ECRH system distributed control and dynamic scheduling method and system

Abstract

The invention discloses an ECRH system distributed control and dynamic scheduling method and system, and the method comprises the steps: dividing ECRH equipment into a plurality of IOC nodes which operate independently based on an EPICS distributed control architecture, and enabling each IOC node to be responsible for local data collection and control; after each IOC node preprocesses the collected data, the priority score of each piece of data is calculated according to the access frequency and the real-time requirement, and the priority level of the data is marked; issuing and sharing the data marked with the priority level by adopting a CA protocol; each IOC node subscribes to the shared data according to the correlation and importance of the IOC node and the data; according to the shared data and the data subscribed by each IOC node, in combination with a given task, dynamic adjustment control is carried out, intelligent scheduling distribution of the task is achieved, high-frequency key data can be rapidly transmitted and processed in real time, and it is ensured that the system can complete response within microsecond-level time.

Description

Technical Field

[0001] The present invention relates to the technical field of electron cyclotron resonance heating systems, and in particular to a method and system for distributed control and dynamic scheduling of an ECRH system. Background Art

[0002] The electron cyclotron resonance heating (ECRH) system is a crucial subsystem in the nuclear fusion device. It is mainly used for plasma heating and current driving. Its operating performance is directly related to the success or failure of the nuclear fusion experiment. Under the requirements of high-parameter and long-pulse experiments, the ECRH system needs to monitor and control key parameters such as the output power of the RF source, the standing wave ratio of the transmission line, and the reflected power in real time. These parameters usually have high-frequency change characteristics, which puts extremely high demands on the real-time performance and response speed of the control system. However, the traditional centralized control architecture is difficult to meet the microsecond response requirements due to the limitations of data transmission delays and processing bottlenecks. In addition, with the expansion of the scale of experiments, the number of devices and data sampling rates in the ECRH system continue to increase. The computing load and communication bandwidth of the centralized control center have become bottlenecks in system performance, further limiting the system's expansion capabilities.

[0003] The traditional centralized architecture also shows low efficiency in multi-node collaboration. The centralized architecture relies on a single control center to handle all data collection, analysis, and instruction generation. Since all nodes must transmit data to the control center for centralized processing and then send the results to each device, the overall data transmission and processing delay is high, which cannot meet the microsecond dynamic response requirements of the ECRH system. In the centralized architecture, the collaboration between subsystems (such as RF sources, cooling systems, and transmission antennas) is weak. Data transmission and control instructions between subsystems often need to be transferred through a centralized control center, causing communication bottlenecks and affecting the efficiency of real-time collaboration between nodes. For example, subsystems such as RF sources, cooling systems, and transmission antennas in the ECRH system need to work closely together to maintain the stable operation of the system. When the output power of the RF source is abnormal, the cooling system must adjust the cooling flow in real time to ensure the safety of the equipment. However, due to the basic limitations of data sharing and collaboration mechanisms, traditional control systems are difficult to achieve efficient task allocation and dynamic collaboration between nodes, affecting the overall operating efficiency.

[0004] Experimental Physics and Industrial Control Systems (EPICS) have been widely used in the field of nuclear fusion device control in recent years due to its openness and distributed architecture. EPICS achieves real-time control and status monitoring of equipment through distributed Input / Output Controllers (IOCs). Each IOC node runs independently and shares data through the Channel Access (CA) protocol. Its distributed architecture has high flexibility and scalability. However, the traditional EPICS framework still has shortcomings in the high-dynamic experimental scenarios of the ECRH system: on the one hand, the CA protocol has a certain delay in high-frequency data sharing and cannot fully meet the needs of microsecond-level data acquisition and control; on the other hand, the coordination mechanism between IOC nodes is not optimized for complex task allocation and dynamic load balancing, and the multi-node coordination efficiency is low. Summary of the invention

[0005] The technical problem to be solved by the present invention is that the traditional EPICS framework cannot fully meet the requirements of microsecond-level data acquisition and control.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] A distributed control and dynamic scheduling method of an ECRH system, comprising:

