A multi-tenant-based terminal queuing framework system

Abstract

The invention relates to the field of station queuing, and discloses a station queuing framework system based on multiple tenants, which is used to solve the problem of untimely request response when the platform system capacity is insufficient. The system comprises a tenant management module, a resource management module, a fence pool module and a queuing module. The system manages tenants in the system, manages resources of multiple tenants, analyzes capacity index data of the system, and issues an early warning for system capacity. When it is determined that the system capacity is sufficient, it determines whether the driver enters the queuing state, records the flow and update of the driver's state, queues the driver who triggers the queue entry operation according to the queuing priority index, determines whether the driver returns to the queue according to the time when the driver leaves, and pushes the queuing information to the driver, thereby effectively ensuring the user experience and enabling timely processing when the platform system capacity is insufficient.

Description

Technical Field

[0001] The present invention relates to the field of terminal queuing, and more specifically to a terminal queuing framework system based on multiple tenants. Background Art

[0002] Terminal queuing refers to the queuing of tenants when they receive orders on the platform. Tenants queue up in a designated area at the airport or train station, and drivers queue up in the order of entering the area to receive orders from the airport or train station. With the development of online car-hailing business, the number of drivers on each platform has gradually increased.

[0003] The existing queuing method at the station is to dispatch orders based on the waiting time in the queue, but this method does not take into account the driver's income issue and the problem that when the platform system capacity is insufficient, the waiting time in the queue will be increased, affecting the quality of service for tenants.

[0004] In view of the above problems, the present invention proposes a solution. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a multi-tenant-based station queuing framework system to solve the problems existing in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A multi-tenant-based station queuing framework system includes a tenant management module, a resource management module, a fence pool module and a queuing module. The modules are connected by signals. The data processing steps between the modules are as follows:

[0008] Manage tenants in the system through the tenant management module;

[0009] The resource management module manages the resources of multiple tenants in the system and issues early warnings on system capacity by analyzing the system's capacity indicator data.

[0010] When the fence pool module determines that the system capacity is sufficient, it determines whether the driver enters the queue state and records the flow and update of the driver status;

[0011] The queuing module queues drivers who trigger the queue entry operation according to the queue priority index, determines their return to the queue based on the time the driver left, and pushes the queue information to the driver.

[0012] Preferably, the steps for managing tenants in the system are:

[0013] Confirm the tenant's identity, verify their legitimacy and authority, and authenticate them using username and password;

[0014] Assign appropriate permissions to tenants to control their access to and operations on system resources, and manage tenant permissions through the role-based access control permission model;

[0015] Set resource quotas for each tenant to limit their use of system resources and ensure that tenants do not exceed their allocated resource quotas to avoid over-consumption or abuse of system resources;

[0016] Monitor tenants' activities and operations, record their logins, requests, and resource usage, and promptly detect abnormal behaviors or illegal operations;

[0017] Monitor and count tenants' resource usage, generate bills and perform billing, provide bill query and payment functions so that tenants can understand and pay for the resources they use, ensure data security and isolation between different tenants, and use encryption, access control and other technical means to protect the security of tenant data.

[0018] Preferably, the steps of analyzing the capacity indicator data of the system and providing an early warning of the system capacity are:

[0019] Use the Prometheus system monitoring tool to monitor the current system's CPU usage, memory usage, and disk space utilization in real time, and calculate the resource utilization coefficient based on the system's CPU usage, memory usage, and disk space utilization;

[0020] Take the detection time T, collect the time data of each request response in the system within the detection time T as the data point, perform clustering on the data points, and calculate the request response coefficient based on the final clustering result;

[0021] The time points and quantities of abnormal logs are collected within the detection time T, and the abnormal log increase coefficient is calculated based on the time points and quantities of the abnormal logs;

[0022] The system capacity adequacy index is calculated by analyzing the resource utilization coefficient, request response coefficient, and abnormal log increase coefficient;

[0023] The system will issue a capacity warning based on the system capacity sufficiency index, and compare the system capacity sufficiency index with the preset threshold. If the system capacity sufficiency index is greater than the preset threshold, the system capacity is determined to be sufficient and no capacity warning is issued. If the system capacity sufficiency index is less than the preset threshold, the system capacity is determined to be insufficient and a capacity warning is issued to remind backend management personnel to expand resources in a timely manner.

