Distributed system for solving large-scale mathematical optimization problem and implementation method

Abstract

The invention relates to a distributed system for solving a large-scale mathematical optimization problem and an implementation method, and aims to improve the calculation efficiency and the solving capability so as to meet the requirements of enterprises on efficient operation and accurate decision-making under the background of digitization, globalization and double carbon. According to the system, a complex optimization problem is decomposed into a plurality of small sub-problems, and the sub-problems are processed in parallel on a plurality of computing nodes, so that the solving process is accelerated. The system also adopts an innovative synchronization and communication mechanism to ensure the data consistency and calculation accuracy among the nodes. In addition, the invention further provides a self-adaptive load balancing strategy so as to optimize resource allocation and further improve the overall calculation efficiency. According to the method, data analysis, knowledge discovery and interdisciplinary research innovation are promoted while high-performance computing power is provided, and the method has important significance on improvement of enterprise competitiveness and scientific research.

Description

Technical Field

[0001] The present invention relates to the technical field of distributed mathematical optimization solution technology, and specifically to a distributed system and implementation method for solving large-scale mathematical optimization problems. Background Art

[0002] In today's era of digitalization and globalization, companies are not only facing unprecedented market expansion opportunities, but also facing the dual challenges of increasing competitive pressure and environmental protection. This transformation requires companies to make more refined and intelligent operational decisions to cope with the complex and changing market environment and strict environmental protection standards.

[0003] In this context, large-scale mathematical optimization problems have become the key for enterprises to achieve accurate decision-making, optimize resource allocation, and reduce energy consumption and emissions. However, with the explosive growth of data volume and the rapid increase in problem complexity, traditional stand-alone mathematical optimization solvers are unable to cope with the processing speed, computing power, and solution accuracy.

[0004] The rise of distributed computing technology has provided strong technical support for solving this problem. By using multiple computers to work together, distributed solvers can significantly enhance the utilization of computing resources and improve the efficiency and accuracy of solving large-scale mathematical optimization problems. However, existing distributed solvers are still limited in flexibility, ease of use, and scalability, and it is difficult to fully meet the needs of various industries for highly customized and fast-response decision-making systems.

[0005] Therefore, this technology aims to develop an efficient, flexible and highly scalable distributed mathematical optimization solution system and method, aiming to break through the bottleneck of existing technologies and provide more accurate and efficient decision support for various industries. Summary of the invention

[0006] In order to solve the problems of the prior art, the present invention provides a distributed system and implementation method for solving large-scale mathematical optimization problems. The system can not only process ultra-large-scale data sets and solve complex optimization problems, but also be flexibly configured and expanded according to the needs of specific application scenarios.

[0007] In order to solve the above technical problems, the present invention is implemented by the following technical solutions: In the first aspect, a distributed system for solving large-scale mathematical optimization problems comprises:

[0008] Head node, equipped with processor, memory and storage unit, used for global control and task scheduling, as well as executing global strategies in branch and bound algorithm;

[0009] Multiple worker nodes, each of which is also equipped with a processor, memory, and storage unit, are used to execute the computational branch tasks in the branch and bound algorithm in parallel and return the results to the master node;

[0010] The global controller + meta server module is used to manage the metadata and global status of the framework to ensure data consistency and stable operation of the framework;

[0011] Shared storage and shared memory (Shared mem) modules are used to share data between the master node and the worker nodes to improve data access speed and efficiency;

[0012] Local storage module, used to store data on working nodes to ensure data persistence and availability;

[0013] The local scheduler module is used to schedule tasks on the working nodes and arrange the execution of tasks based on factors such as task priority, resource requirements, and node status;

[0014] The user interface module is used to receive the mathematical optimization problem definition and parameter settings input by the user, and to display the final optimization results.

[0015] In this embodiment, distributed computing is used to efficiently solve large-scale mathematical optimization problems. Specifically, the system decomposes complex mathematical optimization problems into multiple subtasks by coordinating multiple computing resources and processes them in parallel on multiple working nodes, thereby significantly improving the speed and efficiency of solving the problem.

[0016] During the calculation process, the master node first receives the mathematical optimization problem definition and parameter settings input by the user through the user interface module. Subsequently, the master node uses the global controller + meta server module to manage the metadata and global state of the framework to ensure data consistency and stable operation of the framework in a distributed environment.