[0008] Based on the EPICS distributed control architecture, the ECRH equipment is divided into multiple independently operated IOC nodes, each of which is responsible for local data acquisition and control;

[0009] After preprocessing the collected data, each IOC node calculates the priority score of each data based on the access frequency and real-time requirements, and marks the priority level of the data;

[0010] The CA protocol is used to publish and share data with marked priority levels; each IOC node subscribes to the shared data based on its own relevance and importance to the data;

[0011] Based on the shared data and the data subscribed by each IOC node, combined with the given tasks, dynamic adjustment control is performed to achieve intelligent scheduling and allocation of tasks.

[0012] In one embodiment of the present invention, the IOC node includes a radio frequency source IOC node, a cooling system IOC node and a transmission antenna IOC node.

[0013] In one embodiment of the present invention, the priority of each data is calculated by the following formula:

[0014] P=w t ·F t +w I It ;

[0015] In the formula, P is the priority score of the data, F t For real-time requirements, I t is the access frequency, w t is the real-time demand weight parameter, w I is the access frequency weight parameter;

[0016] When the priority score is greater than or equal to the high priority threshold, the data is marked as high-frequency critical data and placed in the real-time processing queue;

[0017] When the priority score is less than the high priority threshold, the data is marked as low-frequency secondary data and placed in the low-priority queue;

[0018] And, periodically re-evaluate data characteristics and update priority assignments.

[0019] In one embodiment of the present invention, the real-time demand weight parameter w t , access frequency weight parameter w I The value of is set according to the application scenario; w is recommended for high dynamic scenarios t >w I , analysis task recommendation w t <w I .

[0020] In one embodiment of the present invention, the correlation between the shared data and each IOC node is obtained by the following formula:

[0021]

[0022] The importance of shared data to each IOC node is obtained through the following formula:

[0023]

[0024] In one embodiment of the present invention, when subscribing, the data is also scored according to the relevance and importance of the IOC node to perform subscription priority screening; wherein the relevance and importance of the data to the IOC node are scored by the following formula:

[0025] S = γ·R + δ·I;

[0026] In the formula, S is the priority subscription score of the data. The higher the score, the higher the subscription priority. γ is the data relevance weight, δ is the data importance weight, and γ+δ=1.

[0027] In one embodiment of the present invention, the intelligent scheduling and allocation of tasks includes:

[0028] Combine the data published by each IOC node and the data subscribed by each IOC node to calculate the load status of each IOC node;

[0029] According to the task set V and the correlation weight E between tasks, a task dependency graph G(V,E) is established;

[0030] Map nodes on the task dependency graph G(V,E), establish the objective function, and assign tasks to the IOC nodes with the lowest load;

[0031] The IOC node monitors its own load status in real time, and publishes the load status together with the collected data for sharing. In one embodiment of the present invention, the load status of each IOC node is calculated by the following formula:

[0032] L i =α·C i +β·N i ;

[0033] Where, L i is the load status of the ith IOC node, C i is the CPU utilization of the ith IOC node, N i is the network traffic of the i-th IOC node, and the weight parameters of α and β are in the range of [0,1];

[0034] And, the objective function of allocating tasks to the IOC node with the lowest load is as follows:

[0035]

[0036] Where F is the objective function and N is the total number of nodes.

[0037] In one embodiment of the present invention, the constraints when allocating tasks are: each task must be allocated to an IOC node; the load of the IOC node must not exceed a preset load threshold.

[0038] The present invention also provides a system using the above ECRH system distributed control and dynamic scheduling method thereof, comprising:

[0039] The distributed architecture module is used for the distributed control architecture based on EPICS, which divides the ECRH equipment into multiple independently operated IOC nodes, each of which is responsible for local data collection and control. After pre-processing the collected data, each IOC node calculates the priority score of each data according to the access frequency and real-time requirements, and marks the priority level of the data.

