Risk monitoring management system and method for distributed computing

Abstract

The present application relates to a risk monitoring management system and method of distributed computing, comprising a risk monitoring system and a distributed risk information computing system; the distributed risk information computing system comprises a dispatching module, an authority management module and an execution module; the execution module comprises a plurality of executors; the dispatching module receives a computing request and dispatches a computing task to a suitable executor in the execution module and sends risk index information to the risk monitoring system; the authority management module authorizes and manages the dispatching module; the suitable executor in the execution module analyzes the computing task, if the computing task is a first type task, carries out risk analysis and calculation based on the computing task and returns risk index information; if the computing task is a second type task, splits the computing task and dispatches the split computing task to other executors in the execution module. The computing system and the monitoring system are separated, have good independence, have small mutual influence, adopt distributed computing, and the computing rate is significantly improved.

Description

Technical Field

[0001] This invention relates to the technical field of risk monitoring in nuclear power plants, and more specifically, to a distributed computing risk monitoring management system and method. Background Technology

[0002] Currently, nuclear power plant risk monitoring systems are implemented online using software systems. Due to business development needs, these software systems have undergone several functional revisions and upgrades. However, previous revisions and upgrades focused primarily on localized optimizations of internal software modules and algorithms, without interaction with the nuclear power plant's existing data systems. Overall, in current online risk monitoring, the computing system and the risk monitor are coupled together, lacking independence and causing mutual interference. Furthermore, when there are many data points to be calculated, the calculation speed is very low because calculations can only be performed one by one. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a risk monitoring and management system and method for distributed computing.

[0004] The technical solution adopted by the present invention to solve its technical problem is: to construct a distributed computing risk monitoring and management system, including: a risk monitoring system and a distributed risk information computing system that interacts with the risk monitoring system;

[0005] The distributed risk information calculation system includes: a distribution module, a permission management module, and an execution module; the execution module includes: multiple executors;

[0006] The dispatch module is used to receive the calculation request sent by the risk monitoring system, generate a calculation task according to the calculation request, and dispatch the calculation task to the executor adapted in the execution module; the dispatch module is also used to receive the risk indicator information returned by the execution module and send the risk indicator information to the risk monitoring system;

[0007] The permission management module is used to manage and authorize the permissions of the dispatch module;

[0008] The executor adapted in the execution module is used to receive the computing task and analyze it. If the computing task is a first type of task, the executor adapted in the execution module performs risk analysis calculation based on the computing task and returns the risk indicator information. If the computing task is a second type of task, the executor adapted in the execution module splits the computing task and dispatches the split computing task to other executors in the execution module through the dispatch module. The other executors in the execution module then perform risk analysis calculation based on the split computing task and return the risk indicator information.

[0009] In the distributed computing risk monitoring and management system of the present invention, the permission management module includes: a running monitoring unit and a permission management unit;

[0010] The operation monitoring unit monitors the operation status of the permission management unit in real time, and restarts the permission management unit when the operation status of the permission management unit is abnormal.

[0011] The permission management unit synchronizes the validity period of the permissions of the dispatch module and responds to the information requests of the dispatch module, and performs permission management and authorization for the dispatch module according to the validity period of the permissions of the dispatch module and the information requests.

[0012] In the distributed computing risk monitoring and management system of the present invention, the dispatch module includes: multiple dispatchers;

[0013] The permission management module synchronizes the validity period of permissions for each dispatcher and responds to the information requests of each dispatcher. Based on the information requests and validity periods of permissions for each dispatcher, it determines the master dispatcher and authorizes the master dispatcher.

[0014] In the distributed computing risk monitoring and management system of the present invention, the master dispatcher is used to generate a computing task according to the computing request when receiving a computing request sent by the risk monitoring system, determine the appropriate executor in the execution module according to the load of each executor, and send the computing task to the appropriate executor.

[0015] In the distributed computing risk monitoring and management system described in this invention, each executor includes: multiple state machines;

[0016] When each executor receives a computation task sent by the dispatch module, it analyzes the computation task. If the computation task is a first type of task, the executor selects the appropriate state machine from its multiple state machines to execute the computation task. If the computation task is a second type of task, the executor splits the computation task and dispatches the split computation tasks to other executors through the dispatch module.

