Replenishment information generation method and device, equipment, medium and program product

Abstract

The embodiment of the invention discloses a replenishment information generation method and device, equipment, a medium and a program product. A specific embodiment of the method comprises the steps of inputting article circulation data information of a target article into a coding layer of a pre-trained replenishment information generation model to obtain target value information, target flow information and target circulation volume information; inputting the target flow information, the target value information and the target flow information into a decoding layer of a replenishment information generation model to obtain a predicted inventory and a predicted flow set; and according to the predicted stock amount and the predicted flow amount set, generating replenishment information corresponding to the target article. The implementation mode is related to end-to-end replenishment of the goods, the accuracy of the generated replenishment information of the target goods is improved, and the situation that the goods stock is insufficient or the goods stock is overstocked is reduced.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technology, and in particular to a replenishment information generation method, apparatus, device, medium, and program product. Background Art

[0002] By predicting the replenishment quantity of items and performing end-to-end replenishment of items, it is possible to optimize the full-time automatic replenishment of items. At present, when predicting the replenishment quantity of items, the commonly used method is to predict the replenishment quantity of items by predicting the circulation volume of items.

[0003] However, the inventors have discovered that when the above method is used to predict the replenishment quantity of items, the following technical problems often occur: the replenishment quantity of items is predicted only by predicting the turnover of items, without considering predicting the replenishment quantity of items by predicting the flow and value of items, resulting in low accuracy of the predicted replenishment information, leading to insufficient storage of items or excessive backlog of items.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the inventive concept and therefore it may contain information that does not form the prior art that is already known in this country to a person of ordinary skill in the art. Summary of the invention

[0005] The content of this disclosure is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.

[0006] Some embodiments of the present disclosure propose a replenishment information generation method, apparatus, electronic device, computer-readable medium, and program product to solve one or more of the technical problems mentioned in the above background technology section.

[0007] In a first aspect, some embodiments of the present disclosure provide a replenishment information generation method, the method comprising: inputting item circulation data information of a target item into an encoding layer of a pre-trained replenishment information generation model to obtain value coding information, flow coding information and turnover volume coding information; inputting the above-mentioned flow coding information, the above-mentioned value coding information and the above-mentioned turnover volume coding information into a decoding layer of the above-mentioned replenishment information generation model to obtain a predicted inventory quantity and a predicted turnover volume set; and generating replenishment information corresponding to the above-mentioned target item according to the above-mentioned predicted inventory quantity and the above-mentioned predicted turnover volume set.

[0008] Optionally, the replenishment information generation model is trained through the following steps: training a first replenishment information generation model based on the historical circulation information set of the target item; determining a target inventory quantity and a target circulation quantity set corresponding to the first replenishment information generation model; and training a replenishment information generation model based on the historical circulation information set, the target inventory quantity and the target circulation quantity set.

[0009] Optionally, the training of the first replenishment information generation model based on the historical circulation information set of the target item includes: performing the following target training steps based on the historical circulation information set: determining the historical value information, historical flow information and target historical inventory corresponding to the historical circulation information set; inputting the historical circulation information set into the encoding layer of the first initial model to obtain first value coding information, first flow coding information and first circulation volume coding information; inputting the obtained first value coding information, first flow coding information, first circulation volume coding information, historical value information and historical flow information into the decoding layer of the first initial model to obtain the first predicted inventory and the first predicted circulation volume set; determining whether the first initial model reaches a preset optimization target based on the first predicted inventory, the first predicted circulation volume set and the target historical inventory; in response to determining that the first initial model reaches the preset optimization target, determining the first initial model as the target replenishment information model.

[0010] Optionally, the above-mentioned determining whether the first initial model reaches the preset optimization target based on the first predicted inventory quantity, the first predicted turnover volume set and the target historical inventory quantity includes: generating a loss value corresponding to the above-mentioned historical circulation information set based on the first predicted inventory quantity, the first predicted turnover volume set, the target historical inventory quantity and the historical turnover volume set corresponding to the first predicted turnover volume set; determining whether the first initial model reaches the preset optimization target based on the loss value corresponding to the above-mentioned historical circulation information set.

[0011] Optionally, after determining whether the first initial model reaches the preset optimization target based on the first predicted inventory, the first predicted turnover volume set and the target historical inventory, the method further includes: in response to determining that the first initial model does not reach the above-mentioned preset optimization target, adjusting the network parameters of the first initial model, using the adjusted first initial model as the first initial model, and performing the above-mentioned target training step again.

[0012] Optionally, the replenishment information generation model is trained according to the historical circulation information set, the target inventory and the target circulation volume set, including: executing the following training steps according to the historical circulation information set: inputting the historical circulation information set into the encoding layer of the second initial model to obtain second value coding information, second flow coding information and second circulation volume coding information; inputting the obtained second value coding information, second flow coding information and second circulation volume coding information into the decoding layer of the second initial model to obtain a second predicted circulation volume set and a second predicted inventory corresponding to the historical circulation information set; determining whether the second initial model reaches the preset optimization target according to the second predicted circulation volume set, the second predicted inventory and the target historical inventory corresponding to the historical circulation information set; in response to determining that the second initial model reaches the preset optimization target, determining the second initial model as the replenishment information model.

[0013] Optionally, the above-mentioned determination of whether the second initial model reaches the above-mentioned preset optimization target based on the second predicted turnover volume set, the second predicted inventory volume and the target historical inventory volume corresponding to the above-mentioned historical turnover information set includes: generating a first characteristic loss value based on the second predicted turnover volume set, the second predicted inventory volume, the above-mentioned target inventory volume and the above-mentioned target turnover volume set; determining a second characteristic loss value based on the target historical inventory volume of the historical circulation information set, the second predicted inventory volume, the second predicted turnover volume set and the historical turnover volume set corresponding to the second predicted turnover volume set; generating a joint characteristic loss value of the above-mentioned historical circulation information set based on the first characteristic loss value and the second characteristic loss value; and determining whether the above-mentioned second initial model reaches the above-mentioned preset optimization target based on the joint characteristic loss value.

[0014] Optionally, after determining whether the second initial model reaches the preset optimization target based on the second predicted turnover volume set, the second predicted inventory quantity and the target historical inventory quantity corresponding to the historical turnover information set, the method further includes: in response to determining that the second initial model does not reach the preset optimization target, adjusting the network parameters of the second initial model, using the adjusted second initial model as the second initial model, and executing the above training step again.