[0024] Preferably, the step of clustering the data points is:

[0025] Collect the response time of each request within T time as a data point;

[0026] The number of clusters is determined by the silhouette coefficient method, the data points are clustered using the K-means clustering method, and the initial cluster centers corresponding to the determined number of clusters are selected from the data points;

[0027] The distance between each data point and each initial cluster center is calculated by the Euclidean distance algorithm, and it is assigned to a group with the nearest initial cluster center. After traversing all data points, the initial cluster cluster is obtained.

[0028] The new cluster center of the initial cluster is calculated by taking the mean, and the above clustering process is repeated until the cluster center no longer changes, and the final cluster is obtained.

[0029] Preferably, the step of calculating the request response coefficient according to the final clustering result is:

[0030] The overall average request response time is calculated by the averaging method;

[0031] The ratio of the final cluster center to the overall average request response time is calculated and recorded as the response mean coefficient;

[0032] A response mean coefficient threshold is set to filter out requests in clusters whose response mean coefficient is greater than the response mean coefficient threshold;

[0033] The average response time of the filtered requests is calculated by the mean calculation method and recorded as the request response coefficient.

[0034] Preferably, the queuing priority index calculation step is:

[0035] When the driver enters the queue, the timer starts counting the time the driver has been in the queue.

[0036] Retrieve the driver's employment time from the backend database and collect the driver's employment duration;

[0037] The distance of the current order and the distance of the driver's last order are retrieved from the backend database, and the distance coefficient is calculated based on the distance of the current order and the distance of the driver's last order;

[0038] The queue priority index is calculated based on the length of time the driver has been in the queue, the length of time the driver has been employed, and the distance coefficient.

[0039] Preferably, the step of judging the rejoining of the team according to the time when the driver left is:

[0040] When a driver is in a normal queue, if the driver leaves the queue area or collects the vehicle, the suspended state will be triggered. The driver in the suspended state can enjoy the rights and interests of receiving orders normally, and there will be a countdown of several minutes of the suspended state;

[0041] If the queue status returns to normal within the countdown time, the countdown will be deleted and the queue will be normal for order acceptance;

[0042] If the driver fails to return to the normal queue status within the countdown time, he / she will be removed from the queue by the framework. He / she will be re-queued when the conditions for joining the queue are met again. The driver status information will be updated every 5 seconds and pushed every 3 seconds.

[0043] Technical effects and advantages of the present invention:

[0044] By managing tenants in the system, managing the resources of multiple tenants, analyzing the system's capacity indicator data, and issuing early warnings on system capacity, when it is determined that the system capacity is sufficient, the driver's entry into the queue status is judged, the flow and update of the driver's status is recorded, and the drivers who trigger the queue entry operation are queued according to the queue priority index. The return to the queue is determined based on the time the driver left, and the queue information is pushed to the driver, which effectively guarantees the user experience and enables timely processing when the platform system capacity is insufficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is the overall flow chart of the present invention.

[0046] Figure 2 This is a flow chart of the present invention for performing capacity early warning on the system. DETAILED DESCRIPTION

[0047] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. In addition, the forms of the various structures recorded in the following embodiments are merely illustrative. The multi-tenant-based station queuing framework system involved in the present invention is not limited to the various structures recorded in the following embodiments. All other implementations obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0048] The present invention provides a multi-tenant-based station queuing framework system, including a tenant management module, a resource management module, a fence pool module and a queuing module, and the modules are connected by signals;

[0049] Tenant management module, used to manage tenants in the system;

[0050] Tenant management refers to a series of operations and strategies for effectively managing and controlling tenants in the system. In a multi-tenant site queuing system, tenant management involves management work such as user or organization identity authentication, permission control, and resource quota management.

[0051] In this embodiment, it should be specifically explained that the steps for managing tenants in the system are:

[0052] Confirm the tenant's identity, verify their legitimacy and authority, and authenticate them using username and password;

[0053] Assign appropriate permissions to tenants to control their access to and operations on system resources, and manage tenant permissions through the role-based access control permission model;

[0054] Set resource quotas for each tenant to limit their use of system resources and ensure that tenants do not exceed their allocated resource quotas to avoid over-consumption or abuse of system resources;

[0055] Monitor tenants' activities and operations, record their logins, requests, and resource usage, and promptly detect abnormal behaviors or illegal operations to ensure system security and stability;

[0056] Monitor and count tenants' resource usage, generate bills and perform billing, and provide bill query and payment functions so that tenants can understand and pay for the resources they use;

[0057] Ensure data security and isolation between different tenants, prevent data leakage or interference, and use encryption, access control and other technical means to protect the security of tenant data.

