A method and device for processing an order, an electronic device and a computer storage medium

Abstract

The specification discloses a method and device for processing an order, electronic equipment and computer storage medium. According to the order characteristics of the current order obtained and the capacity characteristics of each type of delivery capacity, for each type of delivery capacity, the delivery time of the current order delivered by the delivery capacity of the type is obtained by the time estimation model output, and the delivery probability of the current order delivered by the delivery capacity of the type is obtained by the type estimation model output. Then, according to the delivery time and the delivery probability obtained for each type of delivery capacity, the estimated delivery time of the current order is determined to process the current order. Through the type estimation model, the delivery probability of each type of delivery capacity for delivering the current order is estimated, considering the case that the initial delivery capacity type for processing the order is inconsistent with the final delivery capacity type for processing the order, solving the problem that the estimated delivery time is inaccurate, resulting in the inability to correctly process the order, and improving the accuracy of order processing.

Description

Technical Field

[0001] This specification relates to the field of computer technology, and in particular to a method, apparatus, electronic device, and computer storage medium for processing orders. Background Technology

[0002] With the development of online ordering and offline delivery services (such as food delivery), to facilitate order processing by servers (such as food delivery platforms) and delivery capacity (such as delivery riders) and improve user experience, servers estimate the Estimated Time of Arrival (ETA) after a user places an order and process the order based on the estimated ETA. The estimated ETA not only provides a basis for order processing for servers and delivery capacity but also provides users with a reference time for order delivery, thus improving the user experience to some extent.

[0003] In existing technologies, servers typically determine the type of delivery capacity based on order characteristics, determine the ETA (Earning Time Amount) using a machine learning model corresponding to the delivery capacity type, and then process the order based on its ETA. However, due to complex real-world situations (such as heavy delivery capacity load), delivery capacity may be unable to process orders on time. In such cases, a common approach is to change the delivery capacity type; that is, under certain circumstances, orders that should have been processed by type A delivery capacity are instead processed by type B delivery capacity.

[0004] However, because this processing method changes the delivery capacity type, that is, the delivery capacity type of the initial order processing is inconsistent with the delivery capacity type of the final order processing, the ETA determined by the initial delivery capacity type differs greatly from the actual delivery time, thus making it impossible to process the order correctly based on the incorrect ETA. Summary of the Invention

[0005] This specification provides a method, apparatus, electronic device, and computer storage medium for processing orders, in order to partially solve the aforementioned problems existing in the prior art.

[0006] The embodiments in this specification adopt the following technical solutions:

[0007] This manual provides a method for processing orders, including:

[0008] Obtain the order characteristics of the current order and the capacity characteristics of each type of delivery capacity;

[0009] The order characteristics and the capacity characteristics of each type of delivery capacity are used as input features, and the input features are respectively input into the pre-trained time prediction model and the pre-trained type prediction model;

[0010] For each type of delivery capacity, the delivery time of the current order by that type of delivery capacity is obtained from the time estimation model, and the delivery probability of the current order by that type of delivery capacity is obtained from the type estimation model.

[0011] The estimated delivery time for the current order is determined based on the delivery time and delivery probability obtained for each type of delivery capacity.

[0012] The current order is processed according to the estimated delivery time.

[0013] Optionally, the input features are input into a pre-trained time prediction model, specifically including:

[0014] For each type of delivery capacity, the order characteristics and the capacity characteristics of that type of delivery capacity are input into a pre-trained time prediction model corresponding to that type of delivery capacity.

[0015] Optionally, the estimated delivery time of the current order is determined based on the delivery time and delivery probability obtained for each type of delivery capacity, specifically including:

[0016] For each type of delivery capacity, the delivery time of the current order is weighted by the probability of delivery by that type of delivery capacity.

[0017] The estimated delivery time for the current order is determined based on the weighted result of each type of delivery capacity.

[0018] Optionally, a time prediction model is pre-trained, specifically including:

[0019] Pre-acquire historical order characteristics and capacity characteristics of historical delivery capacity for each type; among the various types of historical delivery capacity, determine the type of historical delivery capacity that actually delivered the historical orders in the past, and designate it as the specified type;

[0020] Using the historical order characteristics and the capacity characteristics of each type of historical delivery capacity as samples, input them into the time prediction model to be trained, and obtain the optimized delivery time of the historical order to be delivered by the historical delivery capacity of that type, as output by the time prediction model to be trained for each type of historical delivery capacity.

[0021] The time prediction model is trained with the goal of minimizing the difference between the expected delivery time of the historical order delivered by the specified type of historical delivery capacity and the historical delivery time of the historical order.