[0015] Optionally, the above-mentioned generating replenishment information corresponding to the above-mentioned target item based on the above-mentioned predicted inventory quantity and the above-mentioned predicted circulation quantity set includes: determining the inventory quantity corresponding to the above-mentioned target item from the above-mentioned item circulation data information; determining the replenishment inventory quantity corresponding to the above-mentioned target item based on the above-mentioned predicted inventory quantity and the above-mentioned inventory quantity; and determining the above-mentioned replenishment inventory quantity and the above-mentioned predicted circulation quantity set as the replenishment information corresponding to the above-mentioned target item.

[0016] Optionally, after generating the replenishment information corresponding to the target item based on the predicted inventory quantity and the predicted turnover volume set, the method further includes: acquiring the inventory location information corresponding to the target item; controlling the associated robotic arm to grab each item corresponding to the replenishment information based on the inventory location information and the replenishment information; and controlling the associated robotic arm to perform item scheduling operations on each grabbed item.

[0017] In a second aspect, some embodiments of the present disclosure provide a replenishment information generating device, the device comprising: a first input unit, configured to input item circulation data information of a target item into an encoding layer of a pre-trained replenishment information generating model, to obtain value coding information, flow coding information and circulation volume coding information; a second input unit, configured to input the above-mentioned flow coding information, the above-mentioned value coding information and the above-mentioned circulation volume coding information into a decoding layer of the above-mentioned replenishment information generating model, to obtain a predicted inventory quantity and a predicted circulation volume set; a generating unit, configured to generate replenishment information corresponding to the above-mentioned target item according to the above-mentioned predicted inventory quantity and the above-mentioned predicted circulation volume set.

[0018] Optionally, the first input unit is further configured as follows: the replenishment information generation model is trained through the following steps: training a first replenishment information generation model based on a historical circulation information set of the target item; determining a target inventory quantity and a target circulation quantity set corresponding to the first replenishment information generation model; and training a replenishment information generation model based on the historical circulation information set, the target inventory quantity and the target circulation quantity set.

[0019] Optionally, the first input unit is further configured to: train the first replenishment information generation model based on the historical circulation information set of the target item, including: executing the following target training steps based on the historical circulation information set: determining the historical value information, historical flow information and target historical inventory corresponding to the historical circulation information set; inputting the historical circulation information set into the encoding layer of the first initial model to obtain first value coding information, first flow coding information and first circulation volume coding information; inputting the obtained first value coding information, first flow coding information, first circulation volume coding information, historical value information and historical flow information into the decoding layer of the first initial model to obtain the first predicted inventory and the first predicted circulation volume set; determining whether the first initial model reaches the preset optimization target based on the first predicted inventory, the first predicted circulation volume set and the target historical inventory; in response to determining that the first initial model reaches the preset optimization target, determining the first initial model as the target replenishment information model.

[0020] Optionally, the first input unit is further configured to: determine whether the first initial model reaches the preset optimization target based on the first predicted inventory, the first predicted turnover volume set and the target historical inventory, including: generating a loss value corresponding to the historical circulation information set based on the first predicted inventory, the first predicted turnover volume set, the target historical inventory and the historical turnover volume set corresponding to the first predicted turnover volume set; determine whether the first initial model reaches the preset optimization target based on the loss value corresponding to the historical circulation information set.

[0021] Optionally, the first input unit is further configured to: after determining whether the first initial model reaches the preset optimization target based on the first predicted inventory, the first predicted turnover volume set and the target historical inventory, the method further includes: in response to determining that the first initial model does not reach the preset optimization target, adjusting the network parameters of the first initial model, using the adjusted first initial model as the first initial model, and performing the target training step again.

[0022] Optionally, the first input unit is further configured to: train a replenishment information generation model based on the historical circulation information set, the target inventory and the target circulation volume set, including: performing the following training steps based on the historical circulation information set: inputting the historical circulation information set into the encoding layer of the second initial model to obtain second value coding information, second flow coding information and second circulation volume coding information; inputting the obtained second value coding information, second flow coding information and second circulation volume coding information into the decoding layer of the second initial model to obtain a second predicted circulation volume set and a second predicted inventory corresponding to the historical circulation information set; determining whether the second initial model reaches the preset optimization target based on the second predicted circulation volume set, the second predicted inventory and the target historical inventory corresponding to the historical circulation information set; in response to determining that the second initial model reaches the preset optimization target, determining the second initial model as the replenishment information model.

[0023] Optionally, the first input unit is further configured to: determine whether the second initial model reaches the preset optimization target based on the second predicted turnover volume set, the second predicted inventory and the target historical inventory corresponding to the historical turnover information set, including: generating a first characteristic loss value based on the second predicted turnover volume set, the second predicted inventory, the target inventory and the target turnover volume set; determining a second characteristic loss value based on the target historical inventory of the historical turnover information set, the second predicted inventory, the second predicted turnover volume set and the historical turnover volume set corresponding to the second predicted turnover volume set; generating a joint characteristic loss value of the historical turnover information set based on the first characteristic loss value and the second characteristic loss value; and determining whether the second initial model reaches the preset optimization target based on the joint characteristic loss value.

[0024] Optionally, the first input unit is further configured to: after determining whether the second initial model reaches the preset optimization target based on the second predicted turnover volume set, the second predicted inventory quantity and the target historical inventory quantity corresponding to the historical turnover information set, the method further includes: in response to determining that the second initial model does not reach the preset optimization target, adjusting the network parameters of the second initial model, using the adjusted second initial model as the second initial model, and executing the training step again.

[0025] Optionally, the generation unit is further configured to: generate replenishment information corresponding to the target item based on the predicted inventory quantity and the predicted turnover volume set, including: determining the inventory quantity corresponding to the target item from the item turnover data information; determining the replenishment inventory quantity corresponding to the target item based on the predicted inventory quantity and the inventory quantity; and determining the replenishment inventory quantity and the predicted turnover volume set as the replenishment information corresponding to the target item.

[0026] Optionally, the replenishment information generating device further includes: an acquisition unit configured to acquire inventory location information corresponding to the target item; a first control unit configured to control an associated robotic arm to grab each item corresponding to the target item based on the inventory location information and the replenishment information; and a second control unit configured to control the associated robotic arm to perform item scheduling operations on each grabbed item.

[0027] In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner of the above-mentioned first aspect.