[0040] The sharing and collaboration module is used to publish and share data marked with priority levels using the CA protocol; each IOC node subscribes to the shared data based on its own relevance and importance to the data;

[0041] The response and task optimization module is used to dynamically adjust and control the given tasks based on the shared data and the data subscribed by each IOC node, so as to realize the intelligent scheduling and allocation of tasks.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] The current centralized architecture relies on a single control center to handle all data collection, analysis, and command generation. Since all nodes must transmit data to the control center for centralized processing and then send the results to each device, the overall data transmission and processing delay is high, which cannot meet the microsecond-level dynamic response requirements of the ECRH system.

[0044] The present invention adopts a distributed control architecture based on EPICS, dividing the equipment of the ECRH system into multiple independently operated IOC nodes, each of which completes local control and data acquisition. At the same time, data sharing and collaboration across nodes are realized through the CA protocol of EPICS, which greatly reduces data transmission delay. The parallel processing capability of the distributed architecture significantly improves the control response speed, enabling the system to meet microsecond-level dynamic control requirements.

[0045] In a centralized architecture, the coordination capabilities between subsystems (such as RF sources, cooling systems, and transmission antennas) are weak. Data transmission and control instructions between subsystems often need to be transferred through a centralized control center, causing communication bottlenecks and affecting the efficiency of real-time collaboration between nodes.

[0046] The present invention realizes data sharing and real-time collaboration among nodes through the CA protocol of EPICS. Each IOC node can directly publish or subscribe to key parameters such as reflected power and cooling flow without relying on the control center for transfer, thereby achieving efficient data synchronization and collaborative operation. Through this mechanism, each node can quickly adjust the control strategy to improve the overall operation efficiency.

[0047] The scalability of the centralized control architecture is limited. When new devices are added to the system or the amount of data increases significantly, the computing and communication load of the centralized control center will increase significantly, resulting in performance degradation or even system instability.

[0048] The present invention adopts a distributed architecture design, and each IOC node runs independently and processes specific equipment tasks. When the system is expanded, only the corresponding IOC nodes need to be added, avoiding the performance bottleneck of the centralized control center. At the same time, the independence between nodes ensures that the system can still maintain efficient and stable operation when expanded.

[0049] The centralized control architecture lacks dynamic scheduling capabilities for data processing and cannot classify data priorities according to real-time requirements. Low-priority data may occupy valuable transmission and processing resources, affecting the response time of critical data.

[0050] The present invention introduces dynamic priority scheduling and processes data in a hierarchical manner according to real-time requirements and access frequency. High-priority key data such as reflected power are transmitted and processed first, and low-priority data such as environmental monitoring parameters are processed with delay, thus optimizing the utilization of system resources.

[0051] The present invention has the following characteristics:

[0052] 1. Advantages of a single module: In the RF source IOC node, the high-speed data acquisition board PXI-6356 and localized processing significantly reduce data transmission delays, and increase the response speed from the millisecond level of the traditional architecture to the microsecond level. At the same time, the dynamic priority scheduling algorithm processes key data such as reflected power and standing wave ratio in real time to ensure the stability and accuracy of RF power output. In the cooling system IOC node, the cooling water flow and temperature are monitored in real time, and the thermal balance and stable operation of the equipment are ensured through intelligent adjustment. At the same time, the delay processing mechanism of low-frequency secondary data effectively reduces resource consumption. In the transmission antenna IOC node, power feedback data is collected in real time and the antenna direction is adjusted dynamically, which greatly improves the energy transmission efficiency. In the face of excessive node load or failure, the transmission task can be migrated to the backup node to further ensure the continuity of the experiment.