[0017] Next, the master node formulates the global strategy of the branch-and-bound algorithm according to the scale and complexity of the problem, and assigns the computational branch tasks to multiple worker nodes. After receiving the tasks, these worker nodes use their respective processors, memory, and storage units to execute the computational tasks in the branch-and-bound algorithm in parallel.

[0018] In order to speed up the computing process and reduce data transmission delay, the system uses shared storage and shared memory (SharedMem) modules to achieve fast data sharing between the master node and the worker nodes. This sharing mechanism not only improves data access speed, but also optimizes resource utilization.

[0019] At the same time, the local storage module on each working node is used to store intermediate results and necessary data to ensure data persistence and availability. When necessary, this data can be reloaded into memory to support further calculation and analysis.

[0020] In addition, the local scheduler module intelligently arranges the execution order of tasks on each worker node based on factors such as task priority, resource requirements, and node status. This localized scheduling strategy helps optimize resource allocation and improve computing efficiency.

[0021] As the calculation proceeds, the working nodes continuously return the calculation results to the master node. The master node updates the global strategy based on these results and continues to assign new calculation tasks. This process is iterative until the optimal solution is found or the stop condition set by the user is met.

[0022] Finally, when the user interface module receives the optimization results sent by the master node, it will be displayed to the user, thus completing the entire mathematical optimization problem solving process. In this way, the system can efficiently handle large-scale mathematical optimization problems and provide users with accurate and reliable solutions.

[0023] In some specific implementations of the first aspect, the distributed coordinator comprises a plurality of central nodes, a coordination algorithm, and a mechanism for communication between the central nodes;

[0024] Central node: Each central node maintains the address information of all nodes. At startup, each central node will propose itself as the master node (master), and after an election process, a consensus is finally reached.

[0025] Coordination algorithm: The distributed coordinator uses a specific coordination algorithm to ensure the consistency between multiple central nodes and the availability of the overall system. The algorithm used includes Paxos and its variants.

[0026] Communication mechanism: Central nodes communicate with each other through message passing, which requires a reliable communication mechanism to ensure timely delivery of information. In addition, distributed coordinators usually provide functions such as distributed locks, central node selection, and naming services.

[0027] In some specific embodiments of the first aspect, the user interface module further provides a visualization display function of the optimization results, including but not limited to charts, heat maps, and dynamic simulations, so that the user can intuitively understand the optimization process and results;

[0028] The adaptive algorithm module consists of multiple components, including a problem size and complexity monitor, a strategy adjuster, an optimization algorithm library, and a performance evaluator. These components work together to improve solution efficiency by dynamically adjusting the calculation strategy.

[0029] Problem size and complexity monitor: This component is responsible for real-time monitoring of the size and complexity of the optimization problem, such as the amount of data, the number of constraints, the complexity of the objective function, etc. This provides basic data support for subsequent strategy adjustments.

[0030] Strategy Adjuster: Based on the feedback from the problem size and complexity monitor, the strategy adjuster decides which optimization algorithm to use and how to set the algorithm parameters. For example, for large-scale but simple problems, an algorithm with high parallelism and fast computing speed can be used; while for small-scale but complex problems, a more sophisticated search algorithm may be required.

[0031] Optimization algorithm library: This library stores a variety of optimization algorithms and their variants, such as simplex, interior point method, branch and bound, genetic algorithm, simulated annealing algorithm, etc. These algorithms can be called according to the instructions of the strategy adjuster and matched with the corresponding parameter settings.

[0032] Performance evaluator: During the execution of the algorithm, the performance evaluator continuously monitors the performance of the algorithm, such as convergence speed, solution accuracy, etc. If it is found that the current algorithm performs poorly, it can provide feedback to the strategy adjuster to prompt it to replace or adjust the algorithm.

[0033] In some specific implementations of the first aspect, the master control node also has a dynamic load balancing function, which can monitor the load of each working node in real time, and dynamically adjust the computing branch tasks assigned to each working node according to the load condition, so as to ensure that the overall performance and resource utilization of the system are optimized;

[0034] Multiple working nodes can communicate with each other to form a working node network; this network supports data sharing and task collaboration between nodes, further improving the efficiency and flexibility of parallel computing.