[0028] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any implementation manner of the above-mentioned first aspect is implemented.

[0029] In a fifth aspect, some embodiments of the present disclosure provide a computer program product, including a computer program, which implements the method described in any implementation manner of the above-mentioned first aspect when executed by a processor.

[0030] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the replenishment information generation method of some embodiments of the present disclosure, the accuracy of the replenishment information is improved, and the situation of insufficient items or excessive backlog of items is reduced. Specifically, the reason for the low accuracy of the replenishment information, insufficient items or excessive backlog of items is that the replenishment quantity of the item is predicted only by predicting the circulation volume of the item, and the replenishment quantity of the item is predicted by predicting the flow and value of the item, resulting in low accuracy of the predicted replenishment information, insufficient stored items or excessive backlog of items. Based on this, the replenishment information generation method of some embodiments of the present disclosure, first, inputs the item circulation data information of the target item into the encoding layer of the pre-trained replenishment information generation model to obtain value encoding information, flow encoding information and circulation volume encoding information. Thus, encoding information representing the predicted value, flow and circulation volume of the target item can be obtained. Then, the above-mentioned flow encoding information, the above-mentioned value encoding information and the above-mentioned circulation volume encoding information are input into the decoding layer of the above-mentioned replenishment information generation model to obtain the predicted inventory and the predicted circulation volume set. Thus, the predicted inventory and each circulation volume of the target item can be obtained. Finally, according to the above-mentioned predicted inventory quantity and the above-mentioned predicted turnover volume set, the replenishment information corresponding to the above-mentioned target item is generated. In this way, the replenishment information representing the replenishment quantity and turnover volume of the above-mentioned target item can be obtained. The above-mentioned replenishment information can represent the predicted replenishment quantity and turnover volume of the above-mentioned target item. Also, because the replenishment information is not predicted only by the predicted turnover volume, but by the predicted value, predicted flow and predicted turnover volume, the accuracy of the replenishment information is improved, and the situation of insufficient quantity of items and backlog of items can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.

[0032] Figure 1 is a schematic diagram of an application scenario of the replenishment information generation method disclosed in the present invention;

[0033] Figure 2 is a flow chart of some embodiments of the replenishment information generation method according to the present disclosure;

[0034] Figure 3 are flow charts of other embodiments of the replenishment information generation method according to the present disclosure;

[0035] Figure 4 is a schematic structural diagram of some embodiments of a replenishment information generating device according to the present disclosure;

[0036] Figure 5 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0037] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0038] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be combined with each other.

[0039] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0040] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0041] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0042] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0043] Figure 1 It is a schematic diagram of an application scenario of the replenishment information generation method according to some embodiments of the present disclosure.

[0044] exist Figure 1In the application scenario, first, the execution subject 101 may input the object circulation data information 102 of the target object into the encoding layer of the pre-trained replenishment information generation model 103 to obtain the value encoding information 104, the flow encoding information 105 and the circulation volume encoding information 106. Then, the execution subject 101 may input the flow encoding information 105, the value encoding information 104 and the circulation volume encoding information 106 into the decoding layer of the replenishment information generation model 103 to obtain the predicted inventory 107 and the predicted circulation volume set 108. Finally, the execution subject 101 may generate the replenishment information 109 corresponding to the target object according to the predicted inventory 107 and the predicted circulation volume set 108.

[0045] It should be noted that the computing device 101 can be hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or it can be implemented as a single server or a single terminal device. When the computing device is embodied as software, it can be installed in the hardware devices listed above. It can be implemented as multiple software or software modules for providing distributed services, or it can be implemented as a single software or software module. No specific limitation is made here.

[0046] It should be understood that Figure 1 The number of computing devices in the embodiment is only illustrative. Any number of computing devices may be provided according to implementation requirements.

[0047] Continue to refer Figure 2 , shows a process 200 of some embodiments of the replenishment information generation method according to the present disclosure. The replenishment information generation method comprises the following steps:

[0048] Step 201, inputting the object circulation data information of the target object into the encoding layer of the pre-trained replenishment information generation model to obtain value encoding information, flow encoding information and circulation volume encoding information.

[0049] In some embodiments, the execution body of the replenishment information generation method (for example Figure 1The computing device 101 shown can obtain value coding information, flow coding information and circulation coding information by inputting the object circulation data information of the target object into the coding layer of the pre-trained replenishment information generation model. Among them, the above-mentioned object circulation data information may include historical object data information and historical contemporaneous object data information. The above-mentioned historical object data information may be data information of the preset number of days before the prediction time interval and data corresponding to the prediction time interval. The above-mentioned prediction time interval may represent the time period of the predicted replenishment information. The above-mentioned preset number of days may represent the number of days before the prediction time interval. The above-mentioned historical contemporaneous object data information may represent the data information corresponding to the historical contemporaneous time period corresponding to the prediction time interval and the number of days before the prediction time interval. The above-mentioned historical contemporaneous time period may represent the prediction time interval of the same period and the time period corresponding to the preset number of days before the prediction time interval. For example, the above-mentioned prediction time interval may be April 1, 2023-May 1, 2023. The above-mentioned preset number of days before the prediction time interval may correspond to the number of days from 180 days before April 1, 2023 to April 1, 2023. The above historical period may be April 1, 2022 to May 1, 2022 and 180 days before April 1, 2022 to April 1, 2022. The historical period item data information may be data corresponding to April 1, 2022 to May 1, 2022 and 180 days before April 1, 2022 to April 1, 2022. The above historical item data information may be data corresponding to the first 180 days of April 1, 2023. The above value encoding information may represent the predicted value of the corresponding prediction time interval. The above flow encoding information may represent the predicted flow of the corresponding prediction time interval. The above turnover encoding information may represent the predicted turnover of the corresponding prediction time interval. For example, the above predicted flow may be the number of users browsing items online. The above replenishment information generation model may be used to generate replenishment information of the target item. The above replenishment information generation model may take the item historical data information as input and take the predicted inventory and turnover of the target item as output. The above encoding layer may be a time series encoder. In practice, the execution entity may input the item data information into the coding layer of the replenishment information generation model to obtain the circulation volume coding information corresponding to the target item.

[0050] As an example, the above-mentioned circulation volume coding information can represent the circulation quantity of the above-mentioned target item in the predicted time interval.