[0022] Optionally, a type prediction model is pre-trained, specifically including:

[0023] The historical order characteristics and capacity characteristics of each type of historical delivery capacity are obtained in advance; and, among each type of historical delivery capacity, the type of historical delivery capacity that actually delivered the historical order is determined as the designated type, and the types of other historical delivery capacity besides the designated type are designated as non-designated types;

[0024] Using the historical order features and the capacity features of each type of historical delivery capacity as samples, input them into the type prediction model to be trained, and obtain the optimized delivery probability of the historical order being delivered by the historical delivery capacity of that type of historical delivery capacity output by the type prediction model to be trained for each type of historical delivery capacity.

[0025] The type prediction model is trained with the training objective of maximizing the probability of optimizing the delivery of the historical order by the specified type of historical delivery capacity and minimizing the probability of optimizing the delivery of the historical order by the non-specified type of historical delivery capacity.

[0026] Optionally, the time prediction model and the type prediction model are pre-trained, specifically including:

[0027] Pre-acquire historical order characteristics and capacity characteristics of historical delivery capacity for each type of order;

[0028] Using the historical order features and the capacity features of each type of historical delivery capacity as samples, input them into the time prediction model and the type prediction model to be trained, respectively.

[0029] For each type of historical delivery capacity, obtain the optimized delivery time of the historical order delivered by the delivery capacity of that type, output by the time prediction model to be trained, and obtain the optimized delivery probability of the historical order delivered by the historical delivery capacity of that type, output by the type prediction model to be trained.

[0030] Based on the delivery time and delivery probability to be optimized obtained for each type of historical delivery capacity, the estimated delivery time to be optimized for the historical orders is determined.

[0031] The time prediction model and the type prediction model are trained with the goal of minimizing the difference between the estimated delivery time to be optimized of the historical orders and the historical delivery time of the historical orders.

[0032] Optionally, before training the time prediction model and the type prediction model with the training objective of minimizing the difference between the estimated delivery time to be optimized of the historical orders and the historical delivery time of the historical orders, the method further includes:

[0033] The prediction accuracy of the type prediction model to be trained is determined to be no less than a set threshold.

[0034] This specification provides an apparatus for processing orders, comprising:

[0035] The acquisition module is used to acquire the order characteristics of the current order and the capacity characteristics of each type of delivery capacity;

[0036] The input module is used to take the order features and the capacity features of each type of delivery capacity as input features, and input the input features into the pre-trained time prediction model and the pre-trained type prediction model respectively;

[0037] The estimation module is used to obtain, for each type of delivery capacity, the delivery time of the current order by the delivery capacity of that type as output by the time estimation model, and the delivery probability of the current order by the delivery capacity of that type as output by the type estimation model.

[0038] The estimated delivery time determination module is used to determine the estimated delivery time of the current order based on the delivery time and delivery probability obtained for each type of delivery capacity.

[0039] The order processing module is used to process the current order according to the estimated delivery time.

[0040] This specification provides an electronic device for processing orders, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method for processing orders.

[0041] This specification provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for processing orders.

[0042] The above-described at least one technical solution adopted in the embodiments of this specification can achieve the following beneficial effects:

[0043] This method introduces a type prediction model, which estimates the probability of delivering the current order using each type of delivery capacity. Combined with the delivery time for that type of delivery capacity, output by a time prediction model, the estimated delivery time for the current order is obtained. This method addresses the issue of inaccurate estimated delivery times leading to incorrect order processing, even when the final delivery capacity type for the order is unknown. By predicting the probability of delivery for each type of delivery capacity, it considers the possibility of discrepancies between the initial and final delivery capacity types, thus resolving the problem of inaccurate estimated delivery times causing incorrect order processing and improving the accuracy of order processing. Attached Figure Description

[0044] The accompanying drawings, which are included to provide a further understanding of this specification and form part of this specification, illustrate exemplary embodiments and are used to explain this specification, but do not constitute an undue limitation thereof. In the drawings:

[0045] Figure 1 A flowchart illustrating an order processing method provided in an embodiment of this specification;

[0046] Figure 2 A schematic diagram of an order processing device provided in the embodiments of this specification;

[0047] Figure 3 The embodiments provided in this specification correspond to Figure 1 A schematic diagram of an electronic device. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this specification clearer, the technical solutions of this specification will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of them. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0049] The technical solutions provided in the various embodiments of this specification are described in detail below with reference to the accompanying drawings.

[0050] Figure 1 This is a flowchart illustrating an order processing method provided in an embodiment of this specification, which specifically includes the following steps:

[0051] S100: Obtain the order characteristics of the current order and the capacity characteristics of each type of delivery capacity.

