Intelligent Loan Approval Process Abnormality Management and Fault Tolerance System

Abstract

The present invention relates to an intelligent loan approval process exception management and fault tolerance system, aiming to optimize the loan approval process during high-traffic periods and enhance the stability of the system. The system includes multiple key modules: The interception and degradation module is automatically activated when detecting high traffic through the service degradation switch, and decides whether to intercept loan applications according to the preset product code, partner code, and interception ratio parameters. The timeout degradation policy module includes a timeout salvage thread for processing loan applications that have not been approved on time in the system. The traffic shaping and flow limiting module is responsible for reducing the load on the downstream system by suspending the current processing task or pushing it to the message queue when the number of failed data interface requests exceeds the preset threshold. The risk control exception stop-loss module monitors the approval results of the risk control system and calculates key performance indicators. This system effectively responds to emergencies through dynamic policy adjustment and real-time monitoring, ensuring the continuity and efficiency of the loan approval process.

Description

Technical Field

[0001] The present invention relates to the technical field of loan approval, and particularly to an intelligent loan approval process exception management and fault tolerance system. Background Art

[0002] In the field of loan approval, the application of intelligent systems has significantly improved the efficiency and accuracy of the approval process. In the prior art, most loan approval systems rely on automated decision engines, which can quickly approve loans based on preset credit scoring models and information of loan applicants. These systems usually include various modules to process data streams and requests during normal operations, such as risk assessment, credit check, and approval decision modules. The development of these technologies has greatly improved the processing speed and reduced human errors, enabling loan institutions to handle a large number of applications while maintaining high efficiency and low risk levels.

[0003] However, existing approval systems often perform poorly in the face of high-traffic applications or emergencies, which may lead to system delays, performance bottlenecks, or even system crashes. During high-demand periods, such as promotions or economic crises, the system may respond slowly or fail due to being overwhelmed by the sudden increase in applications. In addition, existing systems often lack sufficient flexibility and adaptability to dynamically adjust processing strategies to cope with these challenges when suddenly encountering external factor interferences (such as data interface failures). These problems not only affect the user experience but may also cause incorrect decisions in loan approvals, increasing the operational risks of financial institutions.

[0004] Therefore, it is very necessary to develop an intelligent loan approval process exception management and fault tolerance system. Summary of the Invention

[0005] This application provides an intelligent loan approval process exception management and fault tolerance system to optimize the loan approval process during high-traffic periods and enhance the stability of the system.

[0006] This application provides an intelligent loan approval process exception management and fault tolerance system, including:

[0007] Interception Degradation Module, configured with a service degradation switch, which is activated when a high-traffic loan application is detected; after the service degradation switch is activated, intercept or release the loan application according to the preset product code, partner code, and interception ratio parameter; for the first loan application that is intercepted, record its product code, partner code, business application type, business application order number, business urgency, interception time, whether the approval has started, whether the approval is completed, and the approval completion time in the service degradation task table; regularly extract the recorded loan applications from the service degradation task table, and asynchronously execute the approval process according to the priority determined by the business urgency and interception time of the recorded loan applications; and after the approval is completed, update the corresponding approval result in the service degradation task table.

[0008] Timeout Degradation Policy Module, including a timeout degradation service switch and a timeout salvage thread; when the timeout degradation service switch is turned on, the timeout salvage thread regularly checks the second loan applications in the system that have not completed the approval on time; determine the third loan applications with high urgency among the second loan applications that have not completed the approval on time according to the product code, partner code, and current loan approval time; increase the business urgency of the third loan applications, and save the third loan applications with increased business urgency to the service degradation task table.

[0009] Traffic Shaping and Rate Limiting Module, used to count the number of failed requests to the downstream system by the external data interface within a specified time; if the number of failed requests is greater than the first preset threshold and less than the second preset threshold, suspend the current processing task, notify the downstream system to process, and restart the suspended current processing task after the downstream system finishes processing; if the number of failed requests is greater than the second preset threshold, push the current processing task to the message queue and slow down the pressure on the downstream system by configuring the concurrency parameters of the downstream system.

[0010] Risk Control Abnormal Stop-Loss Module, used to monitor the approval results of the risk control system; calculate key indicators including the approval passing rate and the number of consecutive rejections; compare the calculated key indicators with the preset normal range; when the key indicators deviate from the preset normal range, suspend all loan approval processes and issue a warning.

[0011] Furthermore, the Interception Degradation Module further includes an adaptive interception ratio adjustment unit, used to monitor the system resource utilization rate and request processing time in real time; when the system resource utilization rate exceeds the preset threshold or the average request processing time increases significantly, automatically increase the interception ratio; when the system resource utilization rate drops to a safe level and the average request processing time returns to normal, gradually reduce the interception ratio; the adjustment of the interception ratio adopts a progressive strategy to avoid drastic fluctuations in the system load.

[0012] Further, the adaptive interception ratio adjustment unit includes a dynamic adjustment algorithm, and the dynamic adjustment algorithm adjusts the interception ratio in real time according to the following formula (1):

[0013]

[0014] wherein, is the new interception ratio; is the current interception ratio; is the actual system utilization rate; is the target system utilization rate; is the actual request processing time; is the target request processing time; and are adjustment coefficients used to adjust the influence of system utilization rate and processing time on the interception ratio; is the time difference from the moment when the system resource utilization rate exceeds the preset utilization rate threshold to the moment when the dynamic adjustment algorithm is currently executed; and are progressive adjustment functions, which are implemented respectively by the following formula (2) and formula (3):

[0015]

[0016]

[0017] wherein, and are adjustment coefficients.

[0018] Further, the interception degradation module includes a warning unit, and the warning unit is used to notify the system administrator by email or text message when it is predicted that the system is about to reach the load limit, and automatically start additional cloud resources to cope with the upcoming high traffic, so as to avoid system overload.

[0019] Further, the timeout degradation policy module includes a monitoring interface for displaying all current loan applications that have not been approved on time and their business urgency, wherein the monitoring interface allows the administrator to manually adjust their business urgency.

[0020] Further, the timeout degradation service switch can be automatically activated based on real-time weather or traffic accident data to identify possible approval delays caused by emergencies.

[0021] Further, the traffic shaping and flow limiting module includes an automatic failover subsystem, and the failover subsystem switches to the standby system when the number of failures of the downstream system is detected to exceed the first preset threshold, so as to ensure the continuity of the processing task and the high availability of the system.

[0022] Further, the traffic shaping and rate limiting module includes an adaptive threshold adjustment unit for dynamically adjusting a first preset threshold and a second preset threshold according to historical data and current system performance.

[0023] Further, the risk control abnormal stop-loss module includes an interactive risk control adjustment interface that allows an administrator to adjust risk control strategy parameters according to real-time data and system recommendations, including real-time adjustment of the approval passing rate threshold and the rejection times threshold.