[0051] Step 202, input the flow coding information, value coding information and turnover coding information into the decoding layer of the replenishment information generation model to obtain the predicted inventory and predicted turnover sets.

[0052] In some embodiments, the execution entity may input the flow coding information, the value coding information and the turnover coding information into the decoding layer of the replenishment information generation model to obtain the predicted inventory and the predicted turnover set. Among them, the decoding layer may be a full-link decoder. The predicted inventory may represent the inventory corresponding to the prediction time interval. The predicted turnover set may represent each turnover corresponding to the prediction time interval. The turnover may be the turnover for each day in the prediction time interval. In practice, the execution entity may input the turnover coding information into the decoding layer of the replenishment information generation model or into the dynamic inventory model to obtain the inventory information corresponding to the target item.

[0053] As an example, the above dynamic inventory model (Dynamic Target Inventor Adjustment) may be a model that takes the circulation quantity coding information as input and takes the inventory information as output. The above inventory information may represent the inventory quantity and circulation quantity information.

[0054] Step 203: Generate replenishment information for the corresponding target item based on the predicted inventory quantity and the predicted turnover quantity set.

[0055] In some embodiments, the execution entity may generate replenishment information corresponding to the target item based on the predicted inventory and the predicted turnover. In practice, first, the execution entity may determine the replenishment quantity of the target item based on the predicted inventory of the target item. Then, the replenishment quantity may be determined as the predicted turnover. Finally, the turnover and the replenishment quantity are determined as the replenishment information.

[0056] As an example, the replenishment quantity may represent the replenishment quantity of the target item corresponding to the predicted time interval. The replenishment information may represent the replenishment quantity and the turnover volume of the predicted time interval. The turnover volume may represent the turnover quantity corresponding to the predicted time interval.

[0057] In some optional implementations of some embodiments, the execution entity may generate replenishment information corresponding to the target item according to the predicted inventory quantity and the predicted turnover quantity set through the following steps:

[0058] The first step is to determine the inventory quantity corresponding to the target item from the item circulation data information. The inventory quantity may represent the remaining inventory of the target item. The item circulation data information may include the inventory quantity.

[0059] The second step is to determine the replenishment inventory quantity corresponding to the target item based on the predicted inventory quantity and the inventory quantity. The replenishment inventory quantity can represent the replenishment quantity of the target item. In practice, the execution entity can determine the difference between the inventory quantity and the predicted inventory quantity as the replenishment inventory quantity.

[0060] The third step is to determine the replenishment inventory and the predicted turnover set as replenishment information corresponding to the target item. The replenishment information may represent the replenishment quantity and turnover set of the target item.

[0061] Optionally, after generating the replenishment information corresponding to the target item according to the predicted inventory quantity and the predicted turnover quantity set, first, the execution subject may also obtain the inventory location information corresponding to the target item. The inventory location information may represent the storage location of the target item. For example, the storage location may be the 5th floor of the 4th shelf in area A.

[0062] Then, according to the inventory location information and the replenishment information, the associated robotic arm can be controlled to grab each item of the target item corresponding to the replenishment information. In practice, first, the execution subject can determine the replenishment quantity according to the replenishment information. The replenishment information includes the replenishment quantity corresponding to the target item. Then, the associated robotic arm can be controlled to grab each item corresponding to the replenishment quantity.

[0063] Finally, the associated robotic arm can be controlled to perform item dispatching operations on each grabbed item. In practice, first, the above execution subject can obtain the dispatched warehouse area of ​​the target item. Then, the associated robotic arm is controlled to dispatch the item to the corresponding warehouse area.

[0064] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the replenishment information generation of some embodiments of the present disclosure, the accuracy of the replenishment information is improved, and the situation of insufficient items or excessive backlog of items is reduced. Specifically, the reason for the low accuracy of the replenishment information, insufficient items or excessive backlog of items is that the replenishment quantity of the item is predicted only by predicting the circulation volume of the item, and the replenishment quantity of the item is not considered to be predicted by predicting the flow and value of the item, resulting in low accuracy of the predicted replenishment information, resulting in insufficient stored items or excessive backlog of items. Based on this, the replenishment information generation method of some embodiments of the present disclosure, first, inputs the item circulation data information of the target item into the encoding layer of the pre-trained replenishment information generation model to obtain value encoding information, flow encoding information and circulation volume encoding information. Thus, encoding information representing the predicted value, flow and circulation volume of the target item can be obtained. Then, the above-mentioned flow encoding information, the above-mentioned value encoding information and the above-mentioned circulation volume encoding information are input into the decoding layer of the above-mentioned replenishment information generation model to obtain the predicted inventory and predicted circulation volume set. Thus, the predicted inventory and circulation volume of the target item can be obtained. Finally, according to the above-mentioned predicted inventory quantity and the above-mentioned predicted turnover volume set, the replenishment information corresponding to the above-mentioned target item is generated. In this way, the replenishment information representing the replenishment quantity and turnover volume of the above-mentioned target item can be obtained. Also, because the replenishment information is not predicted only by the predicted turnover volume, but the replenishment information is generated by the predicted value, predicted flow and predicted turnover volume, the accuracy of the generated replenishment information is improved, and the situation of insufficient quantity of items and backlog of items can be reduced.

[0065] Further references Figure 3 , which shows a process 300 of another embodiment of a replenishment information generation method. The process 300 of the replenishment information generation method includes the following steps:

[0066] Step 301, inputting the object circulation data information of the target object into the encoding layer of the pre-trained replenishment information generation model to obtain value encoding information, flow encoding information and circulation volume encoding information.

[0067] Step 302, input the flow coding information, value coding information and turnover coding information into the decoding layer of the replenishment information generation model to obtain the predicted inventory and predicted turnover sets.

[0068] Step 303: Generate replenishment information for the corresponding target item based on the predicted inventory quantity and the predicted turnover quantity set.

[0069] In some embodiments, the specific implementation of steps 301-303 and the technical effects brought about can be referred to Figure 2 The corresponding steps 201-203 in the embodiments are not described in detail here.

[0070] Step 304: Train a first replenishment information generation model based on the historical circulation information set of the target item.