[0035] In some specific implementations of the first aspect, the fault-tolerant management module integrates a fault detection unit, a data backup unit, a fault switching unit, and a state recovery unit to achieve continuous operation and rapid recovery of the system in the face of hardware failures or software errors;

[0036] The fault-tolerant management module uses redundant hardware resources and distributed technology to ensure that workloads can be migrated and redistributed between central nodes. At the same time, it maintains the stable operation of the system and the integrity of data through monitoring management and consistency assurance measures.

[0037] Fault detection unit: monitors the status of central nodes in the distributed system, detects and reports any possible errors, such as node downtime, communication timeout, etc.

[0038] Data backup unit and state recovery unit: Regularly back up computing status and data so that they can be quickly restored to normal status when errors occur.

[0039] Failover unit: Once an error is detected, the workload is automatically transferred to the healthy central node to ensure service continuity.

[0040] In some specific embodiments of the first aspect, the user interface further provides a visualization function of the optimization results, including but not limited to charts, heat maps, and dynamic simulations.

[0041] Second, the implementation method for solving large-scale mathematical optimization problems.

[0042] Step 1: The user inputs the definition of the mathematical optimization problem and sets parameters through the user interface module;

[0043] Step 2: The master node combines the global controller + meta server module to analyze the mathematical optimization problem and assigns the computing branch tasks to each working node;

[0044] Step 3: Each working node processes the assigned computing branch tasks in parallel, while the fault-tolerant management module ensures the continuity and reliability of the computing;

[0045] Step 4: The master control node summarizes the calculation results of each working node, forms a global optimization strategy, and displays the final optimization results through the user interface module.

[0046] In some specific embodiments of the second aspect, a post-optimization processing step is further included to verify and improve the accuracy of the calculated results;

[0047] It further includes a data preprocessing module. When the master control node in step 2 allocates computing branch tasks, it will consider the computing power of each working node, the current load, and the dependencies between tasks to optimize the task allocation strategy and reduce the waiting time and resource competition between tasks.

[0048] In some specific implementations of the second aspect, each working node in step 3 will record the execution status and progress of the task in real time when executing the computing branch task, and feed back this information to the master control node. The master control node can monitor the execution of the task in real time based on this information, and promptly discover and handle problems in the execution of the task.

[0049] In some specific implementations of the second aspect, the system may further provide an optimization result analysis and reporting function, which can perform in-depth analysis and interpretation of the optimization results, extract key information and optimization suggestions, and help users better understand the optimization results and make decisions.

[0050] In some specific embodiments of the second aspect, the system further adopts a modular design to support flexible addition or replacement of modules and algorithms according to specific application scenarios and requirements.

[0051] The beneficial effects of the present invention are:

[0052] 1. The present invention uses a distributed computing architecture to break down large-scale and complex optimization problems into multiple subtasks, which are processed in parallel by multiple computing nodes, significantly improving the computing speed and overall performance. It reduces the centralized demand for computing resources, enables the system to efficiently process real-time data, quickly respond to changes, efficiently process ultra-large-scale data sets, solve complex mathematical optimization problems, and significantly improve the speed and accuracy of solving problems. The system uses a branch-and-bound algorithm, combined with a global controller and metaserver module, to ensure data consistency and stable operation of the framework, further improving the efficiency and reliability of solving problems.

[0053] 2. The present invention adopts modular design, supports flexible addition or replacement of modules and algorithms according to specific application scenarios and requirements, and meets the highly customized needs of enterprises. The system can dynamically adjust the calculation strategy and select appropriate optimization algorithms and parameter settings according to the scale and complexity of the problem, which improves the adaptability and flexibility of the system.

[0054] 3. By efficiently solving large-scale mathematical optimization problems, the present invention helps enterprises achieve efficient resource utilization in operational decision-making and reduce energy consumption and emissions. The system can monitor the load of each node in real time, dynamically adjust task allocation, optimize resource utilization, and further reduce operating costs.

[0055] 4. The user interface module provides intuitive visual display functions, including charts, heat maps and dynamic simulations, to help users better understand the optimization process and results. The system provides analysis and reporting functions for optimization results, extracts key information and optimization suggestions, and provides strong support for the company's accurate decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a schematic diagram of the overall system connection structure of the present invention.

[0057] Figure 2 It is a schematic diagram of the internal structure of the overall system of the present invention.