[0024] Further, the risk control abnormal stop-loss module realizes interface docking with an external credit scoring agency and automatically obtains the latest credit information of the applicant from the external credit scoring agency.

[0025] The beneficial effects of the technical solution provided by this application include:

[0026] (1) Through the service degradation switch of the interception and degradation module, the system can automatically adjust the processing capacity when detecting high-traffic loan applications, preventing the system from overloading. This automated interception and release operation enables the system to maintain stable operation even under extreme traffic conditions, thereby improving the overall system responsiveness and stability. (2) The timeout degradation policy module ensures that loan applications with high urgency are processed quickly by regularly checking and preferentially processing loan applications that have not been approved on time. This not only optimizes the allocation of resources but also ensures a quick response to high-priority applications, thereby enhancing customer satisfaction and service efficiency. (3) The traffic shaping and rate limiting module can dynamically adjust the processing strategy according to the failure times of the external data interface. For example, when a certain threshold is reached, the current task is suspended or the task is pushed to the message queue. Such a dynamic management mechanism reduces the pressure on the downstream system and prevents service interruptions caused by system paralysis, enhancing the overall system robustness. (4) The risk control abnormal stop-loss module monitors key indicators of the approval results and compares them with preset standards. Once a deviation is found, all loan approval processes can be suspended and a warning can be issued in a timely manner, which helps to correct possible risk control strategy mistakes in a timely manner and ensure the compliance and approval quality of financial institutions. This immediate monitoring and reaction mechanism greatly reduces the potential risks and financial losses caused by approval errors. Description of the Drawings

[0027] Figure 1 is a schematic diagram of an intelligent loan approval process abnormal management and fault tolerance system provided by the first embodiment of this application. Detailed Embodiments

[0028] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application. However, the present application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the connotation of the present application. Therefore, the present application is not limited by the specific implementations disclosed below.

[0029] The first embodiment of the present application provides an intelligent loan approval process exception management and fault tolerance system. Please refer to Figure 1 , which is a schematic diagram of the first embodiment of the present application. The following will be combined with Figure 1 to describe in detail an intelligent loan approval process exception management and fault tolerance system provided by the first embodiment of the present application. The intelligent loan approval process exception management and fault tolerance system includes an interception and degradation module 101, a timeout degradation policy module 102, a traffic shaping and flow limiting module 103, and a risk control exception stop loss module 104.

[0030] The interception and degradation module 101 is configured with a service degradation switch, and the service degradation switch is activated when a high-flow loan application is detected; after the service degradation switch is activated, the loan application is intercepted or released according to preset product codes, partner codes, and interception ratio parameters; for the first loan application that is intercepted, its product code, partner code, business application type, business application order number, business urgency, interception time, whether the approval has started, whether the approval is completed, and the approval completion time are recorded in the service degradation task table; the recorded loan applications are regularly extracted from the service degradation task table, and the approval process is asynchronously executed according to the priority determined by the urgency and interception time of the recorded loan applications; and after the approval is completed, the corresponding approval result in the service degradation task table is updated.

[0031] In this embodiment, the interception and degradation module 101 is the core part of the intelligent loan approval process exception management and fault tolerance system, and its design aims to maintain the stability and responsiveness of the system under high-load conditions. This module has a service degradation switch, and the main function of this switch is to automatically activate when a high-flow loan application is detected. This function is crucial for handling sudden high-load application flows because it can prevent the system from overloading and thus avoid possible service interruptions.

[0032] Once the service degradation switch is activated, the interception and degradation module 101 will decide whether to intercept the incoming loan application based on several pre-set key parameters. These parameters include product codes, partner codes, and interception ratio parameters, which are all set in advance according to the types of loan products and the specific requirements of partners. In this way, the system can flexibly distinguish and process applications from different sources or types, ensuring that key or urgent loan applications can be given priority for approval.

[0033] The product code refers to the unique identifier assigned to different loan products. This helps the system distinguish various loan products, such as personal loans, mortgage loans, or auto loans, etc. Each type of loan may have different risk assessment criteria and approval processes. For example, a personal loan may have the code PL001, while a mortgage loan may be coded as HL002.

[0034] The partner code, on the other hand, refers to the unique identifier between the lending institution and different partners, which may be other banks, financial institutions, or sales channels. For example, the code for cooperation with a large retail bank may be CB100, and the code for cooperation with an auto dealership may be AD200. This helps the system track and manage the application flow from different channels.

[0035] The business application type generally refers to the specific type of loan, such as a new loan, a renewal loan, or a structured loan, etc. This helps the system understand the specific nature of the application and process it accordingly.

[0036] The business application serial number is the unique identifier for each loan application, ensuring that each application can be accurately tracked and processed.

[0037] The business urgency reflects the priority of the application, which is usually determined by the urgency of the applicant's needs or the characteristics of the loan product. For example, the urgency of an emergency medical loan may be set higher than that of an ordinary consumer loan.

[0038] The interception ratio parameter, which is a value dynamically calculated based on the current system load and historical data, is used to determine what percentage of loan applications the system should intercept within a specific time.

[0039] After intercepting a loan application, each intercepted application will be recorded in a table called the service degradation task table. The information recorded includes the product code, partner code, business application type, business application serial number, business urgency, interception time, whether approval has started, whether approval is completed, and the approval completion time. These detailed records not only provide the necessary information for the subsequent approval process but also offer data support for system auditing and performance evaluation.

[0040] From the service degradation task table, the system will regularly extract the recorded loan applications and determine the processing priority based on the urgency and interception time of each application. This step is executed asynchronously, meaning that the system can flexibly process each loan application according to the current operating conditions and resource availability. Once the approval of an application is completed, the system will update the approval result of the corresponding record in the service degradation task table to ensure that all data remains up-to-date.

[0041] The system will regularly and automatically check the service degradation task list to determine the loan applications that need to be processed with priority. In this process, the urgency level and the interception time of each application are two key factors determining its processing priority.

[0042] The urgency level is usually determined by the nature of the application itself. For example, loan applications involving high risks or high values may be marked with a higher urgency level. In addition, under specific circumstances, the needs of the customers may also lead to certain applications being marked with a high urgency level, such as emergency medical loans or other emergency personal loans.

[0043] The interception time provides a timestamp indicating the specific time when the application is intercepted by the system. This information is very useful for managing the application queue because it helps the system identify which applications have been waiting for a long time. When dealing with high loads, it can ensure that no application will be overly delayed even during peak traffic periods.

[0044] The system calculates the priority of each application based on the urgency level and the interception time of these applications. Generally speaking, applications with a high urgency level, or those that have been waiting in the system for a long time, will be given a higher processing priority. This calculation of priority ensures that all applications can be processed fairly and in a timely manner, especially maintaining the efficiency and quality of the service when system resources are strained.

