Plasma Control System Test Platform, Test Method, and Storage Medium

Abstract

The present invention discloses a plasma control system test platform, a test method, and a storage medium, relating to the technical field of test systems. Specifically, it includes: a client, which is provided with a task submission module and an interactive visualization module; a server, a task upload module is used to receive tasks uploaded by the client, and after parsing the tasks, send them to a task queue scheduler. The task queue scheduler inserts or queues the uploaded tasks according to the priority; a multimodal actuator is connected to the PCS and the PCS-VP. The multimodal actuator obtains the task at the head of the task queue, and according to the task mode, starts the PCS process to execute the task or simultaneously starts the PCS and PCS-VP processes to execute the task; after the task execution is completed, the data analysis module obtains the original execution data, and feeds it back to the client interactive visualization module after data processing. It aims to realize the full-process dynamic interaction between the plasma control system and the master control system, and meet the complex requirements of modern experimental tests.

Description

Technical Field

[0001] The present invention relates to the technical field of testing systems, and in particular to a plasma control system testing platform, a testing method and a storage medium. Background Art

[0002] Plasma control systems play a crucial role in fusion reactions. Their primary task is to ensure stable fusion reactions by precisely controlling plasma parameters (such as temperature, density, and magnetic field). However, the high temperature, high energy, and complexity of plasma make the stability and reliability of the control system crucial to the success of fusion experiments. Therefore, improving the efficiency of plasma control system testing and verification, and ensuring its reliability in actual experiments, is a major challenge in current fusion experimental research.

[0003] Currently, traditional fusion experimental testing systems often rely on manual operations for testing and verification. During this process, testers need to manually modify the test configuration files and execute a series of test processes one by one. This manual operation method is not only cumbersome and time-consuming, but also easily affected by human factors, resulting in errors and uncertainties in the testing process, which in turn affects the accuracy and reliability of the test results. In addition, traditional testing methods usually lack the support of automated testing tools and rely on a single test mode. They are unable to establish full-process dynamic interaction between the plasma control system and the master control system, and are unable to monitor and adjust the interactive behavior of the system in real time. This limitation makes real-time feedback and adjustment during the test process difficult, and it is impossible to fully explore and solve potential system problems, which greatly limits the comprehensiveness and real-time nature of the test.

[0004] Therefore, how to achieve full-process dynamic interaction between the plasma control system and the master control system to meet the complex needs of modern experimental testing has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The main purpose of the present invention is to provide a plasma control system test platform, test method and storage medium, aiming to realize full-process dynamic interaction between the plasma control system and the master control system, and meet the complex needs of modern experimental testing.

[0006] In order to achieve the above objectives, the present invention proposes a plasma control system test platform, comprising:

[0007] The client is provided with a task submission module and an interactive visualization module. The task submission module is used to support single task form configuration and / or batch task file upload;

[0008] The server is provided with a task upload module, a task queue scheduler, a multimodal executor, and a data analysis module; the task upload module is used to receive tasks uploaded by the client, parse the tasks and send them to the task queue scheduler, and the task queue scheduler queues or inserts the uploaded tasks according to the priority; the multimodal executor is connected to PCS and PCS-VP, the multimodal executor obtains the task at the head of the task queue, and according to the task mode, starts the PCS process to execute the task or starts the PCS and PCS-VP processes simultaneously to execute the task; after the task execution is completed, the data analysis module obtains the original execution data, processes the data and feeds it back to the client interactive visualization module.

[0009] In an embodiment of the present application, queuing or inserting the uploaded tasks according to the priority includes:

[0010] The task queue scheduler obtains the task priority, determines whether the current priority is a privileged task, if it is a privileged task, queries whether there is an unfinished privileged task in the current queue; if there is an unfinished privileged task, inserts the new privileged task after the nearest privileged task; if there is no unfinished privileged task, inserts it at the head of the queue;

[0011] If it is not a privileged task, insert the task at the end of the queue, delete the task at the head of the queue after the task execution is completed and update the database status field.

[0012] In an embodiment of the present application, the task modes include: scenario playback mode and simulation test mode;

[0013] When the multimodal executor determines that the mode is the scenario playback mode, modify the PCS configuration file to mode, start the PCS process, call the simulated central control system module to generate a hardware trigger signal, and the signal outputs a high-level pulse through the NII / O card;

[0014] When the multimodal executor determines that the mode is the simulation test mode, modify the PCS configuration file to mode, start the PCS and PCS-VP processes; transmit the control signal to PCS-VP through shared memory, and receive the simulated acquisition signal and feed it back to PCS to form a closed loop.

[0015] In an embodiment of the present application, feeding back the processed data to the client interactive visualization module includes:

[0016] The data analysis module reads the experimental shot data and the historical reference shot data from the database storing the original execution data, identifies the flat-top segments in the experimental shot data and the historical reference shot data, performs waveform analysis of the flat-top segments and calculates the Euclidean distance of the boundary offset, stores the intermediate results as a target format file, records the path in the database test shot table, and feeds back the analysis results to the client's interactive visualization module, which displays a visualization chart.

[0017] In one embodiment of the present application, the client is further provided with a document generation module, which obtains the path of the interactive visualization module displaying the visualization chart, the analysis result summary, and the custom annotations, calls the general large language model to integrate the visualization chart, the analysis result summary, and the custom annotations to generate a structured document.

[0018] In one embodiment of the present application, the server is also provided with a status monitor module, which subscribes to the PCS health information published by the open source distributed control system framework, integrates the platform's own operating status, and pushes the integrated status data to the client through WebSocket, triggering real-time refresh of the client interface.

[0019] In one embodiment of the present application, the client is provided with an identity authentication module, and the server is provided with an identity verification module. The identity verification module receives the identity data of the identity authentication module. When the identity data is for a new user registration, the identity data is stored and the registration result is fed back. When the identity data is historical data, the identity data is compared with its own database and corresponding permissions are distributed.

[0020] In one embodiment of the present application, the permissions include: administrator, core developer, and ordinary developer.

[0021] This application also discloses a plasma control system testing method, comprising the following steps:

[0022] S1. The user submits a single task form configuration and / or batch task file through the client;

[0023] S2. The server verifies permissions and schedules the task to the queue.

[0024] S3. The multimodal executor dynamically adjusts the PCS operating parameters according to the task configuration;

[0025] S4. The status monitor collects PCS health data in real time and pushes it to the client;

[0026] S5. The document generation module processes the test results and generates visual charts and automated reports.

[0027] The present application also discloses a computer-readable storage medium storing a computer program, which when executed by a processor implements the steps of the method described above.

[0028] With the above technical solutions, the plasma control system test platform of the present invention can efficiently manage and execute the test tasks of the plasma control system by integrating multiple functions such as task submission, queue scheduling, task execution, and data analysis. Through clear functional division and close cooperation between the client and the server, users can conveniently submit tasks and view test results in real time. At the same time, the system can reasonably schedule according to the priority and type of tasks, thereby improving the test efficiency and resource utilization rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The present invention will be described in detail below with reference to specific embodiments and drawings, wherein:

[0030] Figure 1 is a schematic diagram of the system structure of the first embodiment of the present invention;

[0031] Figure 2 is a schematic diagram of the process structure of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0032] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be described in detail below with reference to the drawings and embodiments. It should be understood that the following specific embodiments are only used to explain the present invention and do not constitute a limitation to the present invention.

[0033] As Figure 1 shown, in order to achieve the above objectives, the present invention provides a plasma control system test platform, including:

[0034] A client, provided with a task submission module and an interactive visualization module, where the task submission module is used to support single-task form configuration and / or batch task file upload;