[0053] 2. The synergistic advantage of the combination between modules: The synergy of multiple IOC nodes significantly improves the overall performance of the system. Through the EPICS CA protocol and the optimized publish-subscribe mode, the data sharing mechanism reduces the transmission occupancy of irrelevant data and improves the efficiency of data sharing. The multi-node collaborative mechanism ensures that each node can respond quickly in critical scenarios. For example, when the RF source power is abnormal, the cooling system and the transmission antenna node can be adjusted in real time to ensure the safety of the equipment. The task allocation algorithm based on the task dependency graph and node load balances the system's computing and communication loads and avoids single-point bottlenecks. At the same time, the dynamic task adjustment mechanism significantly improves the stability and efficiency of task execution.

[0054] 3. Comprehensive advantages of the overall technology: Compared with the traditional centralized architecture, the distributed design and dynamic task optimization achieve higher real-time response capabilities and load balancing capabilities; compared with the basic EPICS framework, the introduced priority scheduling and screening algorithms greatly optimize the data sharing efficiency and the coordination capabilities between nodes. In addition, this architecture and algorithm are not only applicable to ECRH systems, but can also be extended to other industrial control scenarios that require real-time response and efficient coordination, such as intelligent manufacturing, nuclear fusion devices, and energy distribution networks, and have broad technical competitiveness and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 The present invention is a flow chart of a method for distributed control and dynamic scheduling of an ECRH system according to an embodiment of the present invention.

[0056] Figure 2 A block diagram of a distributed control and dynamic scheduling system of an ECRH system according to an embodiment of the present invention.

[0057] Figure 3 This is a block diagram of a distributed architecture module according to an embodiment of the present invention.

[0058] Figure 4 This is a block diagram of the sharing and collaboration module of an embodiment of the present invention.

[0059] Figure 5 This is a block diagram of a response and task optimization module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] In order to facilitate those skilled in the art to understand the technical solution of the present invention, the technical solution of the present invention is further described in conjunction with the accompanying drawings of the specification.

[0061] The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0062] See also Figure 1 As shown, the present invention provides a distributed control method of an ECRH system and a dynamic scheduling method thereof, comprising:

[0063] S10, based on the distributed control architecture of EPICS, divides the ECRH equipment into multiple independently operated IOC nodes, each of which is responsible for local data acquisition and control.

[0064] In one embodiment of the present invention, the EPICS-based distributed control architecture divides the physical devices of the ECRH system into multiple independently running IOC nodes, each of which is responsible for localized data acquisition and real-time control, thereby significantly improving the control response speed and meeting the needs of high-dynamic experiments in the ECRH system.

[0065] In one embodiment of the present invention, the plurality of independently operated IOC nodes include a radio frequency source IOC node, a cooling system IOC node, and a transmission antenna IOC node.

[0066] The RF source IOC node collects high-speed feedback signals, such as reflected power and standing wave ratio, and controls the power output of the RF source.

[0067] The IOC node of the cooling system collects cooling water flow and temperature signals and is responsible for dynamically adjusting the flow output of the cooling system.

[0068] The transmitting antenna IOC node collects antenna direction information and adjusts the antenna direction to optimize energy transmission.

[0069] S20, each IOC node pre-processes the collected data, calculates the priority score of each data according to the access frequency and real-time requirements, and marks the priority level of the data.

[0070] In one embodiment of the present invention, IOC node signal acquisition and preprocessing are completed by PXI series boards, for example, PXI-6259 is responsible for slow signal acquisition, such as temperature and flow of the cooling system, with a sampling rate of about 1kHz. PXI-6356 is responsible for high-speed signal acquisition, such as power feedback and standing wave ratio of the RF source, with a sampling rate of up to 100kHz. The signals collected by the board are transmitted to the Linux driver layer through a high-speed data channel. The Linux driver performs preliminary formatting on the collected raw data, such as removing invalid data and normalizing the signal amplitude to ensure that the data can be directly processed by the IOC node.

[0071] In one embodiment of the present invention, the collected data is graded according to the access frequency and real-time requirements: high-frequency key data, such as reflected power and standing wave ratio, are crucial to the real-time control of the system and have high evaluation values. They are marked as high priority and transferred to the real-time processing queue and directly processed by the corresponding IOC node. Low-frequency secondary data, such as cooling water temperature, have a low access frequency and short-time delays have little effect on control. They are marked as low priority and assigned to the low-priority queue to be processed when the system is idle. The allocation of dynamic priorities enables a rapid response to high-frequency data while reducing unnecessary processing delays.

