，对上述目标物品数据集对应的物品进行调货处理。
在一些实施例中，步骤401-403、405的具体实现及其所带来的技术效果，可以参考图2对应的实施例中的步骤201-203、205，在此不再赘述。
从图4中可以看出，与图2对应的一些实施例的描述相比，图4对应的一些实施例中的调货方法的流程400更加突出了确定物品流转量变换信息的具体步骤。由此，这些实施例描述的方案通过预定条件。 可以更为精准的确定出物品流转量变换信息。
参考图5，作为对上述各图所示方法的实现，本公开提供了一种调货装置的一些实施例，这些装置实施例与图2所示的那些方法实施例相对应，该装置具体可以应用于各种电子设备中。
如图5所示，一种调货装置500包括：获取单元501、第一生成单元502、确定单元503、第二生成单元504和调货处理单元505。其中，获取单元501，被配置成获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型，其中，上述第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型是基于目标物品数据集中的第一物品数据子集和预先设置的置信水平训练的；第一生成单元502，被配置成根据上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，生成上述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集；确定单元503，被配置成根据上述物品流转量预测区间集，确定上述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集；第二生成单元504，被配置成根据上述残差信息集，生成物品流转量变换信息。调货处理单元505，被配置成根据上述物品流转量变换信息，对上述目标物品数据集对应的物品进行调货处理。
在一些实施例的一些可选的实现方式中，上述装置500还包括：第三确定单元(图中未显示)。其中，上述第三确定单元可以被配置成：根据上述物品流转量变换信息，确定上述第二物品数据子集中各个物品数据对应的物品流转量区间信息。
在一些实施例的一些可选的实现方式中，上述装置500中的第一生成单元502可以被配置成：将上述物品数据分别输入至上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，得到第一数值和第二数值；根据上述第一数值和上述第二数值，生成上述物品数据对应的物品流转量预测区间。
在一些实施例的一些可选的实现方式中，上述装置500中的确定 单元503可以被配置成：对于上述第二物品数据子集中每个物品数据，执行以下残差信息确定步骤：确定上述物品数据对应的物品流转量预测区间；确定上述物品流转量预测区间相对应的第一数值和第二数值；将上述物品流转量预测区间相对应的第一数值和上述物品流转量预测区间相对应的第二数值分别与上述物品数据对应的真实物品流转量数值进行作差，得到第一作差数值和第二作差数值；将上述第一作差数值与上述第二作差数值之间的最大值确定为上述残差信息。
在一些实施例的一些可选的实现方式中，上述装置500中的第二生成单元504可以被配置成：根据上述置信水平，从上述残差信息集中筛选出满足预定条件的残差信息，作为上述物品流转量变换信息。
在一些实施例的一些可选的实现方式中，上述第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型是通过以下步骤训练的：对上述目标物品数据集进行预处理，得到预处理后的物品数据集；将上述预处理后的物品数据集划分为上述第一物品数据子集和上述第二物品数据子集；根据上述置信水平，确定上述第一物品流转量分位点确定模型的模型结构和上述第二物品流转量分位点确定模型的模型结构；根据上述第一物品数据子集、上述第一物品流转量分位点确定模型的模型结构和上述第二物品流转量分位点确定模型的模型结构，训练上述第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型，得到上述训练后的第一物品流转量分位点确定模型和上述训练后的第二物品流转量分位点确定模型。
可以理解的是，该装置500中记载的诸单元与参考图2描述的方法中的各个步骤相对应。由此，上文针对方法描述的操作、特征以及产生的有益效果同样适用于装置500及其中包含的单元，在此不再赘述。
下面参考图6，其示出了适于用来实现本公开的一些实施例的电子设备(例如图1中的电子设备101)600的结构示意图。图6示出的电子设备仅仅是一个示例，不应对本公开的实施例的功能和使用范围带来任何限制。
如图6所示，电子设备600可以包括处理装置(例如中央处理器、图形处理器等)601，其可以根据存储在只读存储器(ROM)602中的程序或者从存储装置608加载到随机访问存储器(RAM)603中的程序而执行各种适当的动作和处理。在RAM 603中，还存储有电子设备600操作所需的各种程序和数据。处理装置601、ROM 602以及RAM 603通过总线604彼此相连。输入/输出(I/O)接口605也连接至总线604。
通常，以下装置可以连接至I/O接口605：包括例如触摸屏、触摸板、键盘、鼠标、摄像头、麦克风、加速度计、陀螺仪等的输入装置606；包括例如液晶显示器(LCD)、扬声器、振动器等的输出装置607；包括例如磁带、硬盘等的存储装置608；以及通信装置609。通信装置609可以允许电子设备600与其他设备进行无线或有线通信以交换数据。虽然图6示出了具有各种装置的电子设备600，但是应理解的是，并不要求实施或具备所有示出的装置。可以替代地实施或具备更多或更少的装置。图6中示出的每个方框可以代表一个装置，也可以根据需要代表多个装置。
特别地，根据本公开的一些实施例，上文参考流程图描述的过程可以被实现为计算机软件程序。例如，本公开的一些实施例包括一种计算机程序产品，其包括承载在计算机可读介质上的计算机程序，该计算机程序包含用于执行流程图所示的方法的程序代码。在这样的一些实施例中，该计算机程序可以通过通信装置609从网络上被下载和安装，或者从存储装置608被安装，或者从ROM 602被安装。在该计算机程序被处理装置601执行时，执行本公开的一些实施例的方法中限定的上述功能。