[0058] Figure 3 It is a schematic diagram of the interior of the node set of the present invention.

[0059] Figure 4 It is a schematic diagram of the connection structure between the central node and the computing node of the present invention.

[0060] Figure 5 It is an internal schematic diagram of the fault-tolerant management of the present invention. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] like Figures 1 to 5 The distributed system for solving large-scale mathematical optimization problems shown in the figure includes: a head node, equipped with a processor, memory and storage unit, used for global control and task scheduling, and executing the global strategy in the branch and bound algorithm; multiple work nodes, each of which is also equipped with a processor, memory and storage unit, used to execute the calculation branch tasks in the branch and bound algorithm in parallel, and return the results to the head node; a global controller + meta server module, used to manage the metadata and global state of the framework, ensure the consistency of data and the stable operation of the framework; a shared storage and shared memory (Shared mem) module, used to share data between the head node and the work node, and improve the data access speed and efficiency; a local storage module, used to store data on the work node, to ensure the persistence and availability of data; a local scheduler module, used to schedule tasks on the work node, and arrange the execution of tasks according to factors such as task priority, resource requirements and node status; a user interface module, used to receive the mathematical optimization problem definition and parameter settings input by the user, and to display the final optimization results.

[0063] The user interface module further provides a visualization function of the optimization results, including but not limited to charts, heat maps and dynamic simulations, so that users can intuitively understand the optimization process and results. It also includes a fault-tolerant management module, which integrates fault detection, data backup, fault switching and state recovery functions to ensure the stable operation of the system when a node fails; the fault-tolerant management module uses redundant hardware resources and distributed technology to ensure that the workload can be migrated and redistributed between working nodes, and at the same time maintains the stable operation of the system and the integrity of the data through monitoring management and consistency assurance measures. The master control node also has a dynamic load balancing function, which can monitor the load of each working node in real time, and dynamically adjust the computing branch tasks assigned to each working node according to the load situation to ensure that the overall performance and resource utilization of the system are optimized. Multiple working nodes can communicate with each other to form a working node network; the network supports data sharing and task collaboration between nodes, further improving the efficiency and flexibility of parallel computing.

[0064] The implementation method for solving large-scale mathematical optimization problems includes the following steps: Step 1: The user inputs the definition of the mathematical optimization problem and sets parameters through a user interface module; Step 2: The master control node analyzes the mathematical optimization problem in conjunction with a global controller + metaserver module, and assigns calculation branch tasks to each working node; Step 3: Each working node processes the assigned calculation branch tasks in parallel, while a fault-tolerant management module ensures the continuity and reliability of the calculation; Step 4: The master control node summarizes the calculation results of each working node, forms a global optimization strategy, and displays the final optimization results through the user interface module.

[0065] When allocating computing branch tasks, the master node in step 2 will consider the computing power of each working node, the current load, and the dependencies between tasks to optimize the task allocation strategy and reduce the waiting time and resource competition between tasks. When each working node in step 3 executes the computing branch task, it will record the execution status and progress of the task in real time and feed this information back to the master node. Based on this information, the master node can monitor the execution of the task in real time, and promptly discover and handle problems in the execution of the task. The further system can also provide an analysis and reporting function for the optimization results. This function can deeply analyze and interpret the optimization results, extract key information and optimization suggestions, and help users better understand the optimization results and make decisions. The further system adopts a modular design, which supports the flexible addition or replacement of modules and algorithms according to specific application scenarios and requirements.

[0066] Implementation methods for solving large-scale mathematical optimization problems,

[0067] Step 1: The user defines the mathematical optimization problem and sets parameters through the user interface; Step 2: The distributed coordinator receives the problem definition, analyzes the problem scale and type, and allocates computing resources; Step 3: The adaptive algorithm module adjusts the algorithm parameters according to the characteristics of the problem; Step 4: Each computing node processes the assigned computing tasks in parallel, and ensures the continuity and reliability of the calculation through the fault-tolerant management module; Step 5: The calculation results are summarized and returned to the user interface for display.

