Power station real-time control and intelligent calculation divide-conquer-collaboration system and method

Abstract

The invention discloses a power station real-time control and intelligent computing divide-conquer-collaboration system and method, and the system comprises an intelligent computing server cluster, and an intelligent computing network which is connected with the intelligent computing server cluster and is used for achieving the internal high-speed computing communication of a server, and achieving the data transmission and task distribution. The data scheduling network is respectively connected with the intelligent computing network, the intelligent scheduling service and the control cluster, the intelligent scheduling service is used for coordinating task distribution and data exchange between the intelligent computing server cluster and the control cluster, the control cluster is connected with the data scheduling network, and the input / output unit is connected with the control cluster. Through an innovative system architecture and technical means, deep fusion of power station control and intelligent calculation is realized, the operation efficiency and reliability of the power station are remarkably improved, and the problem of real-time control and intelligent calculation collaboration of an existing industrial control system under the condition of high-proportion new energy access of a novel power system is solved; the method has obvious practical value and wide application prospect.

Description

Technical Field

[0001] The present invention belongs to the technical field of new power systems, and in particular relates to a power station real-time control and intelligent computing divide-and-conquer collaborative system and method. Background Art

[0002] With the advancement of global energy transformation and power market reform, the operating environment of the power system has become more complex and changeable, which has put forward higher requirements for the flexibility, data processing capabilities, integration and security of the control system. The intermittent and instability of renewable energy poses a huge challenge to the stable operation of the power system. Traditional power systems rely on fossil fuel power generation, and the operating conditions are relatively fixed, making it difficult to adapt to the complex changes of a high proportion of renewable energy access. The opening and competition of the power market have intensified, and electricity prices have fluctuated frequently, which has put forward higher requirements for the flexibility and response speed of the power system. Users' requirements for power quality and reliability are also increasing, especially industrial users and data centers have extremely high requirements for the continuity and stability of power supply. Therefore, the power system needs more intelligent and automated control methods to cope with the high proportion of renewable energy access, the opening of the power market and the diversification of user needs.

[0003] Traditional power industrial control systems face many challenges in the context of new power systems. First, traditional industrial control systems are mainly designed for stable operation under fixed conditions, and it is difficult to quickly respond to grid fluctuations and load changes. Especially when a high proportion of renewable energy is connected, the operating conditions of the power system change frequently, and traditional industrial control systems are difficult to adapt to this dynamic environment. Secondly, the data processing capabilities of traditional industrial control systems are relatively weak, and it is difficult to process large-scale, high-frequency real-time data. The data analysis capabilities are insufficient, and it is impossible to make full use of advanced technologies such as big data and machine learning to achieve comprehensive monitoring and optimization control of the system status. In addition, the integration between the subsystems of traditional industrial control systems is low, the information island phenomenon is serious, and the data exchange and collaborative control capabilities between the subsystems are weak, making it difficult to achieve cross-system collaborative optimization and intelligent control. Finally, the hardware architecture of traditional industrial control systems is relatively fixed, with poor scalability, and it is difficult to adapt to the ever-changing needs of the power system. It takes a lot of time and cost to add new functions and modules, which affects the flexibility and adaptability of the system. The software architecture is rigid, and it is difficult to quickly integrate new control algorithms and applications, which limits the innovation and development of the system.

[0004] In response to the above problems, at present, domestic and foreign research mainly introduces intelligent control technology to the periphery of traditional industrial control systems to achieve local optimization and improve system performance. For example, by combining machine learning and optimization algorithms, the control strategy of traditional industrial control systems is optimized. Using predictive control algorithms, the future system state can be predicted based on historical data and real-time data, and control parameters can be adjusted in advance to improve the response speed and stability of the system. Without affecting the overall system architecture, key subsystems are locally optimized, such as optimizing the boiler combustion control system to improve combustion efficiency and reduce emissions, and optimizing the turbine control system to improve the operation efficiency and reliability of the turbine. Although the local optimization technology of intelligent control based on traditional industrial control systems can improve the performance of the system to a certain extent, it also has some limitations and shortcomings. First, local optimization can only improve the performance of specific subsystems, and it is difficult to comprehensively improve the flexibility and response speed of the entire system. Global optimization requires deeper system transformation and integration. Secondly, integrating intelligent control algorithms and edge computing devices into traditional industrial control systems requires overcoming compatibility and communication issues between different systems, which increases the complexity and maintenance difficulty of the system. Intelligent control relies on high-quality and complete data, but data acquisition and transmission in traditional industrial control systems may be delayed, lost or erroneous, affecting the effect of intelligent control. The introduction of intelligent control technology and equipment requires additional investment, including hardware purchase, software development and system integration, which increases the initial cost of the project. Some intelligent control algorithms and technologies are still under development, and their stability and reliability have not been fully verified, which may pose risks in actual applications. Summary of the invention