[0017] In the distributed computing risk monitoring and management system described in this invention, the distributed risk information computing system uses the cut set method and the re-solution method to perform computational task analysis.

[0018] In the distributed computing risk monitoring and management system described in this invention, the risk monitoring system and the distributed risk information computing system interact using a message queue.

[0019] The distributed computing risk monitoring and management system of the present invention also includes: a database;

[0020] The database is connected to both the risk monitoring system and the distributed risk information calculation system. The database is used to store data from the risk monitoring system and the distributed risk information calculation system, and to provide data to both the risk monitoring system and the distributed risk information calculation system.

[0021] In the distributed computing risk monitoring and management system described in this invention, the risk monitoring system and the distributed risk information computing system are independent of each other.

[0022] This invention also provides a risk monitoring and management method for distributed computing, applied to the aforementioned risk monitoring and management system for distributed computing. The risk monitoring and management system for distributed computing includes: a risk monitoring system and a distributed risk information computing system that interacts with the risk monitoring system; the distributed risk information computing system includes: a distribution module, a permission management module, and an execution module; the execution module includes: multiple executors.

[0023] The risk monitoring and management method for distributed computing includes the following steps:

[0024] The risk monitoring system sends a calculation request to the distributed risk information calculation system;

[0025] The permission management module manages and authorizes the dispatch module;

[0026] The dispatch module receives the computation request, generates a computation task based on the computation request, and dispatches the computation task to the appropriate executor in the execution module.

[0027] The executor adapted in the execution module receives the computation task and analyzes it. If the computation task is a first type of task, the executor adapted in the execution module performs risk analysis calculations based on the computation task and returns the risk indicator information. If the computation task is a second type of task, the executor adapted in the execution module splits the computation task and dispatches the split computation tasks to other executors in the execution module through the dispatch module. The other executors in the execution module then perform risk analysis calculations based on the split computation tasks and return the risk indicator information.

[0028] The dispatch module receives the risk indicator information returned by the execution module and sends the risk indicator information to the risk monitoring system.

[0029] The risk monitoring and management system and method for distributed computing implemented in this invention have the following beneficial effects: It includes a risk monitoring system and a distributed risk information calculation system; the distributed risk information calculation system includes a dispatch module, a permission management module, and an execution module; the execution module includes multiple executors; the dispatch module receives calculation requests and dispatches the calculation tasks to the appropriate executors in the execution module and sends risk indicator information to the risk monitoring system; the permission management module authorizes and manages the dispatch module; the appropriate executors in the execution module analyze the calculation tasks, and if the calculation task is a first-type task, perform risk analysis calculations based on the calculation task and return risk indicator information; if the calculation task is a second-type task, split the calculation task and dispatch the split calculation tasks to other executors in the execution module. This invention separates the calculation system from the monitoring system, ensuring good independence and minimal mutual influence; the use of distributed computing significantly improves the calculation speed. Attached Figure Description

[0030] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings:

[0031] Figure 1 This is a schematic diagram of the distributed computing risk monitoring and management system provided by the present invention;

[0032] Figure 2 This is a schematic diagram of the permission management module provided by the present invention;

[0033] Figure 3 This is a schematic diagram of the distribution module provided by the present invention;

[0034] Figure 4 This is a schematic diagram of the execution module provided by the present invention;

[0035] Figure 5 This is an algorithm flowchart of the distributed risk information calculation system provided by the present invention;

[0036] Figure 6 This is a flowchart illustrating the risk monitoring and management method for distributed computing provided by the present invention. Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] Figure 1 This invention illustrates a preferred embodiment of a distributed computing risk monitoring and management system. This system is based on distributed computing using multiple solution algorithms. Specifically, it distributes problems requiring massive computing power into multiple smaller parts, assigns these parts to multiple executors for processing, and finally combines the results from each executor to obtain the final result. This invention is based on distributed network storage technology, which disperses data across multiple independent machines. The risk monitoring and management system of this invention employs a scalable system architecture, utilizing multiple storage servers to share the storage load and locate stored information. This not only solves the bottleneck problem of a single storage server in traditional centralized storage systems but also improves the system's reliability, availability, and scalability.