[0068] It further includes a post-optimization processing step to verify and improve the accuracy of the calculated results; it further includes a data preprocessing module for cleaning, standardizing and converting the input data before optimization to improve the efficiency and accuracy of the optimization process. The post-optimization processing step includes testing the effectiveness of the optimization results using an external validation data set and adjusting the model parameters based on feedback. The distributed coordinator adopts a decentralized design and realizes direct communication between central nodes through a point-to-point communication protocol to improve the scalability and flexibility of the system; the point-to-point communication protocol uses encryption measures to ensure the security and privacy of data transmission. The system further adopts a modular design to support the flexible addition or replacement of modules and algorithms according to specific application scenarios and requirements.

[0069] Example 1: Distributed computing of large-scale unit combinations in power systems

[0070] System Configuration:

[0071] Master control node: equipped with high-performance processors, large-capacity memory and high-speed storage devices, responsible for global control, task scheduling and execution of algorithm strategies.

[0072] Worker nodes: Multiple worker nodes, each with sufficient computing power and storage resources, are used to parallelize the optimization computing tasks in the power network.

[0073] Global controller + meta server: manages the metadata and global status of the framework to ensure data consistency and stable system operation.

[0074] Shared storage and shared memory: provide high-speed data access channels to accelerate data exchange between the master node and the working nodes.

[0075] Local storage: Stores intermediate data and results on worker nodes to ensure data persistence.

[0076] Local scheduler: Intelligently schedules tasks based on the load and computing requirements of the worker nodes.

[0077] User interface: Provides a graphical interface for inputting problem definitions, parameter settings, and displaying optimization results.

[0078] Implementation method:

[0079] The user inputs the structure, parameters and optimization objectives of the power network through the user interface.

[0080] The master node analyzes the problem, generates computing branch tasks, and assigns them to each working node.

[0081] The working nodes process tasks in parallel and calculate parameters such as power flow and losses.

[0082] The master node aggregates the calculation results, forms a global optimization strategy, and feeds it back to the user.

[0083] Users adjust the power network configuration based on the optimization results.

[0084] Implementation effect:

[0085] It significantly reduces transmission losses in the power grid and improves energy efficiency.

[0086] The output power distribution of the generator sets is optimized and the grid load is balanced.

[0087] Improves the stability and reliability of the power grid and reduces the risk of power outages.

[0088] Application scenarios and extensions:

[0089] Suitable for daily optimization and fault recovery of large power networks.

[0090] Can be expanded to smart grids for real-time monitoring and optimization.

[0091] It can be integrated with other energy management systems to achieve multi-energy complementarity and coordinated optimization.

[0092] Distributed computing for smart city traffic optimization systems

[0093] System Configuration:

[0094] Master node: a high-performance server responsible for global control, task scheduling, and execution of algorithm strategies.

[0095] Work nodes: computing nodes distributed in various areas of the city, used to process traffic data, calculate optimized paths, etc.

[0096] Global controller + meta server: manages the metadata and global status of traffic data to ensure data consistency and stable system operation.

[0097] Shared storage and shared memory: provide high-speed data access channels to accelerate data exchange between the master node and the working nodes.

[0098] Local storage: Stores real-time traffic data and calculation results on working nodes.

[0099] Local scheduler: Intelligently schedules tasks based on traffic conditions and computing requirements.

[0100] User interface: Provides a mobile application or web interface for inputting traffic demands, viewing optimization results and real-time traffic information.

[0101] Implementation method:

[0102] The user inputs transportation requirements, such as the starting point, destination and time of the trip, through the user interface.

[0103] The master node analyzes traffic data, generates computing branch tasks, and distributes them to each working node.

[0104] Working nodes process tasks in parallel, calculating optimal routes, predicting traffic congestion, etc.

[0105] The master control node aggregates the calculation results, forms a global traffic optimization strategy, and feeds it back to the user.

[0106] Users choose the best travel plan based on the optimization results.

[0107] Implementation effect:

[0108] It has significantly reduced urban traffic congestion and improved travel efficiency.

[0109] The allocation of public transportation resources has been optimized and the utilization rate of public transportation has been improved.

[0110] It reduces carbon emissions and promotes environmental protection and sustainable development.

[0111] Application scenarios and extensions:

[0112] Applicable to daily management and emergency response of urban traffic.

[0113] It can be expanded to intelligent transportation systems to realize functions such as vehicle autonomous driving and traffic signal optimization.

[0114] It can be integrated with other smart city systems (such as energy management, environmental monitoring, etc.) to achieve coordinated optimization of the entire city.