[0045] With such a design, the interception degradation module 101 not only improves the efficiency and response speed of the loan approval process, but also significantly enhances the stability and reliability of the system when facing high-traffic applications. In addition, the implementation of this module ensures that the system can effectively manage and respond to potential load fluctuations while maintaining a high service quality, which is of great value to financial service providers.

[0046] Furthermore, the interception degradation module further includes an adaptive interception ratio adjustment unit for real-time monitoring of the system resource utilization rate and the request processing time; when the system resource utilization rate exceeds the preset threshold or the average request processing time increases significantly, automatically increase the interception ratio; when the system resource utilization rate drops to a safe level and the average request processing time returns to normal, gradually reduce the interception ratio; the adjustment of the interception ratio adopts a progressive strategy to avoid drastic fluctuations in the system load.

[0047] In the intelligent loan approval process anomaly management and fault tolerance system, the interception degradation module plays a crucial role, especially its adaptive interception ratio adjustment unit, which is designed to maintain system balance, respond quickly and effectively handle peak traffic. This unit ensures that the system can automatically adjust its processing capacity when facing different operating pressures, thereby optimizing resource usage and avoiding overload.

[0048] The working principle of the adaptive interception ratio adjustment unit is based on two core monitoring metrics: system resource utilization rate and request processing time. The system resource utilization rate reflects the load level of the system, including CPU usage, memory occupancy, network bandwidth usage, etc. The request processing time refers to the time required from receiving a request to returning a result. These two metrics jointly reflect the performance status and processing capacity of the system.

[0049] Preset thresholds are set inside the unit, and these thresholds are pre-configured based on system design and historical performance data. When the system monitors that the resource utilization rate exceeds these thresholds, or the average request processing time significantly increases, it means that the system may be facing high operating pressure or about to be overloaded. At this time, the adaptive interception ratio adjustment unit will automatically increase the interception ratio, that is, increase the interception rate of the system for incoming requests, so as to reduce the occupation of system resources by new requests and prevent the system from crashing due to overloading.

[0050] Specifically, increasing the interception ratio may mean temporarily not processing non-urgent or non-priority loan applications, which can be achieved by delaying the processing of these requests or temporarily storing them in a buffer, etc. This strategy allows the system to concentrate limited resources on processing current high-priority or urgent requests, thus maintaining service quality and response speed.

[0051] When the system detects that the resource utilization rate drops to a safe level and the request processing time returns to the normal range, the interception ratio adjustment unit will gradually reduce the interception ratio. This progressive reduction strategy is to ensure that the system smoothly transitions back to the normal operation state and avoid performance fluctuations or instability caused by suddenly increasing the processing load.

[0052] In addition, the adjustment of the interception ratio adopts a progressive strategy, which means that any adjustment of the interception ratio will not immediately jump from one extreme to another, but is gradually achieved through a series of small and controllable steps to ensure the stability and continuity of the system.

[0053] Through this adaptive adjustment mechanism, the intelligent loan approval process exception management and fault tolerance system can effectively cope with various traffic and load conditions, ensuring that the system can maintain stable operation even in extreme cases and provide consistent and reliable services. The implementation of this technology is extremely valuable for any loan approval system that relies on high availability and high performance.

[0054] Furthermore, the adaptive interception ratio adjustment unit includes a dynamic adjustment algorithm, and the dynamic adjustment algorithm adjusts the interception ratio in real time according to the following formula (1):

[0055]

[0056] Where, is the new interception ratio; is the current interception ratio; is the actual system utilization rate; is the target system utilization rate; is the actual request processing time; is the target request processing time; and is the adjustment coefficient, used to adjust the impact of system utilization rate and processing time on the interception ratio; is the time difference from the moment when the system resource utilization rate exceeds the preset utilization rate threshold to the current moment when the dynamic adjustment algorithm is executed; and is the progressive adjustment function, implemented respectively by the following formulas (2) and (3):

[0057]

[0058]

[0059] where and are the adjustment coefficients.

[0060] In the intelligent loan approval process anomaly management and fault tolerance system, the adaptive interception ratio adjustment unit adopts a dynamic adjustment algorithm, which adjusts the system's interception ratio in real time through a precisely defined mathematical model to respond to changes in system load and fluctuations in request processing time. The following details each component of this algorithm and its implementation method.

[0061] Composition of the dynamic adjustment formula:

[0062] 1. (New interception ratio): This is the adjusted interception ratio, and its value is dynamically calculated based on the current interception ratio and several other factors.

[0063] 2. (Current interception ratio): This is the interception ratio currently set by the system before the new calculation. This value is obtained through real-time monitoring and directly read from the system's operating parameters.

[0064] 3. (Actual system utilization rate): This is the current resource utilization rate of the system, such as CPU utilization rate, memory occupancy, etc., which can be obtained in real time through system performance monitoring tools.

[0065] 4. (Target system utilization rate): This is the ideal resource utilization level set during system design, usually obtained based on optimal operating conditions and performance tests.

[0066] 5. (Actual request processing time): This refers to the time required for the system to actually process a loan request, and this dynamic data can be obtained from logs or monitoring systems.

[0067] 6. (Target request processing time): This is the standard for request processing time that the system should achieve, usually set according to performance requirements during the system design phase.

[0068] 7. and (Adjustment coefficients): These coefficients determine the impact strength on the adjustment of the interception ratio when the system utilization rate and processing time deviate from the target values. These coefficients are usually adjusted by system analysts based on historical data and expected performance.

[0069] 8. (Time difference): This is the time difference from the moment when the system utilization rate exceeds the preset system utilization rate threshold ( ) to the current moment of executing the dynamic adjustment algorithm, usually obtained through the system clock.

[0070] Definition of the progressive adjustment function:

[0071] 1.

[0072] This is a logistic adjustment function used to progressively adjust the interception ratio according to the change of time, where is the adjustment coefficient, which determines the adjustment speed.

[0073] 2. : This is an adjustment function based on the Gaussian distribution, which takes into account the squared influence of time, making the adjustment gradually approach the stable state as time increases. is the adjustment coefficient, controlling the rate and shape of the adjustment. Coefficients and are usually adjusted by system analysts based on historical data and expected performance.

[0074] These formulas and coefficients cooperate with each other to ensure that the system can dynamically, efficiently, and smoothly adjust its interception ratio in the face of changing operating conditions, so as to maintain system performance and responsiveness, and avoid system crashes caused by excessive load. By adjusting the interception ratio in real time, the system can better manage sudden high-traffic situations and ensure the efficiency and fairness of all processing.

[0075] ​Furthermore, the interception and degradation module includes a warning unit, which is used to notify the system administrator via email or text message when it is predicted that the system is about to reach its load limit, and automatically start additional cloud resources to cope with the upcoming high traffic, thus avoiding system overload.