[0005] The purpose of the present invention is to provide a power plant real-time control and intelligent computing divide-and-conquer collaborative system and method, so as to overcome the problem that existing industrial control systems are difficult to solve in the new power system with a high proportion of new energy access conditions. The efficient collaboration of intelligent computing and real-time control, and improve the system's flexibility, scalability, compatibility and global optimization capabilities.

[0006] In order to solve the above problems, the present invention adopts the following technical solutions: A power station real-time control and intelligent computing division-cooperation system, comprising an intelligent computing server cluster and an intelligent dispatching service, wherein the intelligent computing server cluster and the intelligent dispatching service are connected and communicated with each other via an intelligent computing network and a data dispatching network in turn; The data scheduling network is also connected to at least one control cluster, and each control cluster is connected to an input / output unit; The intelligent computing server cluster is used to perform intelligent computing tasks; The intelligent computing network is used to connect the intelligent computing server cluster to realize high-speed computing communication within the server, and provide a communication interface to interact with the data scheduling network to realize data transmission and task allocation; The data scheduling network is used to connect the intelligent scheduling service and several control clusters for real-time data transmission and scheduling; The intelligent scheduling service is used to coordinate task allocation and data exchange between the intelligent computing server cluster and the control cluster; The control cluster includes multiple controllers for real-time control of power plant equipment; The input / output unit is used to collect field data and send control instructions.

[0007] Furthermore, the intelligent scheduling service adopts a master-slave redundancy design and a heartbeat monitoring mechanism, and has the functions of multi-objective optimization scheduling, combination of active and dynamic scheduling, and distributed collaborative scheduling.

[0008] Furthermore, each of the control clusters includes a plurality of control units, and the control unit includes a controller with a redundant design, and the controller has a built-in adaptive control algorithm and supports remote configuration and updating: The control cluster uses adaptive resource pooling technology to transform the minimum control unit into an elastic and flexible control resource pool, supporting dynamic resource scheduling and elastic load.

[0009] Furthermore, the intelligent computing network adopts distributed caching technology and traffic shaping algorithm, the intelligent computing server cluster has a built-in adaptive resource scheduling algorithm and task priority management system, and each server in the intelligent computing server cluster is equipped with a dual redundant trusted encrypted network communication module.

[0010] Furthermore, the data scheduling network supports multi-protocol transmission, adaptive traffic slicing, integrated encryption technology and data integrity verification mechanism, and a built-in intelligent monitoring module.

[0011] The input / output unit integrates AI, AO, DI, DO, and Modbus hardware and software devices, unifies the hardware interface, and defines the input / output mode by software, forming an adaptive module that supports conversion of multiple input / output modes.

[0012] In a second aspect, a power plant real-time control and intelligent computing divide-and-conquer collaborative method is provided, comprising the following steps: The input / output unit collects real-time data on site and transmits the data to the intelligent dispatch service; Intelligent scheduling services generate task allocation instructions based on real-time data and historical data analysis results; The intelligent computing server cluster receives task allocation instructions through the intelligent computing network and data scheduling network, and obtains real-time data on site. The intelligent computing server cluster dynamically adjusts computing resources according to real-time data and task allocation instructions; The intelligent computing server cluster executes intelligent computing tasks and feeds back the computing results to the intelligent scheduling service; The intelligent scheduling service generates control instructions based on the calculation results and sends them to the corresponding control cluster through the data scheduling network; The controller in the control cluster receives the control instructions and sends them to the input / output unit to adjust the operating parameters of the power plant equipment in real time.

