Search

CN-121544131-B - Method, system and equipment for dynamically evaluating distribution capability of distribution service provider

CN121544131BCN 121544131 BCN121544131 BCN 121544131BCN-121544131-B

Abstract

The application provides a method, a system and equipment for dynamically evaluating the distribution capacity of a distribution service provider, wherein the method comprises the steps of acquiring external data and the distribution data of the distribution service provider in real time, and determining multidimensional characteristic data and historical performance scores of the distribution service provider corresponding to a preset prediction time window according to accumulated external data and accumulated distribution data; the multi-dimensional feature data are converted into feature vectors and input into a capacity prediction model to obtain capacity prediction results corresponding to the distribution service providers, capacity scores corresponding to the distribution service providers are determined according to the capacity prediction results, and distribution capacity scores corresponding to the distribution service providers are determined according to the historical performance scores and the capacity scores. The application can dynamically evaluate the real-time delivery capacity of the delivery service providers to carry out delivery scheduling according to the real-time delivery capacity of different delivery service providers, thereby improving the delivery success rate, better meeting the delivery requirements of users, improving the user experience and improving the delivery service quality of the delivery service providers.

Inventors

  • WANG TAIZHOU

Assignees

  • 食亨(浙江)控股有限公司

Dates

Publication Date
20260512
Application Date
20260116

Claims (12)