[0076] In the intelligent loan approval process anomaly management and fault tolerance system, the warning unit built into the interception and degradation module is a key security function designed to ensure the continuous and stable operation of the system when facing potential high traffic loads. The design of this warning unit allows the system to take preventive measures before reaching the load limit, including notifying the system administrator and automatically expanding system resources. The following are the working principles and implementation details of this unit.

[0077] The core function of the warning unit is to monitor the real-time load situation of the system and automatically execute a series of predefined actions to prevent system overload when it detects that the load is about to reach the system processing limit. This includes sending notifications to the system administrator and starting additional cloud resources.

[0078] 1. Real-time load monitoring: The warning unit continuously detects various performance indicators of the system, such as CPU usage, memory usage, network traffic, and request response time. This data is collected in real-time through integrated system monitoring tools to ensure the accuracy and timeliness of the data.

[0079] 2. Load prediction: By analyzing historical data and current performance indicators, the warning unit uses prediction algorithms to evaluate whether the system is about to reach its load limit. This prediction relies on advanced data analysis techniques, such as time series analysis or machine learning models, which can learn from past trends and predict future load situations.

[0080] 3. Notification mechanism: Once it is predicted that the system is about to reach its load limit, the warning unit will immediately send a warning to the system administrator via email or text message. These notifications contain key performance data and recommended countermeasures, enabling the administrator to respond quickly. The sending of notifications depends on the contact information and communication protocols configured within the system.

[0081] 4. Automatic resource expansion: At the same time, the warning unit will trigger automated scripts or API calls to start additional resources on the cloud platform, such as additional server instances or increasing the processing capacity of the database. This process is achieved through integration with the interfaces of cloud service providers to ensure that the processing capacity of the system can be quickly expanded when needed.

[0082] The timeout degradation policy module 102 includes a timeout degradation service switch and a timeout salvage thread; when the timeout degradation service switch is turned on, the timeout salvage thread periodically checks the second loan applications in the system that have not been approved on time; determines the third loan applications with a high degree of urgency among the second loan applications that have not been approved on time according to the product code, partner code, and the current loan approval time; increases the business urgency of the third loan applications, and saves the third loan applications with the increased business urgency to the service degradation task table.

[0083] The timeout degradation policy module 102 is a key component of the intelligent loan approval process exception management and fault tolerance system, aiming to optimize the approval process and ensure the smooth and efficient maintenance of approval activities even in the face of high load or system latency. This module includes a timeout degradation service switch and a timeout salvage thread, which work together to monitor and adjust loan approvals that have not been completed in a timely manner due to processing delays.

[0084] The timeout degradation service switch is the core control unit of the system, responsible for monitoring the operating status of the entire loan approval system. When the system detects that the processing time of the loan approval exceeds the normal expectation, this switch will be automatically triggered. The triggering of the switch is based on real-time analysis of various factors such as the current load of the system, processing time, and the number of applications to be processed.

[0085] Once the timeout degradation service switch is activated, the timeout salvage thread begins to play its role. This thread is specifically designed to periodically check all loan applications that have not been approved within the scheduled time. It evaluates the urgency of each uncompleted application by analyzing specific information of each application, such as product code, partner code, and loan approval time. In this process, the thread will use preset logic and parameters to determine which applications should be given priority, especially those applications that are considered to have a high degree of business urgency.

[0086] In the intelligent loan approval process exception management and fault tolerance system, the timeout salvage thread plays a crucial role, especially when dealing with loan applications that have not been approved on time. This thread evaluates the urgency of each application by precisely analyzing the detailed information of each application, ensuring that the system can prioritize the most urgent cases under limited resources.

[0087] For example, consider a loan application. Its product code may indicate that it is a high-value housing loan (e.g., code HL001), and the partner code may indicate that it is submitted through a major bank channel (e.g., code B001). These pieces of information themselves provide important clues about the nature of the application. High-value housing loans usually have a higher priority because they involve a large amount of funds and are usually related to the urgent needs of customers.

[0088] Furthermore, the loan approval time is also a key factor in determining the urgency of an application. If the application has approached or exceeded the normal processing time window (assuming the normal approval time is two weeks), the system will adjust its urgency to a higher level. This means that even during high-traffic periods, the system will try its best to prioritize the processing of such overdue applications to avoid adverse impacts on customers.

[0089] When analyzing this data, the overdue salvage thread will use preset logic and parameter formulas for calculation. For example, it may use a weighted system where the loan type corresponding to the product code has a weight of 50%, the partner code has a weight of 30%, and each day beyond the normal approval time increases the urgency by 20%. According to such a system, a high-value loan application submitted through a major bank channel and five days overdue will receive an extremely high emergency processing rating.

[0090] Through this method, the overdue salvage thread can systematically identify and mark loan applications that require urgent processing, ensuring that they can be sent into the approval process as soon as possible. This not only optimizes the use of resources but also significantly improves customer satisfaction and the overall efficiency of the system, thus strengthening the reliability and responsiveness of the loan approval process.

[0091] To further improve processing efficiency, the thread not only identifies applications with high urgency but also dynamically adjusts the business urgency of these applications to ensure that they are prioritized for processing when system resources permit. Once the business urgency is increased, these applications will be re-saved in the service degradation task table so that the system can reschedule these tasks for processing.

[0092] In addition, the overdue salvage thread is also responsible for updating the approval status of each application in the service degradation task table, including whether the approval has started, whether the approval is completed, and the time when the approval is completed. This ensures that the system has a detailed record and tracking of the approval progress of all loan applications, making the approval process more transparent and manageable.

[0093] Through such a design, the overdue degradation policy module 102 not only improves the adaptability and flexibility of the loan approval process but also enhances the system's response to uncertain factors and potential delays. The implementation of this policy is crucial for maintaining the efficiency and stability of the loan approval system, especially under high-load or other abnormal operating conditions.

[0094] Furthermore, the overdue degradation policy module includes a monitoring interface for displaying all currently unapproved loan applications on time and their business urgency, where the monitoring interface allows the administrator to manually adjust their business urgency.

[0095] In the intelligent loan approval process anomaly management and fault tolerance system, a key component of the timeout degradation strategy module is the monitoring interface, which provides an intuitive way to view and manage all loan applications that have not been approved on time. The design of this monitoring interface aims to enhance the transparency and operability of the system, enabling administrators to understand the approval progress in real time and intervene when necessary. The following are the function and operation descriptions of the monitoring interface.

[0096] The main function of the monitoring interface is to display in real time all current loan applications that have not been approved on time. The display information for each loan application includes, but is not limited to:

[0097] Detailed information of the loan application: such as application number, application date, applicant information, loan amount, etc.

[0098] Current business urgency: This is a key indicator that indicates the processing priority of each loan application. Applications with a high business urgency indicate that they need to be processed first.