[0072] In this embodiment, the priority score of each data is calculated, the data characteristics are periodically re-evaluated, and the priority allocation is updated. The data characteristics here refer to the access frequency and real-time requirements of the book data. And the priority score of each data is obtained by the following formula:

[0073] P=w t ·F t +w I I t ;

[0074] In the formula, P is the priority score of the data, F t For real-time requirements, I t is the access frequency, w t is the real-time demand weight parameter, w I is the access frequency weight parameter.

[0075] In this embodiment, real-time requirements Where T is the maximum allowed response time. Access frequency

[0076] When the priority score is greater than or equal to the high priority threshold, the data is marked as high-frequency critical data, high priority, and is placed in the real-time processing queue.

[0077] When the priority score is less than the high priority threshold, the data is marked as low-frequency secondary data, low priority, and is placed in the low priority queue.

[0078] In this embodiment, here w t and w I is the normalization parameter, w t Reflects the real-time requirements of parameters, which are directly related to the safety or stability of the equipment. I Reflects the frequency of parameter access and is related to data processing efficiency. t +w i =1,0≤w t , w i <<1.w t =1,w i =0 means the system focuses entirely on real-time performance, and the priority is determined only by the real-time requirement F t This is suitable for highly dynamic experimental scenarios, such as RF source power control. t =0,w i =1 means the system pays full attention to the access frequency, and the priority is completely determined by the data access frequency I t This is suitable for analytical tasks such as long-term trend analysis or archived data retrieval.

[0079] In this embodiment, the high dynamic scene recommendation w t >wi , analysis task recommendation w t <w i For initial value setting, high dynamic demand scenarios, w t =0.9, w i =0.1. Low dynamic demand scenario w t =0.4, w i =0.6. The specific value can be fine-tuned according to the actual situation.

[0080] S30, using the CA protocol to publish and share the data marked with priority levels; each IOC node subscribes to the shared data based on its own relevance and importance to the data.

[0081] In one embodiment of the present invention, data sharing and synchronization are important links in achieving node collaboration in a distributed control architecture. IOC nodes share device operating status and key parameters through the EPICS Channel Access (CA) protocol, and dynamically adjust load distribution based on task optimization, thereby significantly improving the efficiency of data sharing and the stability of system operation.

[0082] In this embodiment, data publishing: In the distributed control architecture based on EPICS, the data sharing mechanism is implemented through the publish-subscribe mode, and each IOC node publishes the key equipment operating parameters collected locally to the shared network in real time. Specifically, the RF source IOC node is responsible for collecting high-dynamic data such as reflected power and standing wave ratio, the cooling system IOC node collects environmental parameters such as cooling water flow and temperature, and the transmission antenna IOC node is responsible for collecting information such as antenna direction and power feedback. These data are broadcasted through the EPICS Channel Access (CA) protocol for other IOC nodes to subscribe as needed.

[0083] In this embodiment, data priority sorting: In order to improve the system's processing speed for important data, the concept of priority queue is introduced in the data sharing mechanism. According to the data label in the previous step, high-priority data such as reflected power and standing wave ratio data will be marked as high priority due to their high frequency of change and critical impact on system operation. When the data is released, real-time push and priority processing are ensured. The environmental monitoring data collected by the cooling system is classified as low priority due to its low update frequency and low real-time requirements. It is only transmitted and processed when the system is idle. Such a priority sorting mechanism ensures the rapid response of high-frequency critical data and the overall operating efficiency of the system while optimizing resource allocation.