  1. 1. A method for dynamically evaluating delivery capabilities of a delivery facilitator, the method comprising: Acquiring external data influencing order delivery and delivery data of a delivery service provider in real time, and determining multi-dimensional characteristic data and historical performance scores of the delivery service provider, which correspond to a preset prediction time window, according to the accumulated external data and the accumulated delivery data which are cut off to the current, wherein the multi-dimensional characteristic data at least comprises a bill receiving rate, average response time, current load rate, regional coverage, time characteristics, weather information and traffic information; Converting the multidimensional feature data into feature vectors and inputting a capacity prediction model to obtain a capacity prediction result of the distribution service provider corresponding to the preset prediction time window, and determining a capacity score of the distribution service provider corresponding to the preset prediction time window according to the capacity prediction result; determining a distribution capacity score of the distribution service provider corresponding to a preset prediction time window according to the historical performance score and the capacity score; For a plurality of distribution service providers, acquiring a capacity prediction result of each distribution service provider corresponding to the preset prediction time window, and predicting distribution demand distribution corresponding to the preset prediction time window according to the acquired accumulated external data and the accumulated distribution data of all distribution service providers; According to the capacity prediction results and the distribution demand distribution of all distribution service providers, which correspond to the preset prediction time window, the distribution time period and the distribution area with unbalanced potential load, which correspond to the preset prediction time window, are identified, and according to the identification results and a preset adjustment strategy, the distribution capacity score, which corresponds to the preset prediction time window, of each distribution service provider is adjusted; acquiring order information of an order to be distributed corresponding to the preset prediction time window, and determining a distribution period, a distribution area and order characteristics corresponding to the order to be distributed according to the order information; And adjusting the distribution capacity score of each distribution service provider corresponding to the distribution time period and the distribution area according to the order characteristics, and determining the distribution service provider for executing the distribution task of the order to be distributed according to the adjusted distribution capacity score of each distribution service provider and a preset selection strategy.
  2. 2. The method of claim 1, wherein determining a historical performance score for the distribution service corresponding to a preset prediction time window comprises: And determining a distribution success rate score, a distribution duration score and a user satisfaction score of the distribution service provider corresponding to a preset prediction time window according to the accumulated distribution data, and calculating the distribution success rate score, the distribution duration score and the user satisfaction score based on a first preset weight to obtain a historical performance score of the distribution service provider corresponding to the preset prediction time window.
  3. 3. The method of claim 2, wherein determining a historical performance score for the distribution service corresponding to a preset predicted time window further comprises: Determining a cost benefit score and a reliability score for the distribution server corresponding to a preset prediction time window based on the accumulated distribution data, wherein, Calculating the distribution success rate score, the distribution duration score and the user satisfaction score based on a first preset weight to obtain a historical performance score of the distribution service provider, wherein the historical performance score corresponds to a preset prediction time window and comprises the following steps: And calculating the distribution success rate score, the distribution duration score and the user satisfaction score based on a first preset weight to obtain a performance score of the distribution service provider corresponding to a preset prediction time window, and taking the performance score, the cost benefit score and the reliability score as a historical performance score of the distribution service provider corresponding to the preset prediction time window.
  4. 4. The method of claim 1, wherein the feature vector comprises at least the elements of a delivery service identifier, a pick-up rate, an average response time, a current load rate, a regional coverage, a time period performance, a weather effect factor, and a traffic effect factor.
  5. 5. The method of claim 1, wherein the constructing of the capacity prediction model comprises: Acquiring historical distribution data of a plurality of distribution service providers and corresponding historical external data influencing distribution, and determining a plurality of historical multidimensional feature data and actual capacity scores corresponding to the preset prediction time window according to the historical distribution data, the historical external data and the preset prediction time window; converting each historical multi-dimensional characteristic data into a characteristic vector, taking the actual capacity score corresponding to the characteristic vector as a true value, taking the true value as a sample, traversing each historical multi-dimensional characteristic data and the actual capacity score corresponding to the historical multi-dimensional characteristic data, and constructing a sample set; And dividing the sample set into a training set, a testing set and a verification set, and training the LSTM neural network to obtain the capacity prediction model.
  6. 6. The method of claim 1, wherein predicting the distribution demand distribution corresponding to the preset prediction time window according to the acquired accumulated external data and the accumulated distribution data of all distribution service providers comprises: Based on the obtained accumulated distribution data of all distribution service providers, predicting distribution demand basic distribution corresponding to the preset prediction time window by adopting a time sequence prediction model; determining an external influence factor corresponding to the preset prediction time window based on the acquired accumulated external data; and determining distribution demand distribution corresponding to the preset prediction time window according to the distribution demand basic distribution and the external influence factor.
  7. 7. The method of claim 1, wherein prior to said adjusting the distribution capability score of each distribution facilitator corresponding to the distribution period, distribution area, according to the order characteristics, the method further comprises: And determining regional dispatcher density scores, regional coverage scores, regional history performance scores and regional adaptation scores of the distribution service providers corresponding to the distribution time periods and the distribution regions according to accumulated distribution data of each distribution service provider, calculating the regional dispatcher density scores, the regional coverage scores, the regional history performance scores and the regional adaptation scores based on second preset weights to obtain regional matching scores of the distribution service providers corresponding to the distribution time periods and the distribution regions, and adjusting distribution capacity scores of the distribution service providers corresponding to the distribution time periods and the distribution regions according to the regional matching scores.
  8. 8. The method according to claim 1, wherein the method further comprises: And updating accumulated delivery data of the corresponding delivery service providers according to the execution result of the delivery task.
  9. 9. A system for dynamically evaluating delivery capabilities of a delivery facilitator, the system comprising: The first module is used for acquiring external data affecting order distribution and distribution data of a distribution service provider in real time, and determining multi-dimensional characteristic data and historical performance scores of the distribution service provider, which correspond to a preset prediction time window, according to the accumulated external data and the accumulated distribution data which are cut off, wherein the multi-dimensional characteristic data at least comprises a bill receiving rate, average response time, current load rate, regional coverage, time characteristics, weather information and traffic information; The second module is used for converting the multidimensional feature data into feature vectors and inputting a capacity prediction model to obtain a capacity prediction result of the distribution service provider corresponding to the preset prediction time window, and determining a capacity score of the distribution service provider corresponding to the preset prediction time window according to the capacity prediction result; a third module, configured to determine a distribution capability score of the distribution server corresponding to a preset prediction time window according to the historical performance score and the capacity score; A fourth module, configured to obtain, for a plurality of distribution service providers, a capacity prediction result corresponding to the preset prediction time window for each distribution service provider, and predict distribution demand distribution corresponding to the preset prediction time window according to the obtained accumulated external data and accumulated distribution data of all distribution service providers; A fifth module, configured to identify, according to the capacity prediction results and the distribution demand distribution of all the distribution service providers that correspond to the preset prediction time window, a distribution period and a distribution area that have unbalanced potential loads and that correspond to the preset prediction time window, and adjust, according to the identification results and a preset adjustment policy, a distribution capacity score of each distribution service provider that corresponds to the preset prediction time window; A sixth module, configured to obtain order information of an order to be distributed corresponding to the preset prediction time window, and determine a distribution period, a distribution area, and an order feature corresponding to the order to be distributed according to the order information; And an eighth module, configured to adjust a distribution capacity score corresponding to the distribution period and the distribution area for each distribution service provider according to the order feature, and determine a distribution service provider that executes the distribution task of the to-be-distributed order according to the adjusted distribution capacity score and a preset selection policy for each distribution service provider.
  10. 10. The system of claim 9, wherein the system further comprises: And a seventh module, configured to determine, according to accumulated distribution data of each distribution service provider, a regional distribution member density score, a regional coverage score, a regional history performance score, and a regional adaptation score of the distribution service provider corresponding to the distribution period and the distribution region, calculate the regional distribution member density score, the regional coverage score, the regional history performance score, and the regional adaptation score based on a second preset weight, obtain a regional matching score of the distribution service provider corresponding to the distribution period and the distribution region, and adjust a distribution capability score of the distribution service provider corresponding to the distribution period and the distribution region according to the regional matching score.
  11. 11. A computer-readable medium comprising, Having stored thereon computer readable instructions which are executed by a processor to implement part or all of the method of any of claims 1 to 8.
  12. 12. An apparatus for dynamically evaluating delivery capabilities of a delivery facilitator, the apparatus comprising: One or more processors, and A memory storing computer readable instructions that, when executed, cause the processor to perform part or all of the operations of the method of any one of claims 1 to 8.