[0099] In addition, the functions provided by the monitoring interface are not limited to information display. It also allows administrators to manually adjust the business urgency of each loan application. This function is of great significance for flexibly responding to emergencies and adjusting the allocation of approval resources.

[0100] The operation process includes:

[0101] Administrators enter the monitoring interface through a secure login process. This usually requires verifying the identity of the administrator to ensure that only authorized personnel can access sensitive data and the control panel.

[0102] All loan applications that have not been approved on time will be listed in real time on the interface. These applications are sorted by factors such as business urgency and application time, so that administrators can quickly identify applications that need to be processed urgently.

[0103] Administrators can directly adjust the business urgency of a specific loan application through tools provided on the interface, such as a slider or input box. After the adjustment, the system will update this information in real time and re-prioritize the relevant application according to the new urgency.

[0104] Any adjustment of the urgency needs to be submitted and confirmed. This is usually done by clicking the "Save" or "Confirm Changes" button. The system will perform necessary validations to ensure that the changes are reasonable and comply with the system's business rules.

[0105] All adjustment activities will be recorded by the system in the operation log for auditing and tracing. At the same time, the system will provide operation feedback to confirm that the changes have been successfully implemented or to alert of errors.

[0106] To ensure the effective operation of the monitoring interface, it is necessary to implement the data processing logic at the backend, the user interface design at the front end, and the data transmission and security mechanisms in the middleware. Technical personnel need to ensure that the user experience design of the interface is simple and intuitive, the data update is highly real-time, and the overall security of the system is guaranteed, including data encryption and access control.

[0107] Through this monitoring interface, the intelligent loan approval process anomaly management and fault tolerance system provides a powerful tool for administrators to manage the loan approval process more effectively, ensuring timely response to various approval challenges and maintaining the efficiency of the approval process.

[0108] Furthermore, the timeout degradation service switch can be automatically activated based on real-time weather or traffic accident data to identify possible approval delays caused by unexpected events.

[0109] In the intelligent loan approval process anomaly management and fault tolerance system, an advanced function of the timeout degradation service switch is its ability to be automatically activated according to real-time weather or traffic accident data. This function is designed to enable the system to respond to unexpected external events that may affect the approval progress, ensuring that the system can take preventive measures in the face of possible delays, thereby maintaining the efficiency and effectiveness of the approval process.

[0110] By integrating real-time data sources, the timeout degradation service switch can monitor real-time events related to weather or traffic. The system has preset specific logic to analyze this data and evaluate its possible impact on the loan approval process. Emergency situations caused by weather or traffic accidents may lead to an increase in the demand for specific loan products. For example, in some cases, the demand for emergency loans related to disasters or loans related to insurance claims will rise, which may indirectly increase the urgency and complexity of approval tasks.

[0111] 1. Data source access: The system needs to access multiple real-time data sources, such as national or regional weather forecast services and traffic management systems. These data sources provide real-time weather updates and traffic condition reports, and the system obtains data by calling these services through APIs.

[0112] 2. Data analysis: The acquired data will be processed in real-time by the analysis module of the system. The analysis module determines whether the current weather or traffic condition reaches the threshold that may potentially affect the approval process according to the preset parameters.

[0113] The automatic activation mechanism includes:

[0114] 1. Condition determination: When the monitored weather or traffic condition meets the preset impact conditions, such as reaching a certain severity level, the timeout degradation service switch will be automatically activated. These conditions may include specific weather warning levels or traffic congestion indices.

[0115] 2. Activation Response: Once the timeout degradation service switch is activated, the system will automatically take a series of preventive measures. This may include temporarily adjusting the approval process, such as extending the approval time window, or enabling more online approval resources when necessary to make up for the shortage of physical approvals.

[0116] Through this design, the intelligent loan approval process anomaly management and fault tolerance system can quickly and effectively respond to external emergencies, minimizing the negative impact of these events on the loan approval process. This not only improves the adaptability and resilience of the system but also ensures the continuity and quality of customer service.

[0117] The traffic shaping and rate limiting module 103 is used to count the number of failed requests to the downstream system by the external data interface within a specified time; if the number of failed requests is greater than the first preset threshold and less than the second preset threshold, the current processing task is suspended, and the downstream system is notified to process. After the downstream system finishes processing, the suspended current processing task is restarted; if the number of failed requests is greater than the second preset threshold, the current processing task is pushed to the message queue, and by configuring the concurrency parameters of the downstream system, the pressure on the downstream system is alleviated.

[0118] The traffic shaping and rate limiting module 103 plays a crucial role in the intelligent loan approval process anomaly management and fault tolerance system. Its main task is to ensure that the system can maintain stability and responsiveness in the face of various external data interface failures or high-traffic requests. The module monitors and analyzes the number of failed requests of the external data interface in real time and takes corresponding strategies to regulate and control the pressure of the data flow to the downstream system, thus avoiding system overload and service quality degradation.

[0119] Specifically, the traffic shaping and rate limiting module 103 first counts all data requests entering the system, which includes but is not limited to data requests received from external systems such as cooperative banks, credit assessment agencies, etc. There are two key preset thresholds inside the module: the first preset threshold and the second preset threshold. These thresholds are set in advance based on historical data and system capacity to indicate when specific actions need to be taken to prevent system overload.

[0120] When the number of failed external requests reaches the first preset threshold but does not exceed the second preset threshold, the module will automatically suspend the currently processing task. The purpose of this measure is to reduce further requests to the downstream system to relieve the immediate pressure. In this case, the module will notify the downstream system to perform necessary processing, such as capacity expansion operations or optimizing database queries, etc. Once it is confirmed that the downstream system has stabilized and is ready to reprocess requests, the module will restart the previously suspended task to ensure that all processing flows can continue smoothly.

[0121] If the number of failures exceeds the second preset threshold, it indicates a more serious situation, and simply pausing to process tasks may not be sufficient to relieve the system pressure. In this case, relevant business requests will be pushed to the message queue (MQ). This step is crucial for peak shaving, as it allows the system to temporarily store requests that cannot be processed immediately due to high traffic, thus avoiding a direct impact on downstream systems. The use of the message queue provides a buffering mechanism, enabling requests to be processed gradually according to the system's processing capacity, rather than overloading the system in a short period. Meanwhile, Redis is used as a tool for traffic splitting and flow control. Through its high-performance key-value storage ability, Redis can quickly sort and manage requests, achieving fast traffic splitting and flow control. This involves temporarily storing incoming requests and processing them according to priorities or other business rules to reduce the direct pressure on the core business system. In addition, the system directly reduces the load on downstream systems by configuring the concurrency parameter of downstream systems. This means dynamically adjusting the number of requests that downstream systems can process simultaneously based on the current system load and the pressure of external requests, thereby ensuring that downstream systems will not crash due to a sudden increase in the number of requests.