[0084] In this embodiment, data subscription: each IOC node will selectively subscribe to highly relevant data published by other nodes according to the task requirements of its own functional module. For example, the cooling system IOC node subscribes to the reflected power data published by the RF source IOC node, and ensures the thermal balance and stable operation of the equipment by adjusting the cooling flow in real time. At the same time, the transmission antenna IOC node subscribes to the power feedback data of the RF source IOC node, optimizes the antenna direction based on real-time data, and thus improves the energy transmission efficiency.

[0085] In this embodiment, the correlation R between data and node tasks is usually determined by the weight in the task dependency graph, and the value range is 0 to 1. And the correlation between the shared data and each IOC node is obtained by the following formula:

[0086]

[0087] The importance of data I reflects the impact of data on the global system, which is set by historical statistics and domain knowledge, and ranges from 0 to 1. In addition, the relative importance of shared data to each IOC node is obtained by the following formula:

[0088]

[0089] In this embodiment, after subscribing to data related to itself, the subscribed data is also prioritized for subscription screening. Specifically, the priority subscription score of the data is calculated based on the relevance and importance of the data. The relevance and importance score of the data to the IOC node is calculated by the following formula:

[0090] S = γ·R + δ·I;

[0091] In the formula, S is the priority subscription score of the data. The higher the score, the higher the subscription priority. γ is the data relevance weight, δ is the data importance weight, and γ+δ=1.

[0092] The cooling system nodes give priority to subscribing to highly correlated reflected power data, reducing the consumption of bandwidth and resources by irrelevant data.

[0093] For example, in the application of subscription priority, the cooling system node preferentially subscribes to highly correlated reflected power data to reduce the consumption of bandwidth and resources by irrelevant data.

[0094] S40, based on the shared data and the data subscribed by each IOC node, combined with the given task, dynamic adjustment control is performed to achieve intelligent scheduling and allocation of tasks.

[0095] In one embodiment of the present invention, data collection and broadcasting: In a distributed control architecture, the implementation of collaborative control relies on data collection, sharing and dynamic adjustment strategies among the IOC nodes to ensure efficient operation and real-time responsiveness of the system. First, each IOC node periodically collects the operating status of the local device and broadcasts it to the shared network through the CA protocol. The RF source IOC node collects key data such as reflected power and standing wave ratio at high speed, and broadcasts it to the shared network in real time. The cooling system IOC node collects environmental parameters such as cooling water flow and temperature at a lower frequency, and also publishes it through the CA protocol. The transmission antenna IOC node collects antenna direction information and shares its status data to the network to provide real-time reference for other nodes.

[0096] In one embodiment of the present invention, the control strategy is dynamically adjusted: in actual operation, each IOC node will adjust its control strategy in real time according to the received data. For example, when the power fluctuation of the RF source exceeds the set threshold, the cooling system IOC node will respond quickly and dynamically adjust the flow output to prevent the equipment from overheating. If the antenna direction deviates from the target, the transmission antenna IOC node will optimize the antenna parameters in real time according to the feedback data to ensure the accuracy of beam transmission and energy utilization efficiency.

[0097] In one embodiment of the present invention, priority scheduling: To prioritize critical data, the system introduces a priority scheduling mechanism. According to the data tag, high-priority critical data such as reflected power is directly sent to the real-time processing queue to ensure fast response, while low-priority data such as cooling system temperature enters the delay queue and waits for the system to be idle before processing. This priority sorting greatly optimizes the utilization of system resources and the responsiveness to high-frequency critical tasks.

[0098] In one embodiment of the present invention, in order to achieve efficient data sharing and load balancing, the present invention adopts task optimization based on distributed collaboration, which can realize intelligent allocation and dynamic adjustment of tasks according to node load conditions and task dependencies. Specifically, the intelligent scheduling and allocation of tasks includes:

[0099] S41, combining the data published by each IOC node and the data subscribed by each IOC node, calculating the load status of each IOC node.

[0100] In this embodiment, the load status of each IOC node is expressed by the following formula:

[0101] L i =α·C i +β·N i ;

[0102] Where, L i is the load status of the ith IOC node, C iis the CPU utilization of the ith IOC node, N i is the network traffic of the i IOC node, and the weight parameters of α and β range from [0,1], which determines the influence of CPU and network traffic.