[0058] By managing tenants, the system can better understand the needs of different tenants, thereby optimizing and allocating resources. This helps avoid waste of resources and unnecessary overconsumption, and improves resource utilization efficiency. Tenant management can help ensure the security of the system. Reasonable permission control and access restrictions can prevent unauthorized users from accessing system resources, thereby protecting the data security and privacy of the system. Managing tenants can help the system optimize performance, and can also promptly discover and resolve performance bottlenecks, improve the response speed and stability of the system. Managing tenants helps control the operating costs of the system, ensure the effective use of resources, avoid unnecessary waste of resources, and thus reduce the operating costs of the system. Through tenant management, the system can provide personalized services to meet the needs of different tenants, which helps to improve user experience, enhance user satisfaction and loyalty to the system, and better meet compliance and regulatory requirements. For example, tenants' operations and resource usage can be recorded to audit and monitor the system to ensure system compliance.

[0059] Resource management module, used to manage the resources of multiple tenants in the system and issue early warnings on system capacity by analyzing the system's capacity indicator data;

[0060] In a multi-tenant site queuing system, resource management is particularly important because the system needs to meet the resource needs of multiple tenants at the same time and ensure the efficient use and stable operation of system resources. In a multi-tenant site queuing system, resource management can ensure that resources are effectively allocated and utilized, avoid idle or wasted resources, and thus improve resource utilization. Through resource management, resources can be reasonably allocated according to the needs and priorities of tenants, guaranteeing the service quality of important tenants or tasks and ensuring the stable operation of the system.

[0061] Reasonable resource management can avoid over-allocation or under-allocation of resources and improve the performance and throughput of the system. Resource management can flexibly adjust resources according to actual needs and usage, avoid unnecessary waste of resources, thereby reducing system operating costs, better meet the needs of tenants, and improve user satisfaction and experience. It can adjust resource configuration in real time according to system load and tenant needs, realize elastic scaling of the system, improve the flexibility and scalability of the system, and allocate resources according to the needs and priorities of different tenants to ensure the fairness and rationality of resource allocation.

[0062] In this embodiment, it should be specifically explained that the steps for managing the resources of multiple tenants are:

[0063] Identify the various types of resources required by the system, such as computing resources, network bandwidth, and storage space, and analyze system usage and demand to determine the demand and priority of each resource;

[0064] Allocate available resources to various tenants or tasks based on resource requirements and priorities, and configure resources to ensure they can meet the performance and capacity requirements of the system;

[0065] Balance the load between multiple servers or nodes to ensure that requests can be processed evenly. Use load balancing algorithms to distribute requests to different servers or nodes for processing.

[0066] Carry out capacity planning based on system development and user needs to ensure that the system can support future growth, predict resource usage trends, and expand or optimize capacity in a timely manner to meet system needs.

[0067] In this embodiment, it should be specifically explained that Figure 2 As shown, the steps of analyzing the capacity indicator data of the system and providing early warning for the system capacity are as follows:

[0068] Use the Prometheus system monitoring tool to monitor the current system's CPU usage, memory usage, and disk space utilization in real time, and calculate the resource utilization coefficient based on the system's CPU usage, memory usage, and disk space utilization. The calculation formula is: RU represents the resource utilization coefficient, RU1 represents the CPU utilization, RU2 represents the memory utilization, and RU3 represents the disk space utilization. The Prometheus system monitoring tool is an open source monitoring and alarm tool, which is widely used for real-time monitoring and collection of various performance indicators of the system, including CPU utilization, memory utilization, disk space utilization, etc. Prometheus is highly scalable and flexible and can adapt to systems of various sizes and types, including cloud native and distributed systems.