[0039] Specifically, such as Figure 1 As shown, in this embodiment, the distributed computing risk monitoring and management system includes: a risk monitoring system 10 and a distributed risk information computing system 20 that interacts with the risk monitoring system 10.

[0040] In this embodiment, the risk monitoring system 10 includes a risk monitor, which monitors the risks of the nuclear power plant and sends calculation requests to the distributed risk information computing system 20. The calculation request is a request, message, or instruction used to trigger the distributed risk information computing system 20 to perform risk information analysis and calculation. Additionally, in this embodiment, the risk monitoring system 10 also receives risk indicator information returned by the distributed risk information computing system 20 and performs related processing based on the received risk indicator information, such as displaying the risk indicator information, making risk predictions based on the risk indicator information, and generating risk warnings based on the risk indicator information. The risk indicator information may include, but is not limited to, minimum cut sets, the probability of a top event occurring, the frequency of a top event occurring, and importance.

[0041] In this embodiment, the distributed risk information calculation system 20 can also be defined as a calculation engine, which may include: a dispatch module 21, a permission management module 22, and an execution module 23.

[0042] In this embodiment, the permission management module 22 is used to manage and authorize permissions for the dispatch module 21. Optionally, in this embodiment, as... Figure 2 As shown, the permission management module 22 includes: a running monitoring unit and a permission management unit.

[0043] In this embodiment, the operation monitoring unit monitors the operation status of the permission management unit in real time, and restarts the permission management unit when the operation status of the permission management unit is abnormal.

[0044] Specifically, in this embodiment, the operation monitoring unit is a watchdog, whose function is to ensure that the core program of the permission management unit can run stably and continuously during the time when the computer cluster of the distributed risk information computing system 20 is continuously effective. When the permission management unit is abnormal due to various operating states (such as the abnormal exit of the permission management unit process), the operation monitoring unit can restart the permission management unit, thereby ensuring the stability and effectiveness of the permission management unit.

[0045] In this embodiment, the permission management unit synchronizes the validity period of permissions for the dispatch module 21 and responds to information requests from the dispatch module 21. It then manages and authorizes permissions for the dispatch module 21 based on its validity period and information requests. Specifically, in this embodiment, the permission management unit acts as a lessee, its main function being to synchronize the validity period of permissions for all dispatchers in the dispatch module 21 and respond to all dispatchers' information requests. This allows it to determine the master dispatcher based on each dispatcher's validity period and information requests, and to authorize the master dispatcher, ensuring that only one dispatcher (i.e., the master dispatcher) is authorized to dispatch permissions at any given time. It should be noted that within a valid local area network, only one permission management unit can exist in a computer cluster.

[0046] In this embodiment, the dispatch module 21 is used to receive the calculation request sent by the risk monitoring system 10, generate a calculation task according to the calculation request, and dispatch the calculation task to the appropriate executor in the execution module 23. The dispatch module 21 is also used to receive the risk indicator information returned by the execution module 23 and send the risk indicator information to the risk monitoring system 10. That is, in this embodiment, when the distributed risk information calculation system 20 is running normally, the dispatch module 21 undertakes the communication work between the entire cluster, the risk monitoring system 10, and the execution module 23.

[0047] Specifically, such as Figure 3As shown, in this embodiment, the dispatch module 21 includes: multiple dispatchers (dispatcher_0, dispatcher_1, ..., dispatcher_n); the permission management module 22 synchronizes the permission validity time of each dispatcher and responds to the information request of each dispatcher, and determines the master dispatcher and authorizes the master dispatcher according to the information request and permission validity time of each dispatcher.