需要说明的是，本公开的一些实施例上述的计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质或者是上述两者的任意组合。计算机可读存储介质例如可以是——但不限于——电、磁、光、电磁、红外线、或半导体的系统、装置或器件，或者任意以上的组合。计算机可读存储介质的更具体的例子可以包括但不限于：具有至少一个导线的电连接、便携式计算机磁盘、硬盘、随机访问存储器 (RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑磁盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本公开的一些实施例中，计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。而在本公开的一些实施例中，计算机可读信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式，包括但不限于电磁信号、光信号或上述的任意合适的组合。计算机可读信号介质还可以是计算机可读存储介质以外的任何计算机可读介质，该计算机可读信号介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。计算机可读介质上包含的程序代码可以用任何适当的介质传输，包括但不限于：电线、光缆、RF(射频)等等，或者上述的任意合适的组合。
在一些实施方式中，客户端、服务器可以利用诸如HTTP(HyperText Transfer Protocol，超文本传输协议)之类的任何当前已知或未来研发的网络协议进行通信，并且可以与任意形式或介质的数字数据通信(例如，通信网络)互连。通信网络的示例包括局域网(“LAN”)，广域网(“WAN”)，网际网(例如，互联网)以及端对端网络(例如，ad hoc端对端网络)，以及任何当前已知或未来研发的网络。
上述计算机可读介质可以是上述电子设备中所包含的；也可以是单独存在，而未装配入该电子设备中。上述计算机可读介质承载有一个或者多个程序，当上述一个或者多个程序被该电子设备执行时，使得该电子设备：获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型，其中，上述第一物品流转量分位点确定模型和上述第二物品流转量分位点确定模型是基于目标物品数据集中的第一物品数据子集和预先设置的置信水平训练的；根据上述预先训练的第一物品流转量分位点确定模型和上述预先训练的第二物品流转量分位点确定模型，生成上述目标物品数据集中第二物 品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集；根据上述物品流转量预测区间集，确定上述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集；根据上述残差信息集，生成物品流转量变换信息。根据上述物品流转量变换信息，对上述目标物品数据集对应的物品进行调货处理。
可以以一种或多种程序设计语言或其组合来编写用于执行本公开的一些实施例的操作的计算机程序代码，上述程序设计语言包括面向对象的程序设计语言—诸如Java、Smalltalk、C++，还包括常规的过程式程序设计语言—诸如“C”语言或类似的程序设计语言。程序代码可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中，远程计算机可以通过任意种类的网络——包括局域网(LAN)或广域网(WAN)——连接到用户计算机，或者，可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。
附图中的流程图和框图，图示了按照本公开各种实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段、或代码的一部分，该模块、程序段、或代码的一部分包含至少一个用于实现规定的逻辑功能的可执行指令。也应当注意，在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个接连地表示的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。也要注意的是，框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合，可以用执行规定的功能或操作的专用的基于硬件的系统来实现，或者可以用专用硬件与计算机指令的组合来实现。
描述于本公开的一些实施例中的单元可以通过软件的方式实现，也可以通过硬件的方式来实现。所描述的单元也可以设置在处理器中，例如，可以描述为：一种处理器包括获取单元、第一生成单元、确定单元、第二生成单元、调货处理单元。其中，这些单元的名称在某种 情况下并不构成对该单元本身的限定，例如，获取单元还可以被描述为“获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型的单元”。
本文中以上描述的功能可以至少部分地由至少一个硬件逻辑部件来执行。例如，非限制性地，可以使用的示范类型的硬件逻辑部件包括：现场可编程门阵列(FPGA)、专用集成电路(ASIC)、专用标准产品(ASSP)、片上系统(SOC)、复杂可编程逻辑设备(CPLD)等等。
以上描述仅为本公开的一些较佳实施例以及对所运用技术原理的说明。本领域技术人员应当理解，本公开的实施例中所涉及的发明范围，并不限于上述技术特征的特定组合而成的技术方案，同时也应涵盖在不脱离上述发明构思的情况下，由上述技术特征或其等同特征进行任意组合而形成的其它技术方案。例如上述特征与本公开的实施例中公开的(但不限于)具有类似功能的技术特征进行互相替换而形成的技术方案。