[0052] In this specification, the device that performs order processing operations can be a server of a delivery platform or a terminal that initiates an order. For ease of understanding, this specification uses a server as an example to illustrate the specific technical solutions in the embodiments described.

[0053] In practical applications, after receiving an order from a user, the server needs to predict the estimated delivery time of the order. This prediction serves as a commitment to timeliness and a basis for order processing, allowing the server to determine the appropriate processing method based on the estimated delivery time. Therefore, the predicted estimated delivery time significantly impacts the accuracy of subsequent order processing.

[0054] Since the final delivery capacity for this order has not yet been determined, the server will typically obtain the order characteristics of the current order and the capacity characteristics of each type of delivery capacity based on the order information.

[0055] Specifically, after the server receives a new order, it determines the order characteristics of the current order based on the order information. The order information may include user geographic location information, merchant geographic location information, delivery distance information, delivery item information, and environmental information. The order characteristics may include any one or more of the following: delivery item characteristics, user characteristics, merchant characteristics, and environmental characteristics. Delivery item characteristics may include the type and weight of the delivered item; user characteristics may include the user's delivery location; merchant characteristics may include the merchant's geographic location and food preparation time; environmental characteristics may include peak hour characteristics, weekday / holiday characteristics, seasonal characteristics, and weather characteristics, etc.

[0056] The delivery capacity types described in the embodiments of this specification can be manual delivery, such as delivery personnel and merchants, or unmanned delivery equipment. Manual delivery capacity can include dedicated delivery capacity, express delivery capacity, etc. This embodiment does not limit the specific type and quantity of delivery capacity.

[0057] Specifically, the characteristics of dedicated delivery and express delivery services differ significantly. In terms of load capacity, express delivery outsources orders to part-time delivery personnel on a voluntary basis, while dedicated delivery assigns orders to full-time delivery personnel. Therefore, dedicated delivery may have a heavier load than express delivery when delivering a particular order. Regarding delivery range, dedicated delivery has a radius of approximately 3 kilometers from the delivery person, while express delivery has a radius of approximately 5 kilometers. Thus, dedicated delivery has a smaller delivery range compared to express delivery. Finally, in terms of actual delivery time, dedicated delivery orders are delivered relatively faster than express delivery orders.

[0058] Specifically, the capacity information of delivery vehicles can include at least the information of the delivery vehicles themselves. The information of the delivery vehicles themselves is characterized by the capacity characteristics of the delivery vehicles, which include any one or more of the following: delivery capacity, load status, and delivery range.

[0059] Since the order processing method provided in this specification requires inputting the order features and capacity features of each type of delivery capacity into a machine learning model, it is necessary to encode the order features and capacity features of each type of delivery capacity to obtain an order feature vector corresponding to the order features and a capacity feature vector corresponding to the capacity features of each type of delivery capacity. These order feature vectors and capacity feature vectors corresponding to the capacity features of each type of delivery capacity are then used as input to the model. This specification does not limit the encoding method.

[0060] S102: Take the order features and the capacity features of each type of delivery capacity as input features, and input the input features into the pre-trained time prediction model and the pre-trained type prediction model respectively.

[0061] In practical applications, for each type of delivery capacity, there can be one or more pre-trained time prediction models, and each of the multiple time prediction models can correspond to a different type of delivery capacity.

[0062] In the case of using a pre-trained time estimation model: the time estimation model can obtain the delivery time of the current order for each type of delivery capacity by training on a sample containing the capacity characteristics of each type of delivery capacity, and then obtain the estimated delivery time of the current order based on the delivery time.

[0063] For the case of using multiple pre-trained time estimation models corresponding to each type of delivery capacity: For each type of delivery capacity, the input features are input into the pre-trained time estimation model corresponding to that type of delivery capacity to obtain the delivery time of the current order for that type of delivery capacity. Combined with the delivery probability of the current order for that type of delivery capacity output by the type estimation model, the estimated delivery time of the current order is obtained.

[0064] In the embodiments of this specification, for ease of understanding, only the example of multiple pre-trained time estimation models corresponding to each type of delivery capacity is used to illustrate the specific technical solutions.

[0065] S104: For each type of delivery capacity, obtain the delivery time of the current order by the delivery capacity of that type as output by the time estimation model, and obtain the delivery probability of the current order by the delivery capacity of that type as output by the type estimation model.

[0066] Specifically, for each type of delivery capacity, the delivery time for the current order by that type of delivery capacity means: if the delivery capacity of that type is the actual delivery capacity for the order, then the delivery time is the actual time required to deliver the order; the delivery probability of the current order being delivered by that type of delivery capacity means: the actual delivery probability of the current order being delivered by that type of delivery capacity.

[0067] S106: Determine the estimated delivery time of the current order based on the delivery time and delivery probability obtained for each type of delivery capacity.