[0048] In this embodiment, the validity period of a dispatcher's permission is the time during which the dispatcher can be used after being authorized. That is, the dispatcher only has dispatching permission within the validity period; after the validity period expires, the dispatcher loses its dispatching permission. For example, among multiple dispatchers, once any dispatcher is authorized and has dispatching permission, that dispatcher becomes the master dispatcher. During its validity period, the master dispatcher is responsible for communication between the entire cluster, the risk monitoring system 10, and the execution module 23. Specifically, during its validity period, when the master dispatcher receives a computation request from the risk monitoring system 10, it generates a computation task based on the request, determines the appropriate executor in the execution module 23 based on the load of each executor, and sends the computation task to the appropriate executor. The load of the executor can be the computational capacity of the executor during the initialization of the distributed risk information computation system 20, or it can be the current task status, current CPU capacity, current CPU utilization, and current resource status of the executor during operation. By determining the appropriate executor (that is, the executor that can perform the computation task) based on the load of each executor, the computational power can be utilized efficiently, and the overall response time of the computation request can be accelerated.

[0049] In this embodiment, each dispatcher periodically sends an information request to the permission management unit to obtain authorization. When the primary dispatcher reaches the authorization validity period, the permission management unit will assign it to other available dispatchers, allowing them to become the new primary dispatcher, thereby restoring the command response of the computer cluster. It should be noted that in this embodiment, a first-come, first-served approach can be used to determine the primary dispatcher. For example, dispatcher_0, dispatcher_1, dispatcher_2, ..., dispatcher_n will all send information requests to the permission management unit. When the permission management unit receives the information requests from dispatcher_0, dispatcher_1, ..., dispatcher_n, it determines the authorization validity period for each dispatcher based on their respective requests. Simultaneously, it determines the dispatcher that sent the information request first (denoted as dispatcher_0) and analyzes whether dispatcher_0 is available. If dispatcher_0 is available, it is designated as the primary dispatcher. If dispatcher_0 is unavailable or malfunctioning, the next dispatcher is selected and determined sequentially according to the time sequence. It should be noted that the primary dispatcher has priority dispatching rights. That is, in the next cycle, when the primary dispatcher and other dispatchers send information requests to the permission management unit, if the primary dispatcher's permission validity period has not expired and the primary dispatcher is available, then the primary dispatcher will continue to be selected as the primary dispatcher, thereby avoiding frequent changes in the primary dispatcher.

[0050] It should be noted that, in this embodiment, the executors in the dispatch module 21 can also be categorized by function, specifically including: a first type of executor (i.e., RormWorker) and a second type of executor (i.e., LpsaWorker). The first type of executor is used to calculate the risk indicator information of the risk monitoring system 10; that is, the dispatchers mentioned above (dispatcher_0, dispatcher_1, dispatcher..., dispatcher_n) are all first type executors in this embodiment. The second type of executor is used to generate the PSA risk model (i.e., the probabilistic safety assessment risk model).

[0051] Furthermore, in this embodiment, the master dispatcher also has management functions for the actuators. For example, when an actuator malfunctions or fails, the master dispatcher analyzes the abnormal or faulty condition of the actuator and sends a restart signal according to the severity of the abnormal or faulty condition. Alternatively, it can transfer the computational task of the actuator to other actuators for execution.

[0052] In this embodiment, the main function of the execution module 23 is to undertake the tasks of computation and data synchronization. Specifically, in this embodiment, the executor adapted in the execution module 23 is used to receive and analyze the computation tasks. If the computation task is a first type of task, the executor adapted in the execution module 23 performs risk analysis calculations based on the computation task and returns risk indicator information. If the computation task is a second type of task, the executor adapted in the execution module 23 splits the computation task and dispatches the split computation tasks to other executors in the execution module 23 through the dispatch module 21. The other executors in the execution module 23 then perform risk analysis calculations based on the split computation tasks and return risk indicator information.

[0053] In this embodiment, the first type of task refers to a computational task that is relatively simple (e.g., with a small computational load), and the second type of task refers to a computational task that is relatively complex (e.g., with a large computational load). The computational load can be determined based on the actual computational task, and this invention does not impose a specific limitation.

[0054] like Figure 4 As shown, in this embodiment, the execution module 23 includes multiple executors (executor_0, executor_1, executor..., executor_n). Each executor includes multiple state machines. When each executor receives a computation task sent by the dispatch module 21, it analyzes the computation task. If the computation task is a first type of task, the executor selects an appropriate state machine from its multiple state machines to execute the computation task. If the computation task is a second type of task, the executor splits the computation task and dispatches the split computation tasks to other executors through the dispatch module 21.