[0071] In some embodiments, the execution subject may train a first replenishment information generation model based on a historical circulation information set of the target item. The first replenishment information model is a model that takes a historical circulation information set of the target item as input and a predicted inventory and a predicted circulation set as output. The historical circulation information in the historical circulation information set may include historical item data information and historical contemporaneous item data information. The historical item data information may be data information of a preset number of days before the prediction time interval and data corresponding to the prediction time interval. The prediction time interval may represent a time period of the predicted replenishment information. The preset number of days may represent the number of days before the prediction time interval. The historical contemporaneous item data information may represent data information corresponding to the historical contemporaneous time period corresponding to the prediction time interval and the number of days before the prediction time interval. The historical contemporaneous time period may represent the prediction time interval of the same period and the time period corresponding to the preset number of days before the prediction time interval. It should be noted that the first replenishment information generation model has the same model structure as the replenishment information generation model, and therefore, it will not be described in detail here.

[0072] In some optional implementations of some embodiments, the execution subject may train a first replenishment information generation model according to the historical flow information set of the target item through the following steps:

[0073] The first step is to perform the following target training steps based on the above historical flow information set:

[0074] The first sub-step is to determine the historical value information, historical flow information and target historical inventory corresponding to the above-mentioned historical circulation information set. Among them, the above-mentioned historical value information can represent the value information corresponding to the predicted time interval. The above-mentioned historical flow information can represent the flow information corresponding to the predicted time interval. The above-mentioned target historical inventory can represent the historical optimal target inventory corresponding to the predicted time interval. The above-mentioned historical optimal target inventory can represent the historical optimal inventory of the target item. In practice, the above-mentioned execution subject can obtain the above-mentioned target historical inventory through a dynamic programming model. The above-mentioned execution subject can input the historical circulation volume and historical inventory volume of the preset time period corresponding to the above-mentioned target item into the above-mentioned dynamic programming model to obtain the target historical inventory volume corresponding to the above-mentioned target item. The above-mentioned dynamic programming model can be a model that takes the historical circulation volume and historical inventory volume within a preset time period as input and the target historical inventory volume of the item as output. The above-mentioned dynamic programming model can be a model required to solve the optimization problem of the decision-making process through the dynamic programming method, and is a model that uses reverse thinking to solve problems. It should be noted that the above-mentioned historical value information and the above-mentioned historical flow information represent the historical data of the prediction time interval. The real value of the prediction time interval is used in training the first initial model, which can be used to improve the accuracy of the turnover and inventory predicted by training the second initial model.

[0075] The second sub-step is to input the above-mentioned historical circulation information set into the coding layer of the first initial model to obtain the first value coding information, the first flow coding information and the first circulation volume coding information. Among them, the above-mentioned first value coding information can represent the value information of the corresponding prediction time interval. The above-mentioned first flow coding information can represent the flow of the corresponding prediction time interval. The above-mentioned first circulation volume coding information can represent the circulation volume of the corresponding prediction time interval. It should be noted that the first value coding information and the first flow coding information output by the timing prediction layer included in the coding layer reflect that the predicted inventory and circulation volume take into account the predicted value information and flow information, thereby improving the accuracy of the predicted inventory and circulation volume.

[0076] In the third sub-step, the obtained first value coding information, first flow coding information, first turnover coding information, historical value information and historical flow information are input into the decoding layer of the first initial model to obtain the first predicted inventory and the first predicted turnover set. The first predicted inventory can be the predicted inventory of the predicted time interval. The predicted turnover set can be each turnover of the predicted time interval.

[0077] The fourth sub-step is to determine whether the first initial model reaches a preset optimization target based on the first predicted inventory, the first predicted turnover set, and the target historical inventory. The preset optimization target may indicate that the loss value of the first initial model satisfies a preset difference condition. The preset difference condition may indicate that the loss value is different from each loss value of the first initial model history.

[0078] In the second step, in response to determining that the first initial model achieves the above-mentioned preset optimization target, the first initial model is determined as the target replenishment information model.

[0079] In some optional implementations of some embodiments, the execution entity may determine whether the first initial model reaches a preset optimization target according to the first predicted inventory, the first predicted turnover set, and the target historical inventory through the following steps:

[0080] The first step is to generate a loss value corresponding to the above historical circulation information set based on the first predicted inventory, the first predicted circulation set, the target historical inventory and the historical circulation set corresponding to the first predicted circulation set. The above historical circulation set can characterize the actual circulation of the above target item in the predicted time interval. In practice, first, the above execution entity can input the above first predicted inventory, the first predicted circulation set, the historical circulation set and the target historical inventory into the hard target loss function to obtain the loss value corresponding to the above historical circulation information set.

[0081] As an example, the above hard loss function can be:

[0082] hardloss=∑[y s -y t ]+∑[y b -y a ].

[0083] Among them, y s It can be a first predicted flow amount in the first predicted flow amount set. t It can be a historical flow volume in the above historical flow volume set. The first predicted flow volume in the above first predicted flow volume set corresponds to the historical flow volume in the above historical flow volume set. b The first predicted inventory quantity can be a It can be the above-mentioned target historical inventory. Thus, the loss value corresponding to the above-mentioned historical circulation information set can be obtained.

[0084] The second step is to determine whether the first initial model has reached the preset optimization target according to the loss value corresponding to the above-mentioned historical flow information set. In practice, first, the above-mentioned execution subject can determine whether the above-mentioned loss value satisfies the above-mentioned preset difference condition. The above-mentioned each historical loss value can be each loss value corresponding to each preset number of training times before the number of training times corresponding to the loss value. For example, the number of training times of the first initial model corresponding to the above-mentioned loss value is the 10th time. The above-mentioned each preset number of training times can be the ninth time, the eighth time, and the seventh time. Then, in response to determining that the number of training times of the first initial model satisfies the preset order condition and the loss value does not satisfy the above-mentioned preset difference condition, it is determined that the above-mentioned first initial model has reached the preset optimization target. The above-mentioned preset order condition can be that the number of training times is less than the preset number of training rounds. For example, the above-mentioned preset number of training rounds can be 15 rounds. Finally, in response to determining that the number of training times of the first initial model satisfies the above-mentioned preset order condition and the loss value satisfies the above-mentioned preset difference condition, it is determined that the above-mentioned first initial model has not reached the preset optimization target.

[0085] Optionally, after determining whether the first initial model has reached the preset optimization target based on the first predicted inventory, the first predicted turnover set, and the target historical inventory, the execution subject may also adjust the network parameters of the first initial model in response to determining that the first initial model has not reached the preset optimization target, use the adjusted first initial model as the first initial model, and perform the target training step again. In practice, the execution subject may adjust the network parameters by back propagation, stochastic gradient descent, and other methods.