[0068] Optionally, for each type of delivery capacity, the delivery time of the current order delivered by that type of delivery capacity is weighted by the probability of delivery of the current order by that type of delivery capacity. Then, based on the weighted result of each type of delivery capacity, the estimated delivery time of the current order is determined. For example, there are two types of delivery capacity that can be matched for the current order: type A and type B. For these two types of delivery capacity, the delivery time t of the current order delivered by type A delivery capacity, output by the time estimation model, is obtained. A The delivery time t of the current order is determined by delivery capacity of type B. B ; Obtain the delivery probability p of the current order delivered by the delivery capacity of type A, as output by the prediction model of the aforementioned type. A The delivery probability p of the current order being delivered by delivery capacity of type B. B Based on the delivery time and delivery probability obtained for each type of delivery capacity, the estimated delivery time T = p for the current order is determined. A ×t A +p B ×t B .

[0069] S108: Process the current order according to the estimated delivery time.

[0070] In practical applications, the estimated delivery time serves to allow the server to schedule the order based on the estimated delivery time, such as prioritizing delivery, delaying delivery, or transferring the order. The estimated delivery time can also be sent to the terminal that initiated the order and displayed there, as a promise of punctuality to the user.

[0071] This method introduces a type prediction model, which estimates the probability of delivering the current order using each type of delivery capacity. Combined with the delivery time for that type of delivery capacity, output by a time prediction model, the estimated delivery time for the current order is obtained. This method addresses the issue of inaccurate estimated delivery times leading to incorrect order processing, even when the final delivery capacity type for the order is unknown. By predicting the probability of delivery for each type of delivery capacity, it considers the possibility of discrepancies between the initial and final delivery capacity types, thus resolving the problem of inaccurate estimated delivery times causing incorrect order processing and improving the accuracy of order processing.

[0072] In the embodiments described in this specification, such as Figure 1 The time prediction model and type prediction model shown in step S102 can be trained separately or jointly.

[0073] In the embodiments described in this specification, such as Figure 1 The time prediction model shown in step S102 can be trained separately using the following methods.

[0074] Specifically, it is necessary to obtain the historical order characteristics of historical orders and the capacity characteristics of historical delivery capacity for each type in advance.

[0075] For each historical order delivery record, the historical order characteristics corresponding to that historical order delivery record are obtained. The method for obtaining historical order characteristics is the same as the method for obtaining current order characteristics described above, and will not be repeated here. At the same time, the capacity characteristics of historical delivery capacity for each type are obtained. The method for obtaining historical delivery capacity capacity characteristics is the same as the method for obtaining delivery capacity capacity characteristics described above, and will not be repeated here.

[0076] For each historical order delivery record, the historical order characteristics and historical delivery capacity characteristics obtained above are used as input features generated based on the historical order delivery record, and the input features generated separately for each historical order delivery record are determined.

[0077] For each type of historical delivery capacity, the type of historical delivery capacity that actually delivered that historical order is designated as the specified type, and other types of historical delivery capacity are designated as unspecified types. In other words, for each historical order delivery record, each type of historical delivery capacity can be divided into specified type historical delivery capacity and unspecified type historical delivery capacity. The specified type of historical delivery capacity refers to the delivery capacity that actually delivered that historical order in the past.

[0078] Using the historical order characteristics and the capacity characteristics of each type of historical delivery capacity as samples, input them into the time prediction model to be trained, and obtain the optimized delivery time of the historical order to be delivered by the historical delivery capacity of that type, as output by the time prediction model to be trained for each type of historical delivery capacity.

[0079] The time prediction model is trained with the goal of minimizing the difference between the expected delivery time of the historical order delivered by the specified type of historical delivery capacity and the historical delivery time of the historical order.

[0080] In practical applications, the difference between the delivery time to be optimized for the historical order and the historical delivery time of the historical order is evaluated using a loss function. The time prediction model is trained using gradient descent and backpropagation algorithms, with the minimization of this loss function value as the training objective. The loss function can include: 0-1 loss function, absolute value loss function, logarithmic loss function, exponential loss function, and Hinge loss function.

[0081] In the embodiments described in this specification, such as Figure 1 The type prediction model shown in step S102 can be trained separately using the following methods.

[0082] The system pre-acquires historical order characteristics and capacity characteristics for each type of historical delivery capacity. Furthermore, it categorizes historical delivery capacity into designated and undesignated types: the type of historical delivery capacity that actually delivered the historical order is designated, and other types of historical delivery capacity are designated as undesignated.