[0055] Specifically, when the main dispatcher receives a calculation request from the risk monitoring system 10, it generates a corresponding calculation task based on the request. Simultaneously, the main dispatcher determines the appropriate executor based on the load of each executor and dispatches the generated calculation task to the appropriate executor (which can be one or more). Upon receiving the calculation task, the appropriate executor first analyzes it to determine its type. If the task is of type 1, the appropriate executor selects a suitable state machine on its current node to execute the task, obtains the corresponding risk indicator information, and returns it to the main dispatcher, which then returns it to the risk monitoring system 10. If the task is of type 2, the appropriate executor first splits the task, then returns the split task to the main dispatcher, which then forwards it to other executors. These other executors perform the calculations, obtain the corresponding risk indicator information, and return it to the main dispatcher, which then returns it to the risk monitoring system 10. Of course, it is understandable that if the computation task is a second type of task, the adapted executor can split the computation task and perform a portion of the split computation task on itself, while sending the remaining computation task to the main dispatcher, which will then transfer the remaining computation task to other executors for computation.

[0056] In this embodiment, the multiple state machines equipped in each actuator may include, but are not limited to, an active state machine, a failed state machine, and a standby state machine. It should be noted that in this embodiment, the actuator can be a computing terminal or a computing server. Similarly, the state machine can also be a computing terminal or a computing server.

[0057] Furthermore, such as Figure 1 As shown in this embodiment, the distributed computing risk monitoring and management system further includes a database 30. The database 30 is connected to both the risk monitoring system 10 and the distributed risk information computing system 20. The database 30 is used to store data from both the risk monitoring system 10 and the distributed risk information computing system 20, and to provide data to both systems respectively.

[0058] In this embodiment, the distributed risk information calculation system 20 employs the cut-set method and the re-solution method for calculation and analysis. The cut-set method uses minimal cut sets to update the calculation, while the re-solution method uses an event tree / fault tree model to solve the problem. By using the cut-set method and the re-solution method, appropriate methods can be selected according to set rules in different scenarios (especially when planning risk calculations) to quickly obtain the required analysis results.

[0059] In this embodiment, the risk monitoring system 10 and the distributed risk information computing system 20 interact using a message queue, wherein the risk monitoring system 10 and the distributed risk information computing system 20 are independent of each other.

[0060] In this embodiment, as Figure 5 As shown, the distributed risk information computing system 20 can be divided into four layers in terms of algorithm flow. The first layer is the data loading layer, which serves the loading of the memory model within the distributed risk information computing system 20, converting the relational model of the database 30 into a memory model (i.e., converting the relational model of the database 30 into a model object package through the SQL (Structured Query Language) interface package to obtain the memory model). The second layer is the communication layer, which provides interfaces and responses for internal or external use within the cluster (i.e., using MQ (Message Queue) interface packages to implement interfaces and responses for internal or external use within the cluster, forming corresponding communication packets). The third layer is the computation execution layer, which consumes the specified computation commands propagated in the cluster and performs execution operations with the loaded memory objects as the execution targets (i.e., performing execution operations through computation packages with the loaded corresponding memory models as the execution targets). The fourth layer is the cluster scheduling layer, which is responsible for receiving relevant instructions sent by the risk monitoring system 10, and for maintaining the stable availability of the distributed risk information computing system 20 while using appropriate load balancing scheduling strategies to allow suitable executors to execute the commands from the risk monitoring system 10. Figure 5 The distributed component in the code is the execution module 23, which sets up multiple executors.

[0061] In this embodiment, each executor, during operation, needs to send heartbeats to the main dispatcher to maintain network relationships, and also needs to respond to computing tasks assigned by the main dispatcher at any time. To flexibly balance the workload and computational load within an executor, the possibility of starting multiple computing threads to handle computing tasks is retained.