[0122] Through this comprehensive strategy, the traffic shaping and flow control module 103 can not only protect the system from the impact of sudden traffic at critical moments, but also optimize the overall processing flow through an intelligent adjustment mechanism, improving the system's adaptability to unstable external environments and overall business continuity. The design and implementation of this module provide strong technical support for the stable operation of the loan approval system, ensuring high efficiency and reliable services even in extreme situations.

[0123] In the intelligent loan approval process anomaly management and fault tolerance system, downstream systems refer to various technical components and services that play key roles in the data processing and approval process. These systems generally include, but are not limited to, database servers, application servers, and various software systems that handle specific business logics. They work together to ensure that loan applications can be effectively evaluated, approved, and relevant records and decisions can be accurately stored and executed.

[0124] For example, consider a loan approval system, whose downstream systems may include a core banking system that is responsible for processing account information, customer credit data, and loan transactions. In addition, there may be a risk management system dedicated to evaluating the risks of loans and deciding whether to approve them. These systems rely on a highly reliable database management system (DBMS) to store and query data, such as customers' financial records, historical transactions, and credit scores.

[0125] Furthermore, the traffic shaping and rate limiting module includes an automatic failover subsystem. When the failover subsystem detects that the number of failures of the downstream system exceeds a first preset threshold, it switches to a standby system, thereby ensuring the continuity of processing tasks and the high availability of the system.

[0126] In the intelligent loan approval process anomaly management and fault tolerance system, a key component of the traffic shaping and rate limiting module is the automatic failover subsystem. The main function of this subsystem is to ensure that when the main downstream system has problems, the system can seamlessly switch to a pre-configured standby system, thereby maintaining the continuity of the approval process and the high availability of the entire system.

[0127] The core function of the automatic failover subsystem is to monitor the health status and performance metrics of the downstream system. It mainly focuses on the number of failures of the downstream system, which may be caused by system overload, hardware failure, network problems, or other technical issues. To effectively perform the failover operation, the subsystem has a preset threshold. Once the number of failures of the downstream system exceeds this threshold, the automatic failover mechanism will be triggered.

[0128] The automatic failover subsystem continuously tracks the status and performance of the downstream system through real-time monitoring tools. This includes but is not limited to metrics such as error rate, response time, and system load.

[0129] Based on historical data and system performance standards, the technical team sets the trigger threshold for failover. This threshold is carefully calculated and adjusted according to the fault tolerance of the system and business requirements to ensure that the failover is triggered only when truly necessary, avoiding unnecessary system switches.

[0130] When the number of failures of the downstream system reaches or exceeds the preset threshold, the automatic failover subsystem will immediately start the standby system. This switching process is automatic and is designed with multiple checking and confirmation mechanisms to ensure the correctness and timeliness of the switch.

[0131] The standby system continuously synchronizes the data and configuration of the main system in an inactive state to ensure that it can seamlessly continue to execute all tasks when it needs to take over. The startup and operation of the standby system will also be monitored to ensure that its performance meets system requirements.

[0132] Failover events will be recorded in the system log, and the relevant technical team and management personnel will receive automatic notifications informing them that the failover has occurred and the current system status.

[0133] Furthermore, the traffic shaping and rate limiting module includes an adaptive threshold adjustment unit for dynamically adjusting the first preset threshold and the second preset threshold according to historical data and current system performance.

[0134] In the intelligent loan approval process anomaly management and fault-tolerant system, the adaptive threshold adjustment unit of the traffic shaping and flow limiting module is a key component. It is responsible for dynamically adjusting the system's ability to process requests to adapt to changing requirements and maintain system stability. The unit automatically adjusts the key thresholds for managing system load by analyzing historical data and monitoring the current system performance.

[0135] The main task of the adaptive threshold adjustment unit is to ensure that the system can automatically adjust its processing capacity in the face of different load conditions, prevent overload, and maximize processing efficiency. The two key parameters dynamically adjusted by this unit are the first preset threshold and the second preset threshold, and these thresholds directly affect how the system responds to an increase or decrease in traffic.

[0136] The unit first collects the necessary data from system operation, including but not limited to the number of request processing times, processing time, number of failures, system resource utilization, etc. These data not only reflect the current system state but also include historical performance data.

[0137] By analyzing historical data, the unit can identify patterns and trends in system load, such as peak periods, the most common bottleneck problems, etc. This information is an important basis for adjusting the thresholds.

[0138] At the same time, the unit monitors the current system performance in real time, including the speed of processing requests and the usage of resources. These real-time data help the unit determine whether the system is approaching its performance limit.

[0139] Based on the information obtained from historical data and real-time monitoring, the unit uses a predetermined algorithm to adjust the first preset threshold and the second preset threshold. These thresholds are directly related to when the system starts to reject new requests or push requests into the waiting queue, and when to initiate failover or resource expansion measures.

[0140] After the threshold is adjusted, the system automatically applies the new settings and continues to monitor the actual effect to confirm whether the adjustment has achieved the expected effect. The system also provides feedback to the administrator to allow further manual optimization.

[0141] Furthermore, the traffic shaping and flow limiting module uses the following formulas (3) and (4) to dynamically calculate and adjust the first preset threshold and the second preset threshold :

[0142]

[0143]

[0144] where is the new first preset threshold; is the new second preset threshold; is the current first preset threshold; is the current second preset threshold; and is the adjustment coefficient; determined based on the statistical analysis of system performance monitoring data; and represent the system performance data points at the past time points; and respectively represent the average value and standard deviation of the current system performance data; is the number of time points considered for calculating the average value and standard deviation.

[0145] In the intelligent loan approval process anomaly management and fault tolerance system, the traffic shaping and flow limiting module uses two mathematical formulas to dynamically calculate and adjust the first preset threshold and the second preset threshold. The design of these formulas aims to finely adjust the thresholds according to the current and historical performance data of the system, so as to optimize the system's response to changing loads and demands. The following details each component of these formulas and their calculation methods.

[0146] Formulas (3) and (4) are used to calculate the new first and second preset thresholds ( and ), which directly affect how the system responds to traffic peaks and processing capacity limitations.

[0147]

[0148] 1. is the new first preset threshold.

[0149] 2. is the current first preset threshold, obtained based on system real-time monitoring data.

[0150] 3. and are the adjustment coefficients, determined by the statistical analysis of system performance, and affect the amplitude and speed of threshold adjustment. System performance refers to the efficiency and effectiveness demonstrated by the system when executing tasks and processing data. It includes at least one of the following measurement metrics, such as:

[0151] Response time: The time required for the system to respond to user requests.

[0152] Processing speed: The number of requests processed by the system per unit time.

[0153] Error rate: The frequency of errors occurring when processing requests.