[0103] S42, establishing a task dependency graph G(V, E) according to the task set V and the correlation weights E between tasks.

[0104] In this embodiment, the task dependency graph describes the correlation between tasks, such as the coupling degree between the cooling task and the reflected power monitoring task. The task dependency graph G(V,E) includes a task set V, such as reflected power monitoring and cooling water flow adjustment, and a correlation weight E between tasks. For example, the correlation between the cooling task and the reflected power monitoring task is 0.8.

[0105] S43, perform node mapping on the task dependency graph G(V,E), establish the objective function, and assign the task to the IOC node with the lowest load.

[0106] In this embodiment, the objective function of allocating tasks to the IOC node with the lowest load is as follows:

[0107]

[0108] Where F is the objective function and N is the total number of nodes.

[0109] In this embodiment, the constraints for assigning tasks are: each task must be assigned to an IOC node; the load of the IOC node must not exceed the preset load threshold. Based on the current load distribution scheme, adjust the node allocation of some tasks to ensure load balancing: if the load of a node is too high, migrate some tasks to the backup node. If the network traffic is too large, adjust the task priority and reduce the frequency of non-critical tasks. If the task dependency changes, such as a sudden increase in the cooling system load, the task is reallocated.

[0110] S44, the IOC node monitors its own load status in real time, and publishes the load status together with the collected key data for sharing.

[0111] In this embodiment, each IOC node monitors its own load status in real time, such as CPU utilization, network traffic, etc., and publishes this load information together with the collected key parameters to the shared network. For example, when the RF source IOC node publishes the reflected power data, it also includes the occupancy information of the computing resources for reference by other nodes. After receiving this load information, the cooling system IOC node will dynamically adjust its subscription and task strategy. For highly correlated critical tasks, such as cooling tasks related to reflected power, the node will obtain them first and process them immediately. When the load of a node is too high, such as the RF source IOC node causing computing resources to be tight due to processing high-frequency power data, the cooling system IOC node can actively share part of the data processing tasks, or transfer non-critical tasks to spare nodes. This dynamic adjustment of task allocation effectively avoids load conflicts between nodes and improves the overall efficiency of task execution.

[0112] Please refer to 1 to Figure 5 As shown, the present invention also provides a distributed control system of an ECRH system and a dynamic scheduling system thereof, which applies the above-mentioned distributed control method of an ECRH system and a dynamic scheduling method thereof, including:

[0113] The distributed architecture module is used for the distributed control architecture based on EPICS, which divides the ECRH equipment into multiple independently running IOC nodes, each of which is responsible for local data collection and control. After pre-processing the collected data, each IOC node calculates the priority score of each data according to the access frequency and real-time requirements, and marks the priority level of the data.

[0114] The sharing and collaboration module is used to publish and share data marked with priority levels using the CA protocol; each IOC node subscribes to the shared data based on its own relevance and importance to the data.

[0115] The response and task optimization module is used to dynamically adjust and control the given tasks based on the shared data and the data subscribed by each IOC node, so as to realize the intelligent scheduling and allocation of tasks.

[0116] It is obvious to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, the embodiments should be regarded as exemplary and non-limiting from any point of view, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention, and any reference numerals in the claims should not be regarded as limiting the claims involved.

[0117] The above-described embodiments merely represent implementation methods of the invention. The protection scope of the present invention is not limited to the above-described embodiments. For those skilled in the art, several modifications and improvements may be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A distributed control and dynamic scheduling method of ECRH system, characterized in that: include: Based on the EPICS distributed control architecture, the ECRH equipment is divided into multiple independently operated IOC nodes, each of which is responsible for local data collection and control; After preprocessing the collected data, each IOC node calculates the priority score of each data based on the access frequency and real-time requirements, and marks the priority level of the data; The CA protocol is used to publish and share data with marked priority levels; each IOC node subscribes to the shared data based on its own relevance and importance to the data; Based on the shared data and the data subscribed by each IOC node, combined with the given tasks, dynamic adjustment control is performed to achieve intelligent scheduling and allocation of tasks.