[0083] Using the historical order features and the capacity features of each type of historical delivery capacity as samples, input them into the type prediction model to be trained, and obtain the optimized delivery probability of the historical order to be delivered by the historical delivery capacity of that type, output by the type prediction model to be trained for each type of historical delivery capacity.

[0084] The type prediction model is trained with the training objective of maximizing the probability of optimizing the delivery of the historical order by the specified type of historical delivery capacity and minimizing the probability of optimizing the delivery of the historical order by the non-specified type of historical delivery capacity.

[0085] For example, based on a historical order, the historical delivery capacity of a specified type is designated as type A historical delivery capacity, and the historical delivery capacity of a non-specified type is designated as type B historical delivery capacity. This means that historically, the order was actually delivered by type A historical delivery capacity. Inputting the historical order features and the capacity features of both types A and B historical delivery capacity into the type prediction model to be trained yields a probability of 0.6 for the historical order delivered by type A historical delivery capacity and a probability of 0.4 for the historical order delivered by type B historical delivery capacity. The training objective of the type prediction model is to make the probability of the historical order delivered by type A historical delivery capacity closer to 1 and the probability of the historical order delivered by type B historical delivery capacity closer to 0.

[0086] In the embodiments described in this specification, such as Figure 1 The time prediction model and type prediction model shown in step S102 can also be jointly trained using the following methods.

[0087] Historical order features and delivery capacity features for each type of historical order are obtained in advance and used as samples. These samples are then input into the time prediction model and the type prediction model to be trained, respectively.

[0088] For each type of historical delivery capacity, obtain the optimized delivery time of the historical order delivered by the historical delivery capacity of that type, output by the time prediction model to be trained, and obtain the optimized delivery probability of the historical order delivered by the historical delivery capacity of that type, output by the type prediction model to be trained.

[0089] The estimated delivery time to be optimized for the historical orders is determined based on the estimated delivery time and probability to be optimized for each type of historical delivery capacity.

[0090] In practical applications, for each type of historical delivery capacity, the estimated delivery time of the historical order to be delivered by that type of historical delivery capacity needs to be weighted, using the probability of that type of historical delivery capacity delivering the historical order as the weight. Then, based on the weighted result of each type of historical delivery capacity, the estimated estimated delivery time of the historical order to be optimized is determined.

[0091] The time prediction model and the type prediction model are trained with the goal of minimizing the difference between the estimated delivery time to be optimized and the historical delivery time of the historical orders. The evaluation method for the difference between the estimated delivery time to be optimized and the historical delivery time of the historical orders is the same as the evaluation method used when training the time prediction model separately, and will not be repeated here.

[0092] Before training the time prediction model and the type prediction model with the training objective of minimizing the difference between the expected delivery time to be optimized of the historical order and the historical delivery time of the historical order, the method further includes determining that the prediction accuracy of the type prediction model to be trained is not lower than a set threshold.

[0093] In practical applications, the model to be trained needs quantifiable model prediction accuracy metrics to evaluate its predictive performance. Model prediction accuracy may include, but is not limited to, at least one of the following: accuracy, precision, recall, F1 score, and area under the ROC curve (AUC).

[0094] Determining the prediction accuracy of the type prediction model to be trained specifically includes:

[0095] For each type of historical delivery capacity, the probability of optimizing delivery for that historical order using that type of historical delivery capacity is obtained. Based on the geographical range of each historical order, the determined probability of optimizing delivery for each historical order is grouped by geographical range. For each geographical range of a historical order, the prediction accuracy (AUC) of the type prediction model to be trained for that geographical range is obtained. Furthermore, using the order volume of that geographical range as weight, a weighted average is calculated based on the AUC of that geographical range. The prediction accuracy (Group AUC, GAUC) of the type prediction model to be trained based on geographical grouping is determined according to the weighted average result.

[0096] Specifically, since the service of delivering orders to users is based on geographical location and has regional attributes, when determining the prediction accuracy of the type prediction model to be trained, it is necessary to first group the orders according to their geographical location in order to improve the accuracy of the prediction.

[0097] Specifically, the threshold value for the type prediction model's prediction accuracy is used to determine the degree of training of the type prediction model. If the prediction accuracy is not lower than the threshold value, it indicates that the training effect of the type prediction model has reached a certain level and it can be jointly trained with the time prediction model. If the prediction accuracy is lower than the threshold value, it indicates that the type prediction model has not been sufficiently trained. If the type prediction model is jointly trained with the time prediction model at this time, it may lead to poor convergence and long training time. Therefore, if the prediction accuracy is lower than the threshold value, the type prediction model still needs to be trained separately. In practical applications, the threshold value for the type prediction model's prediction accuracy is set according to the actual model training requirements. The embodiments in this specification do not limit the method and value of setting this threshold.