[0115] These two examples demonstrate the practical application of distributed computing in solving large-scale mathematical optimization problems. Through flexible system configuration and efficient implementation methods, they can significantly improve the performance and efficiency of power grids and urban transportation systems, providing strong support for the construction and development of smart cities. At the same time, these systems also have good scalability and can be integrated with other systems to achieve more extensive urban optimization and management.

[0116] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A distributed system for solving large-scale mathematical optimization problems, characterized in that: include: Head node, equipped with processor, memory and storage unit, used for global control and task scheduling, and executing global strategies in branch and bound algorithm; Multiple worker nodes, each of which is also equipped with a processor, memory, and storage unit, are used to execute the computational branch tasks in the branch-and-bound algorithm in parallel and return the results to the master node; The global controller + meta server module is used to manage the metadata and global status of the framework to ensure data consistency and stable operation of the framework; Shared storage and shared memory (Shared mem) modules are used to share data between the master node and the worker nodes to improve data access speed and efficiency; Local storage module, used to store data on working nodes to ensure data persistence and availability; The local scheduler module is used to schedule tasks on the worker nodes and arrange the execution of tasks based on factors such as task priority, resource requirements, and node status; The user interface module is used to receive the mathematical optimization problem definition and parameter settings input by the user, and to display the final optimization results.

2. The distributed system for solving large-scale mathematical optimization problems according to claim 1, characterized in that: The user interface module further provides a visualization function of the optimization results, including but not limited to charts, heat maps and dynamic simulations, so that users can intuitively understand the optimization process and results.

3. The distributed system for solving large-scale mathematical optimization problems according to claim 1, characterized in that: Further, it includes a fault-tolerant management module that integrates fault detection, data backup, fault switching and state recovery functions to ensure stable operation of the system in the event of node failure; The fault-tolerant management module utilizes redundant hardware resources and distributed technology to ensure that workloads can be migrated and redistributed among working nodes, while maintaining the stable operation of the system and the integrity of data through monitoring management and consistency assurance measures.

4. The distributed system for solving large-scale mathematical optimization problems according to claim 1, characterized in that: The master control node also has a dynamic load balancing function, which can monitor the load status of each working node in real time, and dynamically adjust the computing branch tasks assigned to each working node according to the load status to ensure that the overall performance and resource utilization of the system are optimal.

5. The distributed system for solving large-scale mathematical optimization problems according to claim 1, characterized in that: The multiple working nodes can communicate with each other to form a working node network; the network supports data sharing and task collaboration between nodes, further improving the efficiency and flexibility of parallel computing.

6. An implementation method for solving large-scale mathematical optimization problems, characterized in that The following steps are involved: Step 1: The user inputs the definition of the mathematical optimization problem and sets parameters through the user interface module; Step 2: The master node combines the global controller + meta server module to analyze the mathematical optimization problem and assigns the computing branch tasks to each working node; Step 3: Each working node processes the assigned computing branch tasks in parallel, while the fault-tolerant management module ensures the continuity and reliability of the computing; Step 4: The master control node summarizes the calculation results of each working node, forms a global optimization strategy, and displays the final optimization results through the user interface module.

7. The implementation method for solving large-scale mathematical optimization problems according to claim 6, characterized in that: When allocating computing branch tasks, the master control node in step 2 will consider the computing power of each working node, the current load, and the dependencies between tasks to optimize the task allocation strategy and reduce the waiting time and resource competition between tasks.

8. The implementation method for solving large-scale mathematical optimization problems according to claim 6, characterized in that: When executing the computing branch task in step 3, each working node will record the execution status and progress of the task in real time, and feed this information back to the master control node. The master control node can monitor the execution of the task in real time based on this information, and promptly discover and handle problems in the task execution.

9. The implementation method for solving large-scale mathematical optimization problems according to claim 6, characterized in that: The system can also provide optimization results analysis and reporting functions. This function can conduct in-depth analysis and interpretation of the optimization results, extract key information and optimization suggestions, and help users better understand the optimization results and make decisions.

10. The implementation method for solving large-scale mathematical optimization problems according to claim 6, characterized in that: The system further adopts a modular design, which supports the flexible addition or replacement of modules and algorithms according to specific application scenarios and requirements.

Images