2. The ECRH system distributed control and dynamic scheduling method according to claim 1 is characterized in that: The IOC nodes include the RF source IOC node, the cooling system IOC node and the transmission antenna IOC node.

3. The ECRH system distributed control and dynamic scheduling method according to claim 1 is characterized in that: Calculate the priority of each data using the following formula: P=w t ·F t +w I ·I t ; In the formula, P is the priority score of the data, F t For real-time requirements, I t is the access frequency, w t is the real-time demand weight parameter, w I is the access frequency weight parameter; When the priority score is greater than or equal to the high priority threshold, the data is marked as high-frequency critical data and placed in the real-time processing queue; When the priority score is less than the high priority threshold, the data is marked as low-frequency secondary data and placed in the low-priority queue; And, periodically re-evaluate data characteristics and update priority assignments.

4. The ECRH system distributed control and dynamic scheduling method according to claim 3 is characterized in that: Real-time demand weight parameter w t , access frequency weight parameter w I The value of is set according to the application scenario; High dynamic scene recommendation t >w I , analysis task recommendation w t <w I .

5. The ECRH system distributed control and dynamic scheduling method according to claim 1 is characterized in that: The relevance of the shared data to each IOC node is obtained by the following formula: The importance of shared data to each IOC node is obtained through the following formula:

6. The ECRH system distributed control and dynamic scheduling method according to claim 5 is characterized in that: When subscribing, the data is also scored based on its relevance and importance to the IOC node for subscription priority screening; the relevance and importance of the data to the IOC node is scored using the following formula: S = γ·R + δ·I; In the formula, S is the priority subscription score of the data. The higher the score, the higher the subscription priority. γ is the data relevance weight, δ is the data importance weight, and γ+δ=1.

7. The ECRH system distributed control and dynamic scheduling method according to claim 1 is characterized in that: Intelligent scheduling and allocation of tasks, including: Combine the data published by each IOC node and the data subscribed by each IOC node to calculate the load status of each IOC node; According to the task set V and the correlation weight E between tasks, a task dependency graph G(V,E) is established; Map nodes on the task dependency graph G(V,E), establish the objective function, and assign tasks to the IOC nodes with the lowest load; The IOC node monitors its own load status in real time and publishes the load status together with the collected data for sharing.

8. The ECRH system distributed control and dynamic scheduling method according to claim 7 is characterized in that: The load status of each IOC node is calculated by the following formula: L i =α·C i +β·N i ; Where, L i is the load status of the ith IOC node, C i is the CPU utilization of the ith IOC node, N i is the network traffic of the i-th IOC node, and the weight parameters of α and β are in the range of [0,1]; And, the objective function of allocating tasks to the IOC node with the lowest load is as follows: Where F is the objective function and N is the total number of nodes.

9. The ECRH system distributed control and dynamic scheduling method according to claim 7, characterized in that: The constraints when allocating tasks are: each task must be assigned to an IOC node; the load of the IOC node must not exceed the preset load threshold.

10. A system using the ECRH system distributed control and dynamic scheduling method according to any one of claims 1 to 9, characterized in that: include: The distributed architecture module is used for the distributed control architecture based on EPICS, which divides the ECRH equipment into multiple independently operated IOC nodes, each of which is responsible for local data collection and control. After pre-processing the collected data, each IOC node calculates the priority score of each data according to the access frequency and real-time requirements, and marks the priority level of the data. The sharing and collaboration module is used to publish and share data marked with priority levels using the CA protocol; each IOC node subscribes to the shared data based on its own relevance and importance to the data; The response and task optimization module is used to dynamically adjust and control the given tasks based on the shared data and the data subscribed by each IOC node, so as to realize the intelligent scheduling and allocation of tasks.

Images