[0098] In the embodiments described in this specification, such as Figure 1Step S104, which describes obtaining the delivery probability of the current order by the delivery capacity of that type, as output by the type prediction model, specifically includes:

[0099] Among the various types of delivery capacity, the delivery capacity type capable of delivering the current order is determined based on the order characteristics of the current order and designated as the deliverable type. Other delivery capacity types are designated as non-deliverable types. There can be one or more deliverable types.

[0100] For each type of delivery capacity, obtain the delivery probability of the current order being delivered by that type of delivery capacity, as output by the type prediction model. Specifically, the sum of the delivery probabilities of the current order being delivered by delivery capacity of the deliverable type should be 1, and the delivery probability of the current order being delivered by delivery capacity of the non-deliverable type should all be 0.

[0101] For example, the server determines, based on the merchant's geographic location information in a current order, that the only delivery capacity type matching that merchant's geographic location is dedicated delivery. Therefore, the probability of dedicated delivery for the current order is 1, while the probability of other delivery capacities for the current order is 0.

[0102] The above describes one or more embodiments of the order processing method provided in this specification. Based on the same idea, this specification also provides corresponding order processing apparatus, such as... Figure 2 As shown.

[0103] Figure 2 This specification provides a schematic diagram of the structure of an order processing device, which specifically includes:

[0104] The acquisition module 200 is used to acquire the order characteristics of the current order and the capacity characteristics of each type of delivery capacity;

[0105] Input module 202 is used to take the order features and the capacity features of each type of delivery capacity as input features, and input the input features into the pre-trained time prediction model and the pre-trained type prediction model respectively;

[0106] The estimation module 204 is used to obtain, for each type of delivery capacity, the delivery time of the current order by the delivery capacity of that type output by the time estimation model, and the delivery probability of the current order by the delivery capacity of that type output by the type estimation model;

[0107] The estimated delivery time determination module 206 is used to determine the estimated delivery time of the current order based on the delivery time and delivery probability obtained for each type of delivery capacity.

[0108] The order processing module 208 is used to process the current order according to the estimated delivery time.

[0109] Optionally, the input module 202 is specifically used to input the order characteristics and the capacity characteristics of the delivery capacity of each type into a pre-trained time estimation model corresponding to that type of delivery capacity.

[0110] Optionally, the estimated delivery time determination module 206 is specifically used to, for each type of delivery capacity, weight the delivery time of the current order delivered by that type of delivery capacity as a weight, and determine the estimated delivery time of the current order based on the weighting result of each type of delivery capacity.

[0111] Optionally, the device further includes:

[0112] The time prediction model training module 210 is specifically used to pre-acquire the historical order characteristics of historical orders and the capacity characteristics of historical delivery capacity for each type. Among the various types of historical delivery capacity, it determines the type of historical delivery capacity that actually delivered the historical orders in the past as the specified type. Using the historical order characteristics and the capacity characteristics of each type of historical delivery capacity as samples, it inputs them into the time prediction model to be trained, and obtains the optimized delivery time of the historical orders delivered by the historical delivery capacity of that type for each type of historical delivery capacity output by the time prediction model to be trained. The training objective is to minimize the difference between the optimized delivery time of the historical orders delivered by the specified type of historical delivery capacity and the historical delivery time of the historical orders, and to train the time prediction model.

[0113] Optionally, the device further includes:

[0114] The type prediction model training module 212 is specifically used to pre-acquire the historical order features of historical orders and the capacity features of historical delivery capacity for each type, and, among the historical delivery capacity for each type, determine the type of historical delivery capacity that actually delivered the historical order in history as the specified type, and treat the types of other historical delivery capacity besides the specified type as non-specified types. Using the historical order features and the capacity features of each type of historical delivery capacity as samples, input them into the type prediction model to be trained, and obtain the optimized delivery probability of the historical order delivered by the historical delivery capacity of that type for each type of historical delivery capacity. The training objective is to maximize the optimized delivery probability of the historical order delivered by the specified type of historical delivery capacity and minimize the optimized delivery probability of the historical order delivered by the non-specified type of historical delivery capacity.

[0115] Optionally, the device further includes:

[0116] The joint training module 214 is specifically used to pre-acquire historical order features and capacity features of historical delivery capacity for each type of historical order. Using the historical order features and capacity features of historical delivery capacity for each type of historical order as samples, it is input into the time prediction model and the type prediction model to be trained, respectively. For each type of historical delivery capacity, it obtains the optimized delivery time of the historical order to be delivered by the historical delivery capacity of that type, output by the time prediction model to be trained, and obtains the optimized delivery probability of the historical order to be delivered by the historical delivery capacity of that type, output by the type prediction model to be trained. Based on the optimized delivery time and optimized delivery probability obtained for each type of historical delivery capacity, it determines the optimized estimated delivery time of the historical order. The training objective is to minimize the difference between the optimized estimated delivery time of the historical order and the historical delivery time of the historical order, and train the time prediction model and the type prediction model.