[0069] Take the detection time T, collect the time data of each request response in the system within the detection time T as the data point, perform clustering on the data points, and calculate the request response coefficient based on the final clustering result;

[0070] The time point and number of abnormal logs collected within the detection time T are used to calculate the abnormal log increase coefficient through the time point and number of abnormal logs. The calculation formula is: Where EL represents the abnormal log increase coefficient, NUM1 and NUM2 represent the number of abnormal logs at the beginning of the detection time period T and the number of abnormal logs after the end of T time;

[0071] The system capacity adequacy index is calculated by the resource utilization coefficient, request response coefficient, and abnormal log increase coefficient. The calculation formula is: Where SC represents the system capacity sufficiency index, RU represents the resource utilization coefficient. When the system resource utilization rate is high, it means that the system is approaching or reaching its capacity limit. This will cause the system to be overloaded, delays to increase, response time to become longer, and even cause system crashes or performance degradation. RP represents the request response coefficient. When the system capacity is sufficient, the system resources are sufficient to effectively process requests, and there are enough idle resources to respond to new requests. Therefore, a high system capacity sufficiency index indicates that the system has great flexibility and ability in processing requests and can achieve a shorter response time. EL represents the abnormal log increase coefficient. When the abnormal log increases at a high rate, it means that a large number of abnormal situations have occurred in the system, which may be caused by insufficient system resources or excessive load. In this case, the system capacity sufficiency index is low, indicating that the system resources are close to or have reached saturation and cannot effectively handle abnormal situations. a1, a2, and a3 represent the weight coefficients of the resource utilization coefficient RU, the request response coefficient RP, and the abnormal log increase coefficient EL. This embodiment does not specifically refer to a1, a2, and a3 for specific calculations;

[0072] The system will issue a capacity warning based on the system capacity sufficiency index, and compare the system capacity sufficiency index with the preset threshold. If the system capacity sufficiency index is greater than the preset threshold, the system capacity is determined to be sufficient and no capacity warning is issued. If the system capacity sufficiency index is less than the preset threshold, the system capacity is determined to be insufficient and a capacity warning is issued to remind backend management personnel to expand resources in a timely manner.

[0073] In this embodiment, it should be specifically explained that the steps of clustering the data points are:

[0074] Collect the response time of each request within T time as a data point. For example, the collected data points are [10, 2, 6, 1, 3, 7, 9, 11, 5];

[0075] The number of clusters is determined by the silhouette coefficient method, and the data points are clustered using the K-means clustering method. The initial cluster center corresponding to the determined number of clusters is selected from the data points. For example, if the determined cluster is 3, the initial cluster center is selected as [10, 2, 6]. The silhouette coefficient is a clustering quality evaluation indicator, which can help measure the compactness and separation of clusters. The closer the silhouette coefficient value corresponding to the number of clusters is to 1, the more reasonable the clustering is. By calculating the silhouette coefficients of different numbers of clusters, the number of clusters that maximizes the silhouette coefficient is selected as the optimal number of clusters. The K-means clustering method is a commonly used clustering algorithm for dividing a data set into K different clusters. The goal of this algorithm is to assign data points to K clusters so that the distance between each data point and the center of the cluster to which it belongs is minimized;

[0076] The distance between each data point and each initial cluster center is calculated by the Euclidean distance algorithm, and it is assigned to a group with the nearest initial cluster center. After traversing all data points, the initial cluster clusters are obtained. For example, the data point [3] is assigned and the distance to each initial cluster center is calculated as Then it is assigned together with the initial cluster center [2]. The Euclidean distance algorithm is a common method for calculating the distance between two points. It measures the straight-line distance between points in Euclidean space.

[0077] The new cluster center of the initial cluster is calculated by taking the mean. For example, if the initial cluster 2 is [2, 1, 3], the new cluster center is Then the new cluster center is [2];

[0078] Repeat the above clustering process until the cluster center no longer changes and obtain the final cluster.

[0079] In this embodiment, it should be specifically explained that the step of calculating the request response coefficient according to the final clustering result is:

[0080] The overall average request response time is calculated by the averaging method;

[0081] The ratio of the final cluster center to the overall average request response time is calculated and recorded as the response mean coefficient;

[0082] A response mean coefficient threshold is set to filter out requests in clusters whose response mean coefficient is greater than the response mean coefficient threshold;

[0083] The average response time of the filtered requests is calculated by the mean calculation method and recorded as the request response coefficient.

[0084] The fence pool module is used to judge the driver's entry into the queue state when it is determined that the system capacity is sufficient, and record the flow and update of the driver's status;

[0085] A positioning system is installed in each tenant's driver's car. When the positioning information enters the station queue area, the queue entry operation is triggered, and the queue entry time and driver status information are recorded.