[0086] Step 305: determine a target inventory quantity and a target turnover quantity set corresponding to the first replenishment information generation model.

[0087] In some embodiments, the execution entity may determine a target inventory and a target turnover set corresponding to the first replenishment information generation model. The target inventory and the turnover set may represent a set of inventory and turnover corresponding to a loss value that satisfies a preset optimization target. The target inventory may be the inventory predicted by the first replenishment information generation model. The target turnover set may be each turnover predicted by the first replenishment information generation model.

[0088] Step 306: Train a replenishment information generation model based on the historical circulation information set, the target inventory quantity, and the target circulation quantity set.

[0089] In some embodiments, the execution entity may train a replenishment information generation model based on the historical circulation information set, the target inventory level, and the target circulation volume set.

[0090] In some optional implementations of some embodiments, the execution entity may train a replenishment information generation model according to the historical circulation information set, the target inventory quantity, and the target circulation quantity set through the following steps:

[0091] The first step is to perform the following training steps based on the above historical flow information set:

[0092] In the first sub-step, the historical circulation information set is input into the coding layer of the second initial model to obtain the second value coding information, the second flow coding information and the second circulation volume coding information. The second value coding information, the second flow coding information and the second circulation volume coding information can represent the value, flow and circulation volume of the target item in the predicted time interval.

[0093] The second sub-step is to input the obtained second value coding information, second flow coding information and second circulation coding information into the decoding layer of the second initial model to obtain a second predicted circulation volume set and a second predicted inventory volume corresponding to the historical circulation information set. The second predicted circulation volume set and the second predicted inventory volume can represent the predicted circulation volume and inventory volume corresponding to the target item.

[0094] The third sub-step is to determine whether the second initial model has achieved the above-mentioned preset optimization goal based on the second predicted turnover volume set, the second predicted inventory volume and the target historical inventory volume corresponding to the above-mentioned historical turnover information set.

[0095] In the second step, in response to determining that the second initial model reaches the preset optimization target, the second initial model is determined as the replenishment information model. It should be noted that the method of determining whether the second initial model reaches the preset optimization target is the same as the method of determining whether the first initial model reaches the preset optimization target. Therefore, it will not be repeated here.

[0096] In some optional implementations of some embodiments, the execution subject may determine whether the second initial model achieves the preset optimization goal according to the second predicted turnover volume set, the second predicted inventory volume, and the target historical inventory volume corresponding to the historical turnover information set through the following steps:

[0097] In the first step, a first characteristic loss value is generated according to the second predicted turnover set, the second predicted inventory, the target inventory and the target turnover set. In practice, first, the execution subject may input the second predicted turnover set, the second predicted inventory, the target inventory and the target turnover set into the soft objective function.

[0098] As an example, the soft objective function may be:

[0099] softloss=Σ[y z -y f ]+Σ[y m -y k [.

[0100] Among them, y z It can be a second predicted flow amount in the above second predicted flow amount set. f It can be a target flow volume in the above target flow volume set. k It can be the target inventory level mentioned above. m It can be the second predicted inventory. The second predicted turnover in the second predicted turnover set corresponds to the target turnover in the target turnover set. It should be noted that the soft objective function loss value is determined based on the target inventory and target turnover set predicted by the first replenishment information generation model, which can improve the accuracy of the second initial model predicting the inventory and turnover set.

[0101] The second step is to determine the second characteristic loss value according to the target historical inventory of the historical circulation information set, the second predicted inventory, the second predicted circulation set and the historical circulation set corresponding to the second predicted circulation set. Among them, the historical circulation set corresponding to the first predicted circulation set is consistent with the historical circulation set corresponding to the second predicted circulation set. In practice, first, the execution subject can input the second predicted inventory, the second predicted circulation set, the historical circulation set corresponding to the second predicted circulation set and the target historical inventory into the hard objective function.

[0102] As an example, the hard objective function may be:

[0103] hardloss=∑[y z -y t ]+∑[y m -y a ].

[0104] Among them, y t It can be a historical flow volume in the historical flow volume set.

[0105] The third step is to generate a joint feature loss value of the above historical flow information set based on the first feature loss value and the second feature loss value. In practice, first, the above execution subject can determine the preset soft objective function weight and the preset hard objective function weight. The hard objective function, the soft objective function, the preset soft objective function weight and the preset hard objective function weight are input into the target loss function to obtain the loss value.

[0106] As an example, the above objective loss function can be:

[0107] loss=α×hardloss+(1-α)×softloss.

[0108] Wherein, α can be a preset hard objective function weight.

[0109] The fourth step is to determine whether the second initial model achieves the preset optimization goal based on the joint feature loss value.

[0110] Optionally, after determining whether the second initial model reaches the preset optimization target based on the second predicted turnover volume set, the second predicted inventory quantity and the target historical inventory quantity corresponding to the historical turnover information set, the execution entity may also adjust the network parameters of the second initial model in response to determining that the second initial model does not reach the preset optimization target, use the adjusted second initial model as the second initial model, and execute the training step again.

[0111] from Figure 3 It can be seen that Figure 2 Compared with the description of some corresponding embodiments, Figure 3 The process 300 of the replenishment information generation method in some corresponding embodiments reflects the training process of the replenishment information generation model, and reflects that the second initial model is trained according to the first replenishment information generation model to obtain the replenishment information generation model. It reflects the structure and training steps of the model, and the distillation through the soft objective function. And the prediction of the value and flow of the encoding layer. And it reflects the prediction of the value and flow of the target item, and predicts the inventory and turnover according to the predicted value and flow, so as to improve the accuracy of generating replenishment information.

[0112] Further references Figure 4 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a replenishment information generating device. These device embodiments are similar to Figure 2 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0113] like Figure 4 As shown, the replenishment information generating device 400 of some embodiments includes: a first input unit 401, a second input unit 402 and a generating unit 403. The first input unit 401 is configured to input the object circulation data information of the target object into the encoding layer of the pre-trained replenishment information generating model to obtain the value encoding information, the flow encoding information and the circulation volume encoding information; the second input unit 402 is configured to input the above-mentioned flow encoding information, the above-mentioned value encoding information and the above-mentioned circulation volume encoding information into the decoding layer of the above-mentioned replenishment information generating model to obtain the predicted inventory quantity and the predicted circulation volume set; the generating unit 403 is configured to generate the replenishment information corresponding to the above-mentioned target object according to the above-mentioned predicted inventory quantity and the above-mentioned predicted circulation volume set.