Description

Method, system and equipment for dynamically evaluating distribution capability of distribution service provider Technical Field The application relates to the technical field of computer information processing, in particular to a technology for dynamically evaluating distribution capability of a distribution service provider. Background In the field of instant retail digital operations, an instant retail service aggregation platform typically interfaces with multiple distribution service providers, and existing order allocation strategies typically employ static weight allocation. Along with the rapid development of the instant distribution market and the aggravation of competition among the distribution platforms, the distribution capability of the distribution platforms is different in different time periods and areas, the distribution requirements of consumers, the distribution service quality requirements and the like are higher and higher, and the traditional distribution scheduling mode based on static weight distribution orders cannot be matched and met. Therefore, how to determine the real-time delivery capability of each delivery service provider to accurately, efficiently and flexibly schedule the order delivery to meet the dynamic order delivery requirement is a technical problem to be solved. Disclosure of Invention In order to solve the above technical problems, the present application aims to provide a method, a system and a device for dynamically evaluating the distribution capability of a distribution service provider. According to one aspect of the present application, there is provided a method for dynamically evaluating delivery capabilities of a delivery facilitator, wherein the method comprises: Acquiring external data influencing order distribution and distribution data of a distribution service provider in real time, and determining multidimensional feature data and historical performance scores of the distribution service provider corresponding to a preset prediction time window according to the accumulated external data and the accumulated distribution data which are cut off to the current state; Converting the multidimensional feature data into feature vectors and inputting a capacity prediction model to obtain a capacity prediction result of the distribution service provider corresponding to the preset prediction time window, and determining a capacity score of the distribution service provider corresponding to the preset prediction time window according to the capacity prediction result; And determining the distribution capacity score of the distribution service provider corresponding to a preset prediction time window according to the historical performance score and the capacity score. Optionally, the multi-dimensional characteristic data at least comprises order receiving rate, average response time, current load rate, regional coverage, time characteristics, weather information and traffic information. Optionally, determining the historical performance score of the distribution service provider corresponding to the preset prediction time window includes: And determining a distribution success rate score, a distribution duration score and a user satisfaction score of the distribution service provider corresponding to a preset prediction time window according to the accumulated distribution data, and calculating the distribution success rate score, the distribution duration score and the user satisfaction score based on a first preset weight to obtain a historical performance score of the distribution service provider corresponding to the preset prediction time window. Optionally, determining the historical performance score of the distribution service provider corresponding to the preset prediction time window further includes: Determining a cost benefit score and a reliability score for the distribution server corresponding to a preset prediction time window based on the accumulated distribution data, wherein, Calculating the distribution success rate score, the distribution duration score and the user satisfaction score based on a first preset weight to obtain a historical performance score of the distribution service provider, wherein the historical performance score corresponds to a preset prediction time window and comprises the following steps: And calculating the distribution success rate score, the distribution duration score and the user satisfaction score based on a first preset weight to obtain a performance score of the distribution service provider corresponding to a preset prediction time window, and taking the performance score, the cost benefit score and the reliability score as a historical performance score of the distribution service provider corresponding to the preset prediction time window. Optionally, the feature vector comprises at least the following elements of a delivery service identifier, a pick-up rate, an average response time, a current load ra