[0013] Furthermore, the intelligent scheduling service automatically optimizes task allocation and resource scheduling strategies based on real-time data and historical data analysis results, continuously optimizes scheduling strategies using reinforcement learning algorithms, predicts future task loads based on historical data, and pre-allocates resources in advance.

[0014] Furthermore, the intelligent scheduling service supports multi-objective optimization.

[0015] Furthermore, the input / output unit has self-diagnosis function, fault log recording and remote backtracking function.

[0016] Compared with the prior art, the present invention has the following beneficial technical effects: The present invention provides a power station real-time control and intelligent computing divide-and-conquer collaborative system, comprising an intelligent computing server cluster, an intelligent computing network connected to the intelligent computing server cluster for realizing high-speed computing communication within the server, realizing data transmission and task allocation, a data scheduling network connected to the intelligent computing network, the intelligent scheduling service and the control cluster respectively, an intelligent scheduling service for coordinating task allocation and data exchange between the intelligent computing server cluster and the control cluster, a control cluster connected to the data scheduling network and an input / output unit connected to the control cluster. Through innovative system architecture and technical means, a deep integration of power station control and intelligent computing is achieved, the operation efficiency and reliability of the power station are significantly improved, and the real-time control and intelligent computing collaboration problems of the existing industrial control system under the condition of a high proportion of new energy access to the new power system are solved. The system has significant practical value and broad application prospects.

[0017] Intelligent scheduling service module: adopts active-standby redundant design and heartbeat monitoring mechanism to ensure high availability and reliability of the system. Through multi-objective optimization scheduling, active and dynamic scheduling, and distributed collaborative scheduling, the intelligent scheduling service module can automatically optimize task allocation and resource scheduling strategies based on real-time data and historical data analysis results, predict future task loads in advance and pre-allocate resources to ensure the stable operation of the system under the condition of a high proportion of new energy access. The intelligent scheduling service module supports multi-objective optimization, taking into account multiple dimensions such as performance, energy consumption and cost, to achieve more efficient resource utilization. Through the reinforcement learning algorithm, the scheduling strategy is continuously optimized, the future task load is predicted based on historical data, and resources are pre-allocated in advance to ensure the efficient operation of the system under the condition of a high proportion of new energy access.

[0018] Control cluster: Integrate multiple minimum control units into a control cluster, use adaptive resource pooling technology to form a flexible control resource pool, and realize dynamic allocation and elastic scaling of resources. The control cluster supports dynamic resource scheduling, automatically adjusts resource allocation according to real-time task requirements and system load conditions, ensures that resources can be quickly expanded under high load, and can be released under low load, thereby improving resource utilization.

[0019] Improve system flexibility and scalability: Intelligent computing server cluster: Built-in adaptive resource scheduling algorithm and task priority management system, each server is equipped with dual redundant trusted encrypted network communication modules to ensure the security and reliability of data transmission. Intelligent computing server cluster can dynamically adjust computing resources according to real-time data and task allocation instructions, ensuring that the system can run efficiently under different loads, greatly improving the flexibility and response speed of the system.

[0020] Input / output unit: Integrate the traditional independent AI, AO, DI, DO, and Modbus hardware and software devices, unify the hardware interface, and define the input / output mode with software to form an adaptive module that supports multiple input / output mode conversions. Through the software-defined function, users can flexibly configure the function of each interface according to actual needs, realize the conversion of multiple input / output modes, improve the flexibility and scalability of the system, have self-diagnosis function, fault log recording and remote backtracking function, can monitor its own status in real time and immediately issue an alarm when an abnormality occurs, helping technicians quickly locate the root cause of the problem and shorten maintenance time.

[0021] Enhance system compatibility and global optimization capabilities: Intelligent computing network: Distributed caching technology and traffic shaping algorithms are used to effectively reduce network congestion and ensure the priority transmission of key tasks. The intelligent computing network can efficiently process large-scale, high-frequency real-time data, support advanced technologies such as big data and machine learning, and achieve comprehensive monitoring and optimization control of system status.

[0022] Data dispatch network: supports multi-protocol transmission, integrates encryption technology and data integrity verification mechanism, and has a built-in intelligent monitoring module. The data dispatch network uses adaptive traffic slicing technology to divide the physical network into multiple logical network slices. Each slice is customized according to specific task requirements to ensure that task scheduling, data transmission and control instructions do not affect each other and are highly reliable. By adopting a communication protection mechanism, it effectively ensures the detour of business after a line or equipment failure in the network.