[0117] Optionally, the joint training module 214 is further configured to, before training the time prediction model and the type prediction model with the training objective of minimizing the difference between the expected delivery time to be optimized of the historical order and the historical delivery time of the historical order, determine that the prediction accuracy of the type prediction model to be trained is not lower than a set threshold.

[0118] based on Figure 2 The order processing device shown introduces a type prediction model to estimate the probability of delivery of the current order for each type of delivery capacity. Combined with the delivery time for that type of delivery capacity, output by a time prediction model, the estimated delivery time for the current order is obtained. This method, even without determining the final delivery capacity type of the processed order, estimates the probability of delivery for each type of delivery capacity, considering the possibility of discrepancies between the initial and final delivery capacity types. This solves the problem of inaccurate estimated delivery times leading to incorrect order processing, thus improving the accuracy of order processing.

[0119] This specification also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 The provided order processing methods.

[0120] This instruction manual also provides Figure 3 The diagram shows a schematic structural representation of the electronic device. Figure 3At the hardware level, the order processing device includes a processor, internal bus, network interface, memory, and non-volatile storage, and may also include other hardware required for the business. The processor reads the corresponding computer program from the non-volatile storage into memory and then runs it to achieve the above. Figure 1 The order processing method described herein. Of course, in addition to software implementation, this specification does not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. In other words, the execution subject of the following processing flow is not limited to individual logic units, but can also be hardware or logic devices.

[0121] In the 1990s, improvements to a technology could be clearly distinguished as either hardware improvements (e.g., improvements to the circuit structure of diodes, transistors, switches, etc.) or software improvements (improvements to the methodology). However, with technological advancements, many methodological improvements today can be considered direct improvements to the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved methodology into the hardware circuit. Therefore, it cannot be said that a methodological improvement cannot be implemented using hardware physical modules. For example, a Programmable Logic Device (PLD) (such as a Field Programmable Gate Array (FPGA)) is such an integrated circuit whose logic function is determined by the user programming the device. Designers can program and "integrate" a digital system onto a PLD themselves, without needing chip manufacturers to design and manufacture dedicated integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing integrated circuit chips, this programming is mostly implemented using "logic compiler" software. Similar to the software compiler used in program development, the original code before compilation must be written in a specific programming language, called a Hardware Description Language (HDL). There are many HDLs, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, and RHDL (Ruby Hardware Description Language). Currently, the most commonly used are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should understand that by simply performing some logic programming on the method flow using one of these hardware description languages ​​and programming it into an integrated circuit, the hardware circuit implementing the logical method flow can be easily obtained.

[0122] The controller can be implemented in any suitable manner. For example, it can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicon Labs C8051F320. A memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also recognize that, in addition to implementing the controller in purely computer-readable program code form, the same functionality can be achieved by logically programming the method steps to make the controller take the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the means included therein for implementing various functions can also be considered as structures within the hardware component. Alternatively, the means for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0123] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0124] For ease of description, the above devices are described in terms of function, divided into various units. Of course, in implementing this specification, the functions of each unit can be implemented in one or more software and / or hardware components.

[0125] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0126] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0127] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0128] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0129] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0130] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0131] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0132] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0133] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0134] This specification can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification can also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0135] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0136] The above description is merely an embodiment of this specification and is not intended to limit this specification. Various modifications and variations can be made to this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of the claims of this specification.

Claims

1. A method for processing orders, characterized in that, include: Obtain the order characteristics of the current order and the capacity characteristics of each type of delivery capacity; The order characteristics and the capacity characteristics of each type of delivery capacity are used as input features, and the input features are respectively input into the pre-trained time prediction model and the pre-trained type prediction model; For each type of delivery capacity, the delivery time of the current order by that type of delivery capacity is obtained from the time estimation model, and the delivery probability of the current order by that type of delivery capacity is obtained from the type estimation model. The estimated delivery time for the current order is determined based on the delivery time and delivery probability obtained for each type of delivery capacity. Process the current order according to the estimated delivery time; Specifically, pre-training the type prediction model includes: The historical order characteristics and capacity characteristics of each type of historical delivery capacity are obtained in advance; and, among each type of historical delivery capacity, the type of historical delivery capacity that actually delivered the historical order is determined as the designated type, and the types of other historical delivery capacity besides the designated type are designated as non-designated types; Using the historical order features and the capacity features of each type of historical delivery capacity as samples, input them into the type prediction model to be trained, and obtain the optimized delivery probability of the historical order being delivered by the historical delivery capacity of that type of historical delivery capacity output by the type prediction model to be trained for each type of historical delivery capacity. The type prediction model is trained with the training objective of maximizing the probability of optimizing the delivery of the historical order by the specified type of historical delivery capacity and minimizing the probability of optimizing the delivery of the historical order by the non-specified type of historical delivery capacity.