[0062] Furthermore, in this embodiment, each executor directly relies on a message queue for message maintenance (i.e., after receiving a computation task dispatched by the master dispatcher, each executor queues the received computation tasks), and then sequentially converts each computation task into a corresponding trigger event, which is sent to the corresponding state machine for execution. After the state machine completes its calculation, it directly returns a message through the message queue communication interface. Each executor should contain at least one usable state machine to consume relevant commands from the distributed cluster. On the other hand, each executor is designed to support multiple computation state machines simultaneously; this system uses aggregation relationships.

[0063] The distributed risk information calculation system 20 of this invention adopts a distributed approach, with the dispatch module 21 handling communication between the entire cluster and the front-end and executors. When a calculation task arrives, it can be broken down into multiple tasks and dispatched to multiple executors for calculation.

[0064] In the original design, the computing system was integrated with the risk monitoring system 10. In this invention, the distributed risk information computing system 20 and the risk monitoring system 10 are separated using message queues and can be deployed separately. The distributed risk information computing system 20 and the risk monitoring system 10 are decoupled, and the independence of each module is relatively good, with little mutual influence.

[0065] When optimizing calculations for multiple points, multiple machines can be used for the calculations. Daily planning at nuclear power plants often involves changes in the status of multiple devices at different times. Different combinations of device statuses can lead to changes in the risk values ​​of the nuclear power plant, thus requiring independent calculations. When there are many points to calculate, these points can be distributed across multiple machines, significantly improving computational efficiency.

[0066] The distributed risk information computing system 20 of this invention can achieve large-scale parallel computing of planning optimization tasks. Based on this, it employs thread pool technology on individual computing nodes to split large-scale event trees and long sequences, further improving computational parallelism. Moreover, it combines the re-solution method and the cut-set method, improving the efficiency of minimum cut-set solution while ensuring computational correctness. In addition, the cut-set method utilizes a trie data structure for optimization during the solution process, further improving computational efficiency and providing a new and efficient method for assessing nuclear power plant planning risks.

[0067] The present invention also provides a risk monitoring and management method for distributed computing, wherein the risk monitoring and management method for distributed computing is applied to the risk monitoring and management system for distributed computing disclosed in the embodiments of the present invention.

[0068] Specifically, such as Figure 6 As shown, in this embodiment, the risk monitoring and management method for distributed computing includes the following steps:

[0069] Step S601: The risk monitoring system 10 sends a calculation request to the distributed risk information calculation system 20.

[0070] Step S602: The permission management module 22 manages and authorizes the dispatch module 21.

[0071] Step S603: The dispatch module 21 receives the calculation request, generates a calculation task according to the calculation request, and dispatches the calculation task to the executor adapted in the execution module 23.

[0072] It should be noted that steps S602 and S603 are not in any particular order. For example, step S602 can be executed before or after step S603.

[0073] In step S604, the executor adapted in execution module 23 receives the calculation task and analyzes it. If the calculation task is a first type of task, the executor adapted in execution module 23 performs risk analysis calculation based on the calculation task and returns risk indicator information. If the calculation task is a second type of task, the executor adapted in execution module 23 splits the calculation task and dispatches the split calculation task to other executors in execution module 23 through dispatch module 21. The other executors in execution module 23 then perform risk analysis calculation based on the split calculation task and return risk indicator information (i.e., perform risk indicator information analysis calculation).

[0074] Step S605: The dispatch module 21 receives the risk indicator information returned by the execution module 23 and sends the risk indicator information to the risk monitoring system 10.

[0075] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0076] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0077] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0078] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They do not limit the scope of protection of the present invention. All equivalent changes and modifications made within the scope of the claims of the present invention should fall within the scope of the claims of the present invention.