[0086] Installing a positioning system in the driver's vehicle can obtain the location information of the driver's vehicle in real time. When the driver's vehicle approaches the designated queue entry position, the system can automatically trigger the queue entry operation without the driver's manual operation, thereby improving the timeliness and convenience of the operation. Automatically triggering the queue entry operation can reduce the impact of human factors on the queue entry operation, avoid the situation where the driver forgets or delays the queue entry operation, reduce the possibility of errors, and improve the accuracy and reliability of the operation. Automatically triggering the queue entry operation can accurately control the timing of queue entry according to the real-time location information of the driver's vehicle, avoiding delays caused by the driver's unfamiliarity or unclearness of the queue entry position, and improving the efficiency and speed of queue entry. The system can intelligently dispatch the queue entry order according to the vehicle location and the target station, optimize resource utilization, avoid the situation where vehicles are idle or waiting in line, and improve resource utilization.

[0087] The queuing module is used to queue drivers who have triggered the queue entry operation according to the queue priority index, determine the return to the queue according to the time when the driver left, and push the queue information to the driver;

[0088] In this embodiment, it should be specifically explained that the queuing priority index calculation steps are:

[0089] When the driver enters the queue, the timer starts counting the time the driver has been in the queue.

[0090] Retrieving the driver's employment time from the backend database and collecting the driver's employment duration, the backend database is a system platform database, including tenant information, driver information, and order information;

[0091] The distance of the current order and the distance of the driver's last order are retrieved from the backend database. The distance coefficient is calculated based on the distance of the current order and the distance of the driver's last order. The calculation formula is: Where DT is the distance coefficient, D 当 Represents the distance of the current order, D 上 Indicates the distance of the driver's last order;

[0092] The queue priority index is calculated based on the length of time the driver has been in the queue, the length of time the driver has been employed, and the distance coefficient. The calculation formula is: Wherein, PQ represents the queue priority index, QT represents the time the driver enters the queue state, the longer the driver waits in the queue, the higher the queue priority index, OT represents the driver's entry time, the platform generally provides support for newcomers, so the longer the driver's entry time, the lower the queue priority index, DT represents the distance coefficient, the larger the distance coefficient is, the greater the distance difference between the current order and the driver's previous order. In order to ensure that the driver repeatedly receives short-distance orders, the larger the distance coefficient is, the higher the queue priority index is, b1, b2, and b3 represent the weight coefficients of the driver's entry time QT, the driver's entry time OT, and the distance coefficient DT, and this embodiment does not perform specific calculations on the specific values ​​of b1, b2, and b3.

[0093] In this embodiment, it should be specifically explained that the step of queuing according to the queue priority index is to sort the queue priority indexes of the drivers entering the queue state from high to low, and give priority to dispatching orders starting from the driver with the highest queue priority index.

[0094] In this embodiment, it should be specifically explained that the step of determining the return to the team according to the time when the driver left is:

[0095] When a driver is in a normal queue, if the driver leaves the queue area or collects the vehicle, the suspended state will be triggered. The driver in the suspended state can enjoy the rights and interests of receiving orders normally, and there will be a countdown of several minutes of the suspended state;

[0096] If the queue status returns to normal within the countdown time, the countdown will be deleted and the queue will be normal for order acceptance;

[0097] If the driver fails to return to the normal queue status within the countdown time, he / she will be removed from the queue by the framework. He / she will be re-queued when the conditions for joining the queue are met again. The driver status information will be updated every 5 seconds and pushed every 3 seconds.

[0098] By queuing through the queuing module, the system can intelligently dispatch orders based on real-time station queuing conditions and driver location information, thereby improving the order delivery efficiency, better matching drivers and passengers, and intelligently matching according to passenger needs and driver characteristics, improving the personalization and quality of services, helping the system to better utilize resources, avoiding empty or idle waiting between drivers and passengers, and improving the utilization rate of vehicles and drivers. Comprehensive evaluation of driver dispatch can reduce operating costs and improve the economic benefits of the system by improving efficiency, reducing waiting time and optimizing resource utilization.

[0099] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0100] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A multi-tenant-based station queuing framework system, characterized in that: It includes tenant management module, resource management module, fence pool module and queuing module. The modules are connected by signals. The data processing steps between the modules are as follows: Manage tenants in the system through the tenant management module; The resource management module manages the resources of multiple tenants in the system and issues early warnings on system capacity by analyzing the system's capacity indicator data. When the fence pool module determines that the system capacity is sufficient, it determines whether the driver enters the queue state and records the flow and update of the driver status; The queuing module queues drivers who trigger the queue entry operation according to the queue priority index, determines their return to the queue based on the time the driver left, and pushes the queue information to the driver.