Claims (10)

  1. 一种调货方法，包括：
    获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型，其中，所述第一物品流转量分位点确定模型和所述第二物品流转量分位点确定模型是基于目标物品数据集中的第一物品数据子集和预先设置的置信水平训练的；
    根据所述预先训练的第一物品流转量分位点确定模型和所述预先训练的第二物品流转量分位点确定模型，生成所述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集；
    根据所述物品流转量预测区间集，确定所述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集；
    根据所述残差信息集，生成物品流转量变换信息；
    根据所述物品流转量变换信息，对所述目标物品数据集对应的物品进行调货处理。
  2. 根据权利要求1所述的方法，其中，所述方法还包括：
    根据所述物品流转量变换信息，确定所述第二物品数据子集中各个物品数据对应的物品流转量区间信息。
  3. 根据权利要求1或2所述的方法，其中，所述根据所述预先训练的第一物品流转量分位点确定模型和所述预先训练的第二物品流转量分位点确定模型，生成所述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，包括：
    将所述物品数据分别输入至所述预先训练的第一物品流转量分位点确定模型和所述预先训练的第二物品流转量分位点确定模型，得到第一数值和第二数值；
    根据所述第一数值和所述第二数值，生成所述物品数据对应的物品流转量预测区间。
  4. 根据权利要求1-3之一所述的方法，其中，所述根据所述物品流转量预测区间集，确定所述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集，包括：
    对于所述第二物品数据子集中每个物品数据，执行以下残差信息确定步骤：
    确定所述物品数据对应的物品流转量预测区间；
    确定所述物品流转量预测区间相对应的第一数值和第二数值；
    将所述物品流转量预测区间相对应的第一数值和所述物品流转量预测区间相对应的第二数值分别与所述物品数据对应的真实物品流转量数值进行作差，得到第一作差数值和第二作差数值；
    将所述第一作差数值与所述第二作差数值之间的最大值确定为所述残差信息。
  5. 根据权利要求1-4之一所述的方法，其中，所述根据所述残差信息集，生成物品流转量变换信息，包括：
    根据所述置信水平，从所述残差信息集中筛选出满足预定条件的残差信息，作为所述物品流转量变换信息。
  6. 根据权利要求1-5之一所述的方法，其中，所述第一物品流转量分位点确定模型和所述第二物品流转量分位点确定模型是通过以下步骤训练的：
    对所述目标物品数据集进行预处理，得到预处理后的物品数据集；
    将所述预处理后的物品数据集划分为所述第一物品数据子集和所述第二物品数据子集；
    根据所述置信水平，确定所述第一物品流转量分位点确定模型的模型结构和所述第二物品流转量分位点确定模型的模型结构；
    根据所述第一物品数据子集、所述第一物品流转量分位点确定模型的模型结构和所述第二物品流转量分位点确定模型的模型结构，训 练所述第一物品流转量分位点确定模型和所述第二物品流转量分位点确定模型，得到训练后的第一物品流转量分位点确定模型和训练后的第二物品流转量分位点确定模型。
  7. 一种调货装置，包括：
    获取单元，被配置成获取预先训练的第一物品流转量分位点确定模型和预先训练的第二物品流转量分位点确定模型，其中，所述第一物品流转量分位点确定模型和所述第二物品流转量分位点确定模型是基于目标物品数据集中的第一物品数据子集和预先设置的置信水平训练的；
    第一生成单元，被配置成根据所述预先训练的第一物品流转量分位点确定模型和所述预先训练的第二物品流转量分位点确定模型，生成所述目标物品数据集中第二物品数据子集中每个物品数据对应的物品流转量预测区间，得到物品流转量预测区间集；
    确定单元，被配置成根据所述物品流转量预测区间集，确定所述第二物品数据子集中每个物品数据对应的残差信息，得到残差信息集；
    第二生成单元，被配置成根据所述残差信息集，生成物品流转量变换信息；
    调货处理单元，被配置成根据所述物品流转量变换信息，对所述目标物品数据集对应的物品进行调货处理。
  8. 根据权利要求7所述的装置，其中，所述第一生成单元进一步被配置成：
    将所述物品数据分别输入至所述预先训练的第一物品流转量分位点确定模型和所述预先训练的第二物品流转量分位点确定模型，得到第一数值和第二数值；
    根据所述第一数值和所述第二数值，生成所述物品数据对应的物品流转量预测区间。
  9. 一种电子设备，包括：
    至少一个处理器；
    存储装置，其上存储有至少一个程序，
    当所述至少一个程序被所述至少一个处理器执行，使得所述至少一个处理器实现如权利要求1-6中任一所述的方法。
  10. 一种计算机可读介质，其上存储有计算机程序，其中，所述程序被处理器执行时实现如权利要求1-6中任一所述的方法。
PCT/CN2022/118574 2022-02-17 2022-09-14 调货方法、装置、电子设备和计算机可读介质 WO2023155425A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210148324.4 2022-02-17
CN202210148324.4A CN116664041A (zh) 2022-02-17 2022-02-17 调货方法、装置、电子设备和计算机可读介质