[0023] The controller has a built-in adaptive control algorithm: The controller can dynamically adjust the control parameters according to real-time feedback to ensure that the equipment always operates in the best state. At the same time, it supports remote configuration and updates, which facilitates centralized management by maintenance personnel and improves the maintenance efficiency and reliability of the system.

[0024] The present invention provides a divide-and-conquer-coordinated method for real-time control and intelligent computing of a power station. The intelligent computing server cluster dynamically adjusts computing resources according to real-time data and task allocation instructions generated by an intelligent scheduling service, executes intelligent computing tasks, and feeds back the computing results to the intelligent scheduling service. The intelligent scheduling service generates control instructions and sends them to the control cluster and input / output units. This divide-and-conquer-coordinated mode not only optimizes resource utilization, but also improves the real-time and flexibility of the system, and further enhances the adaptability of the power station to complex working conditions. In particular, in a new power system with a high proportion of new energy access, it can effectively ensure the stable operation of the power station, reduce operation and maintenance costs, and improve economic benefits, providing strong technical support for the development of new power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a structural diagram of a power station real-time control and intelligent computing divide-and-conquer collaborative system in an embodiment of the present invention; Figure 2 The present invention is a flowchart of a power plant real-time control and intelligent computing divide-and-conquer collaborative method in an embodiment of the present invention.

[0026] In the figure, 1. Intelligent computing server cluster; 2. Intelligent computing network; 3. Data scheduling network; 4. Intelligent scheduling service; 5. Control cluster; 6. Input / output unit. DETAILED DESCRIPTION

[0027] In order to make the technical problems, technical solutions and beneficial effects solved by the present invention more clearly understood, the present invention is further described in detail in the following specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0028] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0029] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. 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.

[0030] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0031] In the description of the embodiments of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0032] A power plant real-time control and intelligent computing divide-and-conquer collaborative system, such as Figure 1 As shown, it includes an intelligent computing server cluster 1, an intelligent computing network 2, a data scheduling network 3, an intelligent scheduling service 4, a control cluster 5 and an input / output unit 6, wherein the intelligent computing server cluster 1, the intelligent computing network 2, the data scheduling network 3 and the intelligent scheduling service 4 are connected in sequence, the input end of the control cluster 5 is connected to the data scheduling network 3, and the output end of the control cluster 5 is connected to the input / output unit 6.

[0033] In this embodiment, both the intelligent scheduling service 4 and the control cluster 5 adopt a redundant design to improve reliability. Specifically, there are two intelligent scheduling services 4, one as the primary and the other as the backup; the data scheduling network 3 connects multiple control clusters 5, each control cluster 5 includes multiple minimum control units, and each minimum control unit includes two intelligent controllers. Similarly, the two intelligent controllers are designed as primary and backup redundancy.

[0034] The controller has a built-in adaptive control algorithm that can dynamically adjust control parameters based on real-time feedback to ensure that the equipment always operates in the best condition. At the same time, it supports remote configuration and updates to facilitate centralized management by maintenance personnel.

[0035] There are multiple input / output units 6, and each input / output unit 6 includes multiple input / output modules for collecting field data and sending control instructions.

[0036] In an optional embodiment, a power plant real-time control and intelligent computing division-cooperation system is provided, including an intelligent computing server cluster 1, an intelligent computing network 2, a data scheduling network 3, an intelligent scheduling service 4, a plurality of control clusters 5, and a plurality of input / output units 6. The intelligent computing server cluster 1 is used to perform complex intelligent computing tasks. The intelligent computing network 2 is used to connect the intelligent computing server cluster 1 to realize high-speed computing communication within the server, and provide a communication interface to interact with the data scheduling network to realize data transmission and task allocation. The data scheduling network 3 connects the intelligent scheduling service 4 and a plurality of control clusters 5 for real-time data transmission and scheduling. The intelligent scheduling service 4 is used to coordinate task allocation and data exchange between the intelligent computing server cluster 1 and the control cluster 5. Each of the plurality of control clusters 5 includes a plurality of controllers for real-time control of power plant equipment. Each of the plurality of input / output units 6 includes a plurality of input / output modules for collecting field data and sending control instructions.