2. According to the multi-tenant-based station queuing framework system of claim 1, it is characterized by: The steps for managing tenants in the system are: Confirm the tenant's identity, verify their legitimacy and authority, and authenticate them using username and password; Assign appropriate permissions to tenants to control their access to and operations on system resources, and manage tenant permissions through the role-based access control permission model; Set resource quotas for each tenant to limit their use of system resources and ensure that tenants do not exceed their allocated resource quotas to avoid over-consumption or abuse of system resources; Monitor tenants' activities and operations, record their logins, requests, and resource usage, and promptly detect abnormal behaviors or illegal operations; Monitor and count tenants' resource usage, generate bills and perform billing, provide bill query and payment functions so that tenants can understand and pay for the resources they use, ensure data security and isolation between different tenants, and use encryption, access control and other technical means to protect the security of tenant data.

3. According to the multi-tenant-based station queuing framework system of claim 1, it is characterized by: The steps of analyzing the capacity indicator data of the system and providing early warning for the system capacity are as follows: Use the Prometheus system monitoring tool to monitor the current system's CPU usage, memory usage, and disk space utilization in real time, and calculate the resource utilization coefficient based on the system's CPU usage, memory usage, and disk space utilization; Take the detection time T, collect the time data of each request response in the system within the detection time T as the data point, perform clustering on the data points, and calculate the request response coefficient based on the final clustering result; The time points and quantities of abnormal logs are collected within the detection time T, and the abnormal log increase coefficient is calculated based on the time points and quantities of the abnormal logs; The system capacity adequacy index is calculated by analyzing the resource utilization coefficient, request response coefficient, and abnormal log increase coefficient; The system will issue a capacity warning based on the system capacity sufficiency index, and compare the system capacity sufficiency index with the preset threshold. If the system capacity sufficiency index is greater than the preset threshold, the system capacity is determined to be sufficient and no capacity warning is issued. If the system capacity sufficiency index is less than the preset threshold, the system capacity is determined to be insufficient and a capacity warning is issued to remind backend management personnel to expand resources in a timely manner.

4. According to the multi-tenant-based station queuing framework system of claim 3, it is characterized by: The steps of clustering the data points are as follows: Collect the response time of each request within T time as a data point; The number of clusters is determined by the silhouette coefficient method, the data points are clustered using the K-means clustering method, and the initial cluster centers corresponding to the determined number of clusters are selected from the data points; The distance between each data point and each initial cluster center is calculated by the Euclidean distance algorithm, and it is assigned to a group with the nearest initial cluster center. After traversing all data points, the initial cluster cluster is obtained. The new cluster center of the initial cluster is calculated by taking the mean, and the above clustering process is repeated until the cluster center no longer changes, and the final cluster is obtained.

5. According to the multi-tenant-based station queuing framework system of claim 3, it is characterized by: The step of calculating the request response coefficient according to the final clustering result is: The overall average request response time is calculated by the averaging method; The ratio of the final cluster center to the overall average request response time is calculated and recorded as the response mean coefficient; A response mean coefficient threshold is set to filter out requests in clusters whose response mean coefficient is greater than the response mean coefficient threshold; The average response time of the filtered requests is calculated by the mean calculation method and recorded as the request response coefficient.

6. According to a multi-tenant-based station queuing framework system according to claim 1, it is characterized by: The queuing priority index calculation steps are: When the driver enters the queue, the timer starts counting the time the driver has been in the queue. Retrieve the driver's employment time from the backend database and collect the driver's employment duration; The distance of the current order and the distance of the driver's last order are retrieved from the backend database, and the distance coefficient is calculated based on the distance of the current order and the distance of the driver's last order; The queue priority index is calculated based on the length of time the driver has been in the queue, the length of time the driver has been employed, and the distance coefficient.

7. The multi-tenant-based station queuing framework system according to claim 1 is characterized in that: The steps of judging the return to the team according to the time when the driver left are as follows: When a driver is in a normal queue, if the driver leaves the queue area or collects the vehicle, the pause state will be triggered. The driver in the pause state can enjoy the rights and interests of receiving orders normally, and there will be a countdown of several minutes of the pause state; If the queue status returns to normal within the countdown time, the countdown will be deleted and the queue will be normal for order acceptance; If the driver fails to return to the normal queue status within the countdown time, he / she will be removed from the queue by the framework. He / she will be re-queued when the conditions for joining the queue are met again. The driver status information will be updated every 5 seconds and pushed every 3 seconds.

Images