2. The method as described in claim 1, characterized in that, The input features are respectively input into the pre-trained time prediction model, specifically including: For each type of delivery capacity, the order characteristics and the capacity characteristics of that type of delivery capacity are input into a pre-trained time prediction model corresponding to that type of delivery capacity.

3. The method as described in claim 1, characterized in that, Based on the delivery time and delivery probability obtained for each type of delivery capacity, the estimated delivery time of the current order is determined, specifically including: For each type of delivery capacity, the delivery time of the current order is weighted by the probability of delivery by that type of delivery capacity. The estimated delivery time for the current order is determined based on the weighted result of each type of delivery capacity.

4. The method according to any one of claims 1 to 3, characterized in that, Pre-trained time prediction models include: Pre-acquire historical order characteristics and capacity characteristics of historical delivery capacity for each type; among the various types of historical delivery capacity, determine the type of historical delivery capacity that actually delivered the historical orders in the past, and designate it as the specified type; Using the historical order characteristics and the capacity characteristics of each type of historical delivery capacity as samples, input them into the time prediction model to be trained, and obtain the optimized delivery time of the historical order to be delivered by the historical delivery capacity of that type, as output by the time prediction model to be trained for each type of historical delivery capacity. The time prediction model is trained with the goal of minimizing the difference between the expected delivery time of the historical order delivered by the specified type of historical delivery capacity and the historical delivery time of the historical order.

5. The method according to any one of claims 1 to 3, characterized in that, Pre-trained time prediction models and type prediction models, specifically including: Pre-acquire historical order characteristics and capacity characteristics of historical delivery capacity for each type of order; Using the historical order features and the capacity features of each type of historical delivery capacity as samples, input them into the time prediction model and the type prediction model to be trained, respectively. For each type of historical delivery capacity, obtain the optimized delivery time of the historical order delivered by the historical delivery capacity of that type, output by the time prediction model to be trained, and obtain the optimized delivery probability of the historical order delivered by the historical delivery capacity of that type, output by the type prediction model to be trained. Based on the delivery time and delivery probability to be optimized obtained for each type of historical delivery capacity, the estimated delivery time to be optimized for the historical orders is determined. The time prediction model and the type prediction model are trained with the goal of minimizing the difference between the estimated delivery time to be optimized of the historical orders and the historical delivery time of the historical orders.

6. The method as described in claim 5, characterized in that, Before training the time prediction model and the type prediction model with the training objective of minimizing the difference between the estimated delivery time to be optimized of the historical orders and the historical delivery time of the historical orders, the method further includes: The prediction accuracy of the type prediction model to be trained is determined to be no less than a set threshold.

7. An apparatus for processing orders, characterized in that, include: The acquisition module is used to acquire the order characteristics of the current order and the capacity characteristics of each type of delivery capacity; The input module is used to take the order features and the capacity features of each type of delivery capacity as input features, and input the input features into the pre-trained time prediction model and the pre-trained type prediction model respectively; The estimation module is used to obtain, for each type of delivery capacity, the delivery time of the current order by the delivery capacity of that type as output by the time estimation model, and the delivery probability of the current order by the delivery capacity of that type as output by the type estimation model. The estimated delivery time determination module is used to determine the estimated delivery time of the current order based on the delivery time and delivery probability obtained for each type of delivery capacity. The order processing module is used to process the current order according to the estimated delivery time; The device further includes: a type prediction model training module, specifically used to pre-acquire historical order features of historical orders and capacity features of historical delivery capacity for each type, and, among the historical delivery capacity of each type, determine the type of historical delivery capacity that actually delivered the historical order in history as a specified type, and treat the types of other historical delivery capacity besides the specified type as non-specified types. Using the historical order features and the capacity features of each type of historical delivery capacity as samples, input them into the type prediction model to be trained, and obtain the optimized delivery probability of the historical order delivered by the historical delivery capacity of that type for each type of historical delivery capacity output by the type prediction model to be trained. The training objective is to maximize the optimized delivery probability of the historical order delivered by the specified type of historical delivery capacity and minimize the optimized delivery probability of the historical order delivered by the non-specified type of historical delivery capacity.

8. An electronic device for processing orders, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method described in any one of claims 1 to 6.

Images