[0037] In this embodiment, the intelligent computing server cluster 1 can allocate tasks according to the intelligent scheduling service 4, with a built-in adaptive resource scheduling algorithm, monitor the task queue in real time, and dynamically adjust the hardware resources such as CPU and GPU to ensure efficient operation under different loads, greatly improving the flexibility and response speed of the system. The intelligent computing server cluster 1 can dynamically configure and deploy 3 to 128 servers according to the actual needs of the power station. Each server is equipped with a dual-redundant trusted encrypted network communication module, which transmits data inward to the server storage module and execution module, and transmits data outward to other servers or data scheduling networks. The dual-redundant communication modules are mutually active and standby during normal operation, working in parallel. When one of the communication modules fails, the other module automatically takes over the work of the failed communication module; Intelligent computing network 2 has high bandwidth and low latency characteristics, ensuring efficient communication between intelligent computing server cluster 1 and intelligent scheduling service 4. Intelligent computing network 2 adopts distributed caching technology and traffic shaping algorithm, effectively reducing network congestion and ensuring priority transmission of key tasks; it supports automatic fault switching and fast recovery, and improves the fault tolerance of the system.

[0038] Data Dispatch Network 3 supports multi-protocol transmission to ensure the security and integrity of real-time data. The Data Dispatch Network not only supports multiple transmission protocols, but also integrates advanced encryption technology and data integrity verification mechanism to ensure the security and consistency of data during transmission; at the same time, the built-in intelligent monitoring module detects and reports any potential security threats in real time.

[0039] Intelligent Scheduling Network 3 supports adaptive traffic slicing technology, which can automatically slice traffic according to the task situation, ensuring that task scheduling, data transmission and control instructions do not affect each other and are highly reliable. The specific implementation method is as follows: Adaptive traffic slicing: Smart Scheduling Network 3 uses adaptive traffic slicing technology to divide the physical network into multiple logical network slices, each of which is customized according to specific task requirements. Through tools such as machine learning and random games, network slices are adaptively and dynamically adjusted to optimize long-term resource utilization, ensuring that task scheduling, data transmission, and control instructions do not affect each other and are highly reliable.

[0040] High reliability design: The overall reliability of the intelligent dispatching network 3 must meet the requirements of network topology reliability, equipment reliability, low network latency, low network-wide routing convergence time, network management stability and reliability, etc. By adopting a communication protection mechanism, it effectively ensures that services are transmitted in a roundabout way after a line or equipment failure in the network.

[0041] Intelligent Scheduling Service 4 can automatically optimize task allocation and resource scheduling strategies based on real-time data and historical data analysis results. Intelligent Scheduling Service uses reinforcement learning algorithms to continuously optimize scheduling strategies, predict future task loads based on historical data, and pre-allocate resources in advance; at the same time, it supports multi-objective optimization, taking into account multiple dimensions such as performance, energy consumption and cost, and achieving more efficient resource utilization.

[0042] Intelligent Scheduling Service 4 has the functions of multi-objective optimization scheduling, active and dynamic scheduling, and distributed collaborative scheduling: Multi-objective optimization scheduling: Supports multi-objective optimization, taking into account multiple dimensions such as performance, energy consumption and cost, to achieve more efficient resource utilization. The intelligent scheduling service module can automatically optimize task allocation and resource scheduling strategies based on real-time data and historical data analysis results, predict future task loads based on historical data, and pre-allocate resources in advance.

[0043] Combination of active and dynamic scheduling: Combining active scheduling and dynamic scheduling strategies, the intelligent scheduling service module can predict the system failure in advance by comparing and analyzing real-time data and historical data before the equipment fails, and on this basis, reschedule the remaining normal equipment. At the same time, for emergencies in production sequencing, such as emergency behavior, the intelligent scheduling service module can adjust the scheduling plan in real time to ensure the efficiency and stability of the production process.

[0044] Distributed collaborative scheduling: supports intelligent interconnection with multiple distributed scheduling centers. Through computational intelligence and feature analysis technology, it conducts group intelligent multi-search operations and knowledge-based searches on the sub-dispatching centers. Finally, it integrates the obtained information and uses knowledge-driven collaborative intelligent algorithms for scheduling.