[0154] Resource utilization rate: The usage of system resources (such as CPU, memory, network bandwidth).

[0155] 4. Represent system performance data points for the past N time points.

[0156] 5. and are the average value and standard deviation of the current system performance data, obtained by performing statistical analysis on past performance data.

[0157] 6. is the number of time points considered, which are used to calculate the average value and standard deviation.

[0158] This formula uses the exponential function and the hyperbolic tangent function to smooth out abnormal fluctuations in performance data and adjusts the threshold according to the distribution of performance data.

[0159]

[0160] 1. is the new second preset threshold.

[0161] 2. is the current second preset threshold.

[0162] 3. and are adjustment coefficients, which affect the intensity and response period of threshold adjustment.

[0163] 4. represents different performance data points for the past N time points similar to .

[0164] This formula uses the sine function and the exponential decay function to respond to changes in system performance by adjusting the threshold, making the system's response more flexible and adaptable.

[0165] By precisely implementing and adjusting these formulas, the system can more effectively handle high-load situations, avoid overload and optimize resource usage, thereby improving the performance and reliability of the entire loan approval system.

[0166] The risk control abnormal stop-loss module 104 is used to monitor the approval results of the risk control system; calculate key indicators including the approval passing rate and the number of consecutive rejections; compare the calculated key indicators with the preset normal range; when the key indicators deviate from the preset normal range, suspend all loan approval processes and issue a warning.

[0167] The risk control and abnormal stop-loss module 104 plays a crucial monitoring and guarantee role in the intelligent loan approval process anomaly management and fault tolerance system. This module is specifically designed to ensure that the risk control in the entire loan approval process complies with established standards and expectations, thus preventing potential risks from causing significant losses.

[0168] The core function of this module is to monitor and analyze key risk control indicators in the loan approval system in real time, such as the approval passing rate and the number of consecutive rejections. These indicators are regarded as key parameters for measuring the risk control effect in the loan approval process. The passing rate directly reflects the looseness or strictness of the approval system, while the number of consecutive rejections may indicate whether the system is too conservative or there may be misjudgments.

[0169] The operation of the risk control and abnormal stop-loss module 104 begins with continuously collecting real-time data from the approval system. This data includes the approval results of each loan application, which are automatically recorded by the system and transmitted to the module for analysis. Advanced algorithms are set inside the module, which can calculate the current approval passing rate and the number of consecutive rejections based on the collected data, and compare these real-time calculation results with the preset normal range.

[0170] To ensure operability, the risk control and abnormal stop-loss module 104 presets thresholds for the normal operation range, which are set based on historical data analysis and risk management strategies. For example, if the normal approval passing rate is 70% to 85%, any measured value below or above this range will trigger the module's alarm system.

[0171] When key indicators such as the approval passing rate or the number of consecutive rejections deviate from the preset normal range, the module will automatically execute a series of predefined countermeasures. The most direct measure is to suspend all current loan approval processes, which is an emergency measure to prevent further risk accumulation. At the same time, the module will send a warning to the management department of the system, prompting them to intervene and conduct further analysis.

[0172] This automatic suspension and alarm mechanism ensures that the loan approval system can respond quickly when facing possible risk deviations and prevent problems from expanding. In this way, the risk control and abnormal stop-loss module 104 not only protects the lending institution from the impact of bad loans but also safeguards customers from the impact of wrong decisions.

[0173] In summary, the risk control and abnormal stop-loss module 104 provides a powerful risk management tool for the loan approval process through precise monitoring and intelligent warning systems, ensuring that the entire approval process is both efficient and safe.

[0174] Furthermore, the risk control exception stop-loss module includes an interactive risk control adjustment interface that allows administrators to adjust risk control strategy parameters based on real-time data and system recommendations, including real-time adjustment of the approval pass rate threshold and the rejection times threshold.

[0175] In the intelligent loan approval process exception management and fault tolerance system, a core function of the risk control exception stop-loss module is the included interactive risk control adjustment interface. This interface enables system administrators to adjust risk control strategy parameters according to real-time data and system recommendations, especially the approval pass rate threshold and the rejection times threshold. The implementation of this function not only enhances the adaptability and flexibility of the system but also improves the efficiency and accuracy of managing risk control strategies.

[0176] The interactive risk control adjustment interface is designed to allow administrators to directly view and modify key parameters of the risk control strategy on the interface in real time. These include:

[0177] Approval pass rate threshold: This is the minimum pass rate that determines whether a loan application can be automatically approved. Adjusting this threshold can affect the looseness or strictness of the system and thus affect the loan approval rate.

[0178] Rejection times threshold: This is a setting used to determine how many times a customer's application can be rejected within a certain period before the system automatically marks the customer's subsequent applications as high-risk or conducts special reviews.

[0179] The interface integrates a real-time data monitoring function that can display the current system's approval situation and relevant risk control indicators. These data include but are not limited to real-time loan application quantities, pass rates, rejection rates, and their changing trends.

[0180] Based on real-time data analysis and a preset risk control model, the system will propose recommended risk control strategy adjustments. For example, if the system detects an abnormal decrease in the pass rate, it may recommend increasing the pass rate threshold to avoid over-rejecting qualified loan applications.

[0181] Administrators can directly adjust the approval pass rate threshold and the rejection times threshold on the interface according to these recommendations or their own judgment. Adjustments are usually made through sliders or input boxes, providing a user-friendly operation method.

[0182] After any parameter adjustment, the system will immediately display the potential impact of the adjustment, such as the expected change in the pass rate. This helps administrators make more informed decisions.

[0183] After the adjustment is completed, a confirmation and possibly an approval process are required to ensure that the adjustment complies with the overall risk control strategy and compliance requirements.

[0184] Through this interactive risk control adjustment interface, system administrators can respond to various risk control issues arising during the loan approval process more flexibly and promptly, optimize the approval results, thereby improving customer satisfaction and reducing risks. The implementation of this function has significantly enhanced the operation efficiency of the system and the dynamic adaptability of risk control management.

[0185] Furthermore, the risk control abnormal stop-loss module has realized the interface docking with external credit rating agencies and automatically obtains the latest credit information of the applicant from the external credit rating agencies.

[0186] In the intelligent loan approval process anomaly management and fault tolerance system, the risk control abnormal stop-loss module has an important function: the interface docking with external credit rating agencies. This function enables the system to automatically obtain the latest credit information of the applicant, which is crucial for evaluating the applicant's loan eligibility and determining loan conditions. The following details the implementation methods and operation processes of this function.

[0187] The core of this function is that the system can access the data provided by external credit rating agencies in real time and obtain the latest credit scores and credit histories of loan applicants. In this way, the risk control module can make more accurate risk assessments based on the latest data, improving the quality and speed of loan approvals.