[0114] Optionally, the first input unit 401 is further configured as follows: the replenishment information generation model is trained through the following steps: training a first replenishment information generation model based on the historical circulation information set of the target item; determining a target inventory quantity and a target circulation quantity set corresponding to the first replenishment information generation model; and training a replenishment information generation model based on the historical circulation information set, the target inventory quantity and the target circulation quantity set.

[0115] Optionally, the first input unit 401 is further configured to: train the first replenishment information generation model based on the historical circulation information set of the target item, including: performing the following target training steps based on the historical circulation information set: determining the historical value information, historical flow information and target historical inventory corresponding to the historical circulation information set; inputting the historical circulation information set into the encoding layer of the first initial model to obtain first value coding information, first flow coding information and first circulation volume coding information; inputting the obtained first value coding information, first flow coding information, first circulation volume coding information, historical value information and historical flow information into the decoding layer of the first initial model to obtain the first predicted inventory and the first predicted circulation volume set; determining whether the first initial model reaches the preset optimization target based on the first predicted inventory, the first predicted circulation volume set and the target historical inventory; in response to determining that the first initial model reaches the preset optimization target, determining the first initial model as the target replenishment information model.

[0116] Optionally, the first input unit 401 is further configured to: determine whether the first initial model reaches a preset optimization target based on the first predicted inventory, the first predicted turnover volume set and the target historical inventory, including: generating a loss value corresponding to the historical circulation information set based on the first predicted inventory, the first predicted turnover volume set, the target historical inventory and the historical turnover volume set corresponding to the first predicted turnover volume set; determine whether the first initial model reaches the preset optimization target based on the loss value corresponding to the historical circulation information set.

[0117] Optionally, the first input unit 401 is further configured to: after determining whether the first initial model reaches the preset optimization target based on the first predicted inventory, the first predicted turnover volume set and the target historical inventory, the method further includes: in response to determining that the first initial model does not reach the preset optimization target, adjusting the network parameters of the first initial model, using the adjusted first initial model as the first initial model, and performing the target training step again.

[0118] Optionally, the first input unit 401 is further configured to: perform the following training steps according to the historical circulation information set: input the historical circulation information set into the encoding layer of the second initial model to obtain second value coding information, second flow coding information and second circulation volume coding information; input the obtained second value coding information, second flow coding information and second circulation volume coding information into the decoding layer of the second initial model to obtain a second predicted circulation volume set and a second predicted inventory quantity corresponding to the historical circulation information set; determine whether the second initial model reaches the preset optimization target according to the second predicted circulation volume set, the second predicted inventory quantity and the target historical inventory quantity corresponding to the historical circulation information set; in response to determining that the second initial model reaches the preset optimization target, determine the second initial model as a replenishment information model.

[0119] Optionally, the first input unit 401 is further configured to: determine whether the second initial model reaches the preset optimization target based on the second predicted turnover volume set, the second predicted inventory and the target historical inventory corresponding to the historical turnover information set, including: generating a first characteristic loss value based on the second predicted turnover volume set, the second predicted inventory, the target inventory and the target turnover volume set; determining a second characteristic loss value based on the target historical inventory of the historical turnover information set, the second predicted inventory, the second predicted turnover volume set and the historical turnover volume set corresponding to the second predicted turnover volume set; generating a joint characteristic loss value of the historical turnover information set based on the first characteristic loss value and the second characteristic loss value; and determining whether the second initial model reaches the preset optimization target based on the joint characteristic loss value.

[0120] Optionally, the first input unit 401 is further configured to: in response to determining that the second initial model does not reach the preset optimization target, adjust the network parameters of the second initial model, use the adjusted second initial model as the second initial model, and perform the training step again.

[0121] Optionally, the generation unit 403 is further configured to: determine the inventory quantity corresponding to the target item from the item circulation data information; determine the replenishment inventory quantity corresponding to the target item based on the predicted inventory quantity and the inventory quantity; and determine the replenishment inventory quantity and the predicted circulation quantity set as the replenishment information corresponding to the target item.

[0122] Optionally, the replenishment information generating device also includes: an acquisition unit, configured to acquire inventory location information corresponding to the above-mentioned target item; a first control unit, configured to control the associated robotic arm to grab each item corresponding to the above-mentioned replenishment information based on the above-mentioned inventory location information and the above-mentioned replenishment information; and a second control unit, configured to control the associated robotic arm to perform item scheduling operations on each grabbed item.

[0123] It is understood that the units described in the device 400 are similar to those described in the reference Figure 2 Therefore, the operations, features and beneficial effects described above for the method are also applicable to the device 400 and the units included therein, and will not be described in detail here.

[0124] Reference below Figure 5 , which shows an electronic device 500 (eg, Figure 1 A schematic diagram of the structure of the computing device in FIG. Figure 5 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0125] like Figure 5 As shown, the electronic device 500 may include a processing device 501 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage device 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0126] Typically, the following devices may be connected to the I / O interface 505: an input device 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; and a communication device 509. The communication device 509 may allow the electronic device 500 to communicate with other devices wirelessly or by wire to exchange data. Figure 5 The electronic device 500 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 5 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0127] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from the network through the communication device 509, or installed from the storage device 508, or installed from the ROM 502. When the computer program is executed by the processing device 501, the above-mentioned functions defined in the method of some embodiments of the present disclosure are executed.

[0128] It should be noted that the computer-readable medium recorded 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 above two. The 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 of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In some embodiments of the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0129] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0130] The computer-readable medium may be included in the electronic device; or it may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: inputs the object circulation data information of the target object into the encoding layer of the pre-trained replenishment information generation model to obtain value coding information, flow coding information and circulation volume coding information; inputs the flow coding information, the value coding information and the circulation volume coding information into the decoding layer of the replenishment information generation model to obtain the predicted inventory and the predicted circulation volume set; generates the replenishment information corresponding to the target object according to the predicted inventory and the predicted circulation volume set.

[0131] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0132] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0133] The units described in some embodiments of the present disclosure may be implemented by software or by hardware. The described units may also be provided in a processor, for example, may be described as: a processor comprising a first input unit, a second input unit, and a generation unit. The names of these units do not, in some cases, constitute limitations on the units themselves, for example, the generation unit may also be described as "a unit for generating replenishment information corresponding to the above-mentioned target item based on the above-mentioned predicted inventory quantity and the above-mentioned predicted turnover quantity set".