[0045] The controller in the control cluster 5 can adjust the control parameters in real time according to the instructions of the intelligent dispatch service 4 to ensure that the power plant equipment operates in the optimal state. The controller has a built-in adaptive control algorithm that can dynamically adjust the control parameters according to real-time feedback to ensure that the equipment always operates in the optimal state; at the same time, it supports remote configuration and updating, which is convenient for maintenance personnel to perform centralized management.

[0046] Control cluster 5 uses adaptive resource pooling technology to form a flexible control resource pool with the smallest control unit, supporting dynamic resource scheduling and elastic load. The specific implementation method is as follows: Conversion from minimum control unit to control cluster: Integrate multiple minimum control units into a control cluster, use adaptive resource pooling technology to form an elastic and flexible control resource pool, realize dynamic allocation and elastic scaling of resources, and manage and schedule according to actual control business needs. At the same time, to ensure security, when the resource pool fails, container independent orchestration technology is used, and each minimum control unit can run as an independent container.

[0047] Dynamic scheduling and elastic load: The control cluster supports dynamic resource scheduling, automatically adjusting resource allocation according to real-time task requirements and system load conditions. Through the built-in resource scheduling algorithm, combined with the adaptive load mechanism, elastic scaling of resources is achieved, ensuring that resources can be quickly expanded under high load and released under low load, thereby improving resource utilization.

[0048] The input / output unit 6 has a self-diagnosis function, which can sound an alarm and record fault information in abnormal situations. The input / output unit uses an embedded intelligent diagnostic chip, which can monitor its own status in real time and immediately sound an alarm in case of abnormalities; at the same time, it has a fault log recording and remote backtracking function to help technicians quickly locate the root cause of the problem and shorten the maintenance time.

[0049] The input / output unit 6 integrates the traditional independent AI, AO, DI, DO, and Modbus hardware and software devices, unifies the hardware interface, and defines the input / output mode with software, forming an adaptive module that supports multiple input / output mode conversions. The specific implementation method is as follows: AI, AO, DI, DO, Modbus integration: The input / output unit 6 adopts a unified hardware interface design and supports the input and output of multiple signal types (such as AI, AO, DI, DO, Modbus). Through software-defined input / output modes, users can flexibly configure the functions of each interface according to actual needs and realize the conversion of multiple input / output modes. For example, through the configuration file or graphical interface, the user can set an interface to AI mode for temperature sensor signal input, or set another interface to DO mode for controlling relays.

[0050] Unified hardware interface: The hardware interface of the input / output unit 6 adopts a standardized design, supports the access of multiple signal types, reduces the types and number of hardware interfaces, and improves the integration and reliability of the system. At the same time, through the unified interface standard, the maintenance and expansion of the system are simplified.

[0051] Software-defined input / output mode: The input / output unit 6 supports software-defined functions. Users can flexibly define the input / output mode of each interface through software configuration files or graphical configuration tools. This software-defined approach enables the input / output unit to adapt to different application scenarios and requirements, improving the flexibility and scalability of the system.

[0052] The present invention also provides a power plant real-time control and intelligent computing divide-and-conquer collaborative method, such as Figure 2 As shown, the following steps are included: The input / output unit 6 collects real-time data on site and transmits the data to the intelligent scheduling service 4; Intelligent scheduling service 4 generates task allocation instructions based on real-time data and historical data analysis results; The intelligent computing server cluster 1 receives the task allocation instruction through the intelligent computing network 2 and the data scheduling network 3, and obtains the on-site real-time data. The intelligent computing server cluster 1 dynamically adjusts the computing resources according to the real-time data and the task allocation instruction; The intelligent computing server cluster 1 executes the intelligent computing task and feeds back the computing result to the intelligent scheduling service 4; The intelligent scheduling service 4 generates control instructions according to the calculation results and sends them to the corresponding control cluster 5 through the data scheduling network 3; The controller in the control cluster 5 sends the control instruction to the input / output unit 6 according to the received control instruction, and adjusts the operating parameters of the power plant equipment in real time.