[0188] The implementation steps include:

[0189] 1. Interface integration:

[0190] The system docks with the systems of external credit rating agencies through pre-defined APIs. These APIs need to support efficient data query and response to ensure the instant update and transmission of data.

[0191] The design of the interface needs to consider the compatibility of data formats to ensure that the data received from external agencies can be correctly parsed and used by the risk control module.

[0192] 2. Data request and reception:

[0193] When a loan application is received by the system and enters the approval process, the risk control module automatically initiates a request to the external credit rating agency to query the applicant's credit score.

[0194] After receiving the data, the system needs to perform data verification to confirm the integrity and accuracy of the data and prevent incorrect information from affecting the approval results.

[0195] 3. Data processing and application:

[0196] The received credit information will be analyzed by the risk control module and matched with the internal risk control standards and models of the system to determine the applicant's credit risk level.

[0197] Based on the credit risk level, the system automatically adjusts the approval decision, such as adjusting the loan amount, interest rate, or directly deciding whether to approve the loan.

[0198] 4. Update and maintenance:

[0199] The system regularly synchronizes data with external credit rating agencies to ensure that the credit data used is always up-to-date.

[0200] It is necessary to continuously monitor the status and performance of the interface to ensure the stability and security of data exchange.

[0201] Through this function of docking with the interface of external credit rating agencies, the intelligent loan approval process exception management and fault tolerance system can more dynamically and accurately evaluate the risks of loan applications, thereby improving the efficiency and security of the entire loan approval process. This not only improves the quality of customer service but also strengthens the system's risk management capabilities.

[0202] Although this application is disclosed above with preferred embodiments, it is not intended to limit this application. Any person skilled in the art can make possible changes and modifications without departing from the spirit and scope of this application. Therefore, the protection scope of this application shall be subject to the scope defined by the claims of this application.

Claims

1. An intelligent loan approval process exception management and fault tolerance system, characterized in that: include: An interception and degradation module is configured with a service degradation switch, wherein the service degradation switch is activated when a high-volume loan application is detected; After the service downgrade switch is activated, the loan application is intercepted or released according to the preset product code, partner code and interception ratio parameters; for the first loan application that is intercepted, its product code, partner code, business application type, business application number, business urgency, interception time, whether approval has been started, whether approval has been completed, and approval completion time are recorded in the service downgrade task table; Extract recorded loan applications from the service degradation task table at regular intervals, and asynchronously execute the approval process based on the business urgency and interception time of the recorded loan applications; After the approval is completed, the corresponding approval result in the service downgrade task table is updated; The timeout downgrade strategy module includes a timeout downgrade service switch and a timeout salvage thread; when the timeout downgrade service switch is turned on, the timeout salvage thread regularly checks the second loan applications in the system that have not been approved on time; determines a third loan application with a high degree of urgency among the second loan applications that have not been approved on time according to the product code, the partner code and the current loan approval time; increases the business urgency in the third loan application, and saves the third loan application with increased business urgency into the service downgrade task table; Traffic shaping and current limiting module, used to count the number of failures of external data interfaces to request downstream systems within a specified time; If the number of failures is greater than the first preset threshold and less than the second preset threshold, suspend the current processing task, notify the downstream system to process, and restart the suspended current processing task after the downstream system completes the processing; If the number of failures is greater than a second preset threshold, the current processing task is pushed to the message queue, and the pressure on the downstream system is relieved by configuring the concurrency parameters of the downstream system; The risk control abnormal stop loss module is used to monitor the approval results of the risk control system; calculate key indicators including approval pass rate and number of consecutive rejections; compare the calculated key indicators with the preset normal range; when the key indicators deviate from the preset normal range, suspend all loan approval processes and issue a warning.

2. The intelligent loan approval process abnormality management and fault tolerance system according to claim 1 is characterized in that: The interception degradation module also includes an adaptive interception ratio adjustment unit for real-time monitoring of system resource utilization and request processing time; When the system resource utilization exceeds the preset threshold or the average request processing time increases significantly, the interception ratio is automatically increased; When the system resource utilization rate drops to a safe level and the average request processing time returns to normal, the interception ratio is gradually reduced; the interception ratio is adjusted using a progressive strategy to avoid drastic fluctuations in system load.

3. The intelligent loan approval process abnormality management and fault tolerance system according to claim 2 is characterized in that: The adaptive interception ratio adjustment unit includes a dynamic adjustment algorithm, which adjusts the interception ratio in real time according to the following formula (1): in, is the new interception ratio; is the current interception ratio; is the actual system utilization; is the target system utilization; is the actual request processing time; is the target request processing time; and is the adjustment factor, which is used to adjust the impact of system utilization and processing time on the interception ratio; It is the time when the system resource utilization exceeds the preset utilization threshold. The time difference to the current execution time of the dynamic adjustment algorithm; and is a gradual adjustment function, which is implemented using the following formulas (2) and (3): in, and is the adjustment coefficient.

4. The intelligent loan approval process abnormality management and fault tolerance system according to claim 1 is characterized in that: The interception and degradation module includes an early warning unit, which is used to notify the system administrator via email or SMS when it is predicted that the system is about to reach the load limit, and automatically start additional cloud resources to cope with the upcoming high traffic, thereby avoiding system overload.

5. The intelligent loan approval process abnormality management and fault tolerance system according to claim 1 is characterized in that: The timeout downgrade strategy module includes a monitoring interface for displaying all loan applications that have not been approved on time and their business urgency, wherein the monitoring interface allows an administrator to manually adjust the business urgency.

6. The intelligent loan approval process abnormality management and fault tolerance system according to claim 1 is characterized in that: The timeout service degradation switch can be automatically activated based on real-time weather or traffic accident data to identify approval delays that may be caused by emergencies.

7. The intelligent loan approval process abnormality management and fault tolerance system according to claim 1 is characterized in that: The traffic shaping and limiting module includes an automatic failover subsystem, which switches to a backup system when detecting that the number of failures of the downstream system exceeds a first preset threshold, thereby ensuring the continuity of processing tasks and the high availability of the system.

8. The intelligent loan approval process abnormality management and fault tolerance system according to claim 1 is characterized in that: The traffic shaping and limiting module includes an adaptive threshold adjustment unit, which is used to dynamically adjust the first preset threshold and the second preset threshold according to historical data and current system performance.

9. The intelligent loan approval process abnormality management and fault tolerance system according to claim 1 is characterized in that: The risk control abnormal stop loss module includes an interactive risk control adjustment interface, which allows administrators to adjust risk control strategy parameters based on real-time data and system recommendations, including real-time adjustment of approval pass rate thresholds and rejection number thresholds.

10. The intelligent loan approval process abnormality management and fault tolerance system according to claim 1 is characterized in that: The risk control abnormal stop loss module realizes interface connection with an external credit scoring agency and automatically obtains the latest credit information of the applicant from the external credit scoring agency.

Images