Publications (1)

Publication Number Publication Date
WO2023155425A1 true WO2023155425A1 (zh) 2023-08-24

Family

ID=87577418

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/118574 WO2023155425A1 (zh) 2022-02-17 2022-09-14 调货方法、装置、电子设备和计算机可读介质

Country Status (2)

Country Link
CN (1) CN116664041A (zh)
WO (1) WO2023155425A1 (zh)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190377724A1 (en) * 2018-06-12 2019-12-12 The Vanguard Group, Inc. Device, method, and computer readable medium for large scale electronic processing
CN111832782A (zh) * 2019-04-17 2020-10-27 北京京东振世信息技术有限公司 确定物品物流属性的方法和装置
CN112766842A (zh) * 2020-12-31 2021-05-07 杭州大萝卜网络科技有限公司 一种跨境电商海外仓精准库存管理系统及方法
CN113191806A (zh) * 2021-04-30 2021-07-30 北京沃东天骏信息技术有限公司 确定流量调控目标的方法及装置
CN114004425A (zh) * 2021-12-29 2022-02-01 北京京东振世信息技术有限公司 物品流转信息预测模型生成方法、信息生成方法、装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190377724A1 (en) * 2018-06-12 2019-12-12 The Vanguard Group, Inc. Device, method, and computer readable medium for large scale electronic processing
CN111832782A (zh) * 2019-04-17 2020-10-27 北京京东振世信息技术有限公司 确定物品物流属性的方法和装置
CN112766842A (zh) * 2020-12-31 2021-05-07 杭州大萝卜网络科技有限公司 一种跨境电商海外仓精准库存管理系统及方法
CN113191806A (zh) * 2021-04-30 2021-07-30 北京沃东天骏信息技术有限公司 确定流量调控目标的方法及装置
CN114004425A (zh) * 2021-12-29 2022-02-01 北京京东振世信息技术有限公司 物品流转信息预测模型生成方法、信息生成方法、装置

Also Published As

Publication number Publication date
CN116664041A (zh) 2023-08-29

Similar Documents

Publication Publication Date Title
CN111427971B (zh) 用于计算机系统的业务建模方法、装置、系统和介质
CN114004425B (zh) 物品流转信息预测模型生成方法、信息生成方法、装置
CN115085196B (zh) 电力负荷预测值确定方法、装置、设备和计算机可读介质
CN114202130A (zh) 流转量预测多任务模型生成方法、调度方法、装置和设备
CN114049072B (zh) 指标确定方法、装置、电子设备和计算机可读介质
CN115271757A (zh) 需求信息生成方法、装置、电子设备和计算机可读介质
CN112182374B (zh) 库存控制方法、装置、电子设备和计算机可读介质
CN113435632A (zh) 信息生成方法、装置、电子设备和计算机可读介质
CN113159453A (zh) 资源数据预测方法、装置、设备及存储介质
CN112598337A (zh) 针对物品的车辆控制方法、装置、设备和计算机可读介质
CN115660244B (zh) 路线信息生成方法、装置、设备和介质
CN112308378A (zh) 调度设备控制方法、装置、电子设备和计算机可读介质
WO2023155425A1 (zh) 调货方法、装置、电子设备和计算机可读介质
CN111709784A (zh) 用于生成用户留存时间的方法、装置、设备和介质
CN111985967A (zh) 物品信息生成方法、装置、电子设备和计算机可读介质
CN111767290A (zh) 用于更新用户画像的方法和装置
CN110619537A (zh) 用于生成信息的方法和装置
CN114792258B (zh) 信息生成方法、装置、电子设备和计算机可读介质
CN118096192B (zh) 基于图神经网络的信息推送方法、装置、设备和介质
CN110633596A (zh) 预测车辆方向角的方法和装置
CN112015997B (zh) 页面显示方法、装置、电子设备和计算机可读介质
WO2024119865A1 (zh) 预测信息生成方法、装置、设备、介质和程序产品
CN112073202B (zh) 信息生成方法、装置、电子设备和计算机可读介质
CN112231585B (zh) 推荐信息显示方法、装置、设备和计算机可读介质
US20230283073A1 (en) Load prediction method and apparatus, electronic device and storage medium

Legal Events

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

Ref document number: 22926730

Country of ref document: EP

Kind code of ref document: A1