Claims

1. A distributed computing-based risk monitoring and management system, characterized in that, include: A risk monitoring system and a distributed risk information computing system that interacts with the risk monitoring system; The distributed risk information calculation system includes: a distribution module, a permission management module, and an execution module; the execution module includes: multiple executors; the distribution module includes: multiple dispatchers; The dispatch module is used to receive the calculation request sent by the risk monitoring system, generate a calculation task according to the calculation request, and dispatch the calculation task to the executor adapted in the execution module; the dispatch module is also used to receive the risk indicator information returned by the execution module and send the risk indicator information to the risk monitoring system; The permission management module is used to manage and authorize permissions for the dispatch module; the permission management module includes a running monitoring unit and a permission management unit, and a permission management unit is set up in a computer cluster of a local area network; the running monitoring unit is used to monitor the running status of the permission management unit in real time, and to restart the permission management unit when the running status of the permission management unit is abnormal; The permission management module is used to synchronize the validity period of permissions for each dispatcher and respond to the information requests of each dispatcher. It determines the master dispatcher based on the information request and the validity period of permissions for each dispatcher and authorizes the master dispatcher. If the validity period of permissions for the master dispatcher has not expired and the master dispatcher is running normally, then the master dispatcher is selected as the new master dispatcher. If the validity period of permissions for the master dispatcher expires, then any dispatcher running normally is selected as the new master dispatcher. The main dispatcher is used to generate a computing task according to the computing request when it receives a computing request sent by the risk monitoring system, determine the appropriate executor in the execution module according to the load of each executor, and send the computing task to the appropriate executor. The executor adapted in the execution module is used to receive the computing task and analyze it. If the computing task is a first type of task, the executor adapted in the execution module performs risk analysis calculation based on the computing task and returns the risk indicator information. If the computing task is a second type of task, the executor adapted in the execution module splits the computing task and dispatches the split computing tasks to other executors in the execution module through the dispatch module. The other executors in the execution module then perform risk analysis calculation based on the split computing tasks and return the risk indicator information. Wherein, if the calculation task is a second type of task, the individual computing node of the distributed risk information calculation system adopts thread pool technology to split the event tree or sequence; the distributed risk information calculation system adopts the cut set method and the re-solution method to perform risk analysis calculation, and optimizes the data structure of the trie in the process of solving the cut set method.

2. The distributed computing risk monitoring and management system according to claim 1, characterized in that, Each actuator includes: multiple state machines; When each executor receives a computation task sent by the dispatch module, it analyzes the computation task. If the computation task is a first type of task, the executor selects the appropriate state machine from its multiple state machines to execute the computation task. If the computation task is a second type of task, the executor splits the computation task and dispatches the split computation tasks to other executors through the dispatch module.

3. The distributed computing risk monitoring and management system according to claim 1, characterized in that, The distributed risk information calculation system uses the cut set method and the re-solution method to perform calculation and analysis tasks.

4. The distributed computing risk monitoring and management system according to claim 1, characterized in that, The risk monitoring system and the distributed risk information calculation system interact using a message queue.

5. The distributed computing risk monitoring and management system according to any one of claims 1-4, characterized in that, Also includes: database; The database is connected to both the risk monitoring system and the distributed risk information calculation system. The database is used to store data from the risk monitoring system and the distributed risk information calculation system, and to provide data to both the risk monitoring system and the distributed risk information calculation system.

6. The distributed computing risk monitoring and management system according to claim 1, characterized in that, The risk monitoring system and the distributed risk information calculation system are independent of each other.

7. A risk monitoring and management method for distributed computing, applied to the risk monitoring and management system for distributed computing as described in any one of claims 1-6, characterized in that, The distributed computing risk monitoring and management system includes: a risk monitoring system and a distributed risk information computing system that interacts with the risk monitoring system; the distributed risk information computing system includes: a distribution module, a permission management module, and an execution module; the execution module includes: multiple executors; The risk monitoring and management method for distributed computing includes the following steps: The risk monitoring system sends a calculation request to the distributed risk information calculation system; The permission management module manages and authorizes the dispatch module; The dispatch module receives the computation request, generates a computation task based on the computation request, and dispatches the computation task to the appropriate executor in the execution module. The executor adapted in the execution module receives the computation task and analyzes it. If the computation task is a first type of task, the executor adapted in the execution module performs risk analysis calculations based on the computation task and returns the risk indicator information. If the computation task is a second type of task, the executor adapted in the execution module splits the computation task and dispatches the split computation tasks to other executors in the execution module through the dispatch module. The other executors in the execution module then perform risk analysis calculations based on the split computation tasks and return the risk indicator information. The dispatch module receives the risk indicator information returned by the execution module and sends the risk indicator information to the risk monitoring system.

Images