[0134] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0135] Some embodiments of the present disclosure further provide a computer program product, including a computer program, which implements any of the above-mentioned replenishment information generation methods when executed by a processor.

[0136] The above descriptions are only some preferred embodiments of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) and the technical solutions formed.

Claims

1. A method for generating replenishment information, comprising: Inputting the object circulation data information of the target object into the encoding layer of the pre-trained replenishment information generation model to obtain value encoding information, flow encoding information and circulation volume encoding information; Input the flow coding information, the value coding information and the circulation volume coding information into the decoding layer of the replenishment information generation model to obtain a predicted inventory quantity and a predicted circulation volume set; Generate replenishment information corresponding to the target item based on the predicted inventory quantity and the predicted turnover quantity set.

2. The method according to claim 1, wherein: The replenishment information generation model is trained by the following steps: According to the historical circulation information set of the target item, a first replenishment information generation model is trained; Determine a target inventory quantity and a target turnover quantity set corresponding to the first replenishment information generation model; A replenishment information generation model is trained based on the historical circulation information set, the target inventory quantity and the target circulation quantity set.

3. The method according to claim 2, wherein: The training of a first replenishment information generation model based on the historical circulation information set of the target item includes: According to the historical flow information set, the following target training steps are performed: Determine the historical value information, historical flow information and target historical inventory corresponding to the historical circulation information set; Input the historical circulation information set into the coding layer of the first initial model to obtain first value coding information, first flow coding information and first circulation volume coding information; Input the obtained first value coding information, first flow coding information, first circulation volume coding information, historical value information and historical flow information into the decoding layer of the first initial model to obtain a first predicted inventory quantity and a first predicted circulation volume set; Determining whether the first initial model achieves a preset optimization goal according to the first predicted inventory quantity, the first predicted turnover quantity set, and the target historical inventory quantity; In response to determining that the first initial model achieves the preset optimization target, the first initial model is determined as the target replenishment information model.

4. The method according to claim 3, wherein: The determining whether the first initial model reaches a preset optimization target according to the first predicted inventory quantity, the first predicted turnover quantity set and the target historical inventory quantity includes: Generate a loss value corresponding to the historical circulation information set according to the first predicted inventory quantity, the first predicted circulation quantity set, the target historical inventory quantity and the historical circulation quantity set corresponding to the first predicted circulation quantity set; According to the loss value corresponding to the historical flow information set, it is determined whether the first initial model reaches the preset optimization goal.

5. The method according to claim 4, wherein: After determining whether the first initial model reaches a preset optimization target according to the first predicted inventory quantity, the first predicted turnover quantity set and the target historical inventory quantity, the method further includes: In response to determining that the first initial model does not reach the preset optimization target, the network parameters of the first initial model are adjusted, the adjusted first initial model is used as the first initial model, and the target training step is performed again.

6. The method according to claim 3, wherein: The training of a replenishment information generation model according to the historical circulation information set, the target inventory quantity and the target circulation quantity set includes: According to the historical flow information set, the following training steps are performed: Inputting the historical circulation information set into the coding layer of the second initial model to obtain second value coding information, second flow coding information and second circulation volume coding information; Input the obtained second value coding information, second flow coding information and second circulation quantity coding information into the decoding layer of the second initial model to obtain a second predicted circulation quantity set and a second predicted inventory quantity corresponding to the historical circulation information set; Determining whether the second initial model achieves the preset optimization goal according to the second predicted turnover volume set, the second predicted inventory volume, and the target historical inventory volume corresponding to the historical turnover information set; In response to determining that the second initial model achieves the preset optimization target, the second initial model is determined as the replenishment information model.

7. The method according to claim 6, wherein: The determining whether the second initial model achieves the preset optimization goal according to the second predicted turnover volume set, the second predicted inventory volume, and the target historical inventory volume corresponding to the historical turnover information set includes: Generate a first characteristic loss value according to the second predicted turnover volume set, the second predicted inventory volume, the target inventory volume and the target turnover volume set; Determine a second characteristic loss value according to the target historical inventory of the historical circulation information set, the second predicted inventory, the second predicted circulation volume set, and the historical circulation volume set corresponding to the second predicted circulation volume set; Generating a joint feature loss value of the historical flow information set according to the first feature loss value and the second feature loss value; According to the joint feature loss value, determine whether the second initial model reaches the preset optimization target.

8. The method according to claim 6, wherein: After determining whether the second initial model reaches the preset optimization target according to the second predicted turnover volume set, the second predicted inventory volume, and the target historical inventory volume corresponding to the historical turnover information set, the method further includes: In response to determining that the second initial model does not reach the preset optimization target, the network parameters of the second initial model are adjusted, the adjusted second initial model is used as the second initial model, and the training step is performed again.

9. The method according to claim 1, wherein: The generating, according to the predicted inventory quantity and the predicted turnover quantity set, corresponding replenishment information of the target item includes: Determine the inventory quantity corresponding to the target item from the item circulation data information; Determining a replenishment inventory quantity corresponding to the target item according to the predicted inventory quantity and the inventory quantity; The replenishment inventory quantity and the predicted turnover quantity set are determined as replenishment information corresponding to the target item.

10. The method according to claim 9, wherein: After generating replenishment information corresponding to the target item according to the predicted inventory quantity and the predicted turnover quantity set, the method further includes: Obtaining inventory location information corresponding to the target item; According to the inventory location information and the replenishment information, controlling the associated robotic arm to grab each item corresponding to the target item; Control the associated robotic arms to perform item scheduling operations on each grabbed item.

11. A replenishment information generating device, comprising: A first input unit is configured to input the object circulation data information of the target object into the encoding layer of the pre-trained replenishment information generation model to obtain value encoding information, flow encoding information and circulation volume encoding information; A second input unit is configured to input the flow coding information, the value coding information and the circulation coding information into a decoding layer of the replenishment information generation model to obtain a predicted inventory quantity and a predicted circulation quantity set; The generating unit is configured to generate replenishment information corresponding to the target item according to the predicted inventory quantity and the predicted turnover quantity set.

12. An electronic device comprising: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 10.

13. A computer readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method according to any one of claims 1 to 10 is implemented.

14. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the method according to any one of claims 1 to 10.

Images