[0053] In detail, the specific steps of the above control method are: Step S1: Collect field data through the input / output unit 6 and transmit the data to the intelligent scheduling service 4; the input / output unit performs preliminary processing on the collected data, such as filtering and compression, to reduce the amount of data transmission and improve the data quality; at the same time, a built-in intelligent filter is used to transmit only useful data, further optimizing bandwidth usage; • Step S2: The intelligent scheduling service 4 generates task allocation instructions based on the analysis results of real-time data and historical data; the intelligent scheduling service uses machine learning algorithms to automatically optimize task allocation and resource scheduling strategies; at the same time, it builds a prediction model based on historical data to plan future tasks in advance to ensure that the system is always in the optimal operating state; • Step S3: The intelligent computing server cluster 1 receives the task assignment instruction and obtains the required data through the intelligent computing network 2; the intelligent computing server cluster dynamically adjusts computing resources according to task requirements to improve resource utilization and response speed; at the same time, the intelligent computing network adopts distributed caching technology to ensure the efficiency and stability of data transmission; • Step S4: The intelligent computing server cluster 1 executes the intelligent computing task and feeds back the result to the intelligent scheduling service 4; the intelligent computing server cluster realizes real-time data transmission through the high-bandwidth, low-latency intelligent computing network; at the same time, the built-in task priority management system ensures that key tasks are processed first and improves overall efficiency; • Step S5: The intelligent scheduling service 4 generates control instructions according to the calculation results and sends them to the corresponding control cluster 5 through the data scheduling network 3; the intelligent scheduling service dynamically adjusts the control instructions according to the real-time feedback data to ensure the stability and response speed of the system; at the same time, the data scheduling network supports intelligent routing selection, optimizes the transmission path, and improves efficiency; •Step S6: The controller in the control cluster 5 adjusts the operating parameters of the power plant equipment in real time according to the received control instructions; the control cluster uses edge computing technology to process part of the data locally, reduces dependence on the central server, and enhances real-time response capabilities; at the same time, a built-in adaptive control algorithm ensures that the equipment always operates in the best state.

[0054] In the context of new power systems, traditional industrial control systems face many challenges, including insufficient flexibility, limited data processing capabilities, low integration, network security risks and poor scalability. In order to meet these challenges, it is necessary to introduce more advanced control technologies and platforms, such as edge computing, cloud computing, big data analysis and artificial intelligence, to improve the flexibility, data processing capabilities, integration and security of the system, and realize intelligent and efficient power production and dispatching. Finally, in order to achieve efficient and reliable operation of the new power system, the real-time control and intelligent computing divide-and-conquer collaborative system and method in the present invention are constructed. The present invention combines real-time control with intelligent computing, handles local optimization problems through a divide-and-conquer strategy, and achieves global optimization through a collaborative mechanism to ensure the overall performance and stability of the system. Specifically, real-time control ensures real-time response and control of key subsystems, improves the flexibility and reliability of the system; intelligent computing uses technologies such as big data analysis and machine learning to achieve global optimization and intelligent decision-making, and improves the overall performance of the system; divide-and-conquer collaborative processes local optimization problems through a divide-and-conquer strategy, and achieves global optimization through a collaborative mechanism to ensure the overall performance and stability of the system. This divide-and-conquer-collaborate system and method can effectively solve the limitations of traditional industrial control systems, provide strong support for the construction of new power systems, and promote the digital transformation of the power industry.

[0055] The preferred embodiments of the present invention are described in detail above, but the present invention is not limited thereto. Within the technical concept of the present invention, the technical solution of the present invention can be subjected to a variety of simple modifications, including the combination of various technical features in any other suitable manner, and these simple modifications and combinations should also be regarded as the contents disclosed by the present invention and belong to the protection scope of the present invention.

Claims

1. A power plant real-time control and intelligent computing divide-and-conquer collaborative system, characterized in that: It comprises an intelligent computing server cluster (1) and an intelligent scheduling service (4), wherein the intelligent computing server cluster (1) and the intelligent scheduling service (4) are connected and communicated with each other via an intelligent computing network (2) and a data scheduling network (3) in sequence; The data scheduling network (3) is also connected to at least one control cluster (5), and each control cluster (5) is connected to an input / output unit (6); The intelligent computing server cluster (1) is used to execute intelligent computing tasks; The intelligent computing network (2) is used to connect the intelligent computing server cluster (1) to realize high-speed computing communication within the server, and provide a communication interface to interact with the data scheduling network (3) to realize data transmission and task allocation; The data dispatching network (3) is used to connect the intelligent dispatching service (4) and a plurality of control clusters (5) for real-time data transmission and dispatching; The intelligent scheduling service (4) is used to coordinate task allocation and data exchange between the intelligent computing server cluster (1) and the control cluster (5); The control cluster (5) comprises a plurality of controllers for performing real-time control of power plant equipment; The input / output unit (6) is used to collect field data and send control instructions.

2. A power plant real-time control and intelligent computing divide-and-conquer collaborative system according to claim 1, characterized in that: The intelligent scheduling service (4) adopts a master-slave redundancy design and a heartbeat monitoring mechanism, and has the functions of multi-objective optimization scheduling, active and dynamic scheduling combination, and distributed collaborative scheduling.

3. A power plant real-time control and intelligent computing divide-and-conquer collaborative system according to claim 1, characterized in that: Each of the control clusters (5) includes a plurality of control units, wherein the control unit includes a controller with a redundant design, the controller has a built-in adaptive control algorithm, and supports remote configuration and updating; The control cluster (5) uses adaptive resource pooling technology to transform the minimum control unit into an elastic and flexible control resource pool, supporting dynamic resource scheduling and elastic load.

4. A power plant real-time control and intelligent computing divide-and-conquer collaborative system according to claim 1, characterized in that: The intelligent computing network (2) adopts distributed caching technology and traffic shaping algorithm, the intelligent computing server cluster (1) has a built-in adaptive resource scheduling algorithm and task priority management system, and each server in the intelligent computing server cluster (1) is equipped with a dual redundant trusted encryption network communication module.

5. A power plant real-time control and intelligent computing divide-and-conquer collaborative system according to claim 1, characterized in that: The data dispatching network (3) supports multi-protocol transmission, adaptive traffic slicing, integrated encryption technology and data integrity verification mechanism, and has a built-in intelligent monitoring module.

6. A power plant real-time control and intelligent computing divide-and-conquer collaborative system according to claim 1, characterized in that: The input / output unit (6) integrates AI, AO, DI, DO, and Modbus hardware and software devices, unifies the hardware interface, and defines the input / output mode with software, thereby forming an adaptive module that supports conversion of multiple input / output modes.

7. A divide-and-conquer collaborative method for real-time control and intelligent computing of a power plant, characterized in that: The power plant real-time control and intelligent computing divide-and-conquer collaborative system according to any one of claims 1 to 6 comprises the following steps: The input / output unit (6) collects real-time data on site and transmits the data to the intelligent dispatching service (4); Intelligent scheduling service (4) generates task allocation instructions based on real-time data and historical data analysis results; The intelligent computing server cluster (1) receives the task allocation instruction through the intelligent computing network (2) and the data scheduling network (3), and obtains the on-site real-time data. The intelligent computing server cluster (1) dynamically adjusts the computing resources according to the real-time data and the task allocation instruction; The intelligent computing server cluster (1) executes the intelligent computing tasks and feeds back the computing results to the intelligent scheduling service (4); The intelligent scheduling service (4) generates control instructions based on the calculation results and sends them to the corresponding control cluster (5) through the data scheduling network (3); The controller in the control cluster (5) sends the control instruction to the input / output unit (6) according to the received control instruction, and adjusts the operating parameters of the power plant equipment in real time through the input / output unit (6).

8. A divide-and-conquer-coordinated method for real-time control and intelligent computing of a power plant according to claim 7, characterized in that: The intelligent scheduling service (4) automatically optimizes task allocation and resource scheduling strategies based on the analysis results of real-time data and historical data, continuously optimizes scheduling strategies using reinforcement learning algorithms, predicts future task loads based on historical data, and pre-allocates resources in advance.

9. A divide-and-conquer-coordinated method for real-time control and intelligent computing of a power plant according to claim 8, characterized in that: The intelligent scheduling service (4) supports multi-objective optimization.

10. A divide-and-conquer-coordinated method for real-time control and intelligent computing of a power plant according to claim 7, characterized in that: The input / output unit (6) has a self-diagnosis function, a fault log recording function, and a remote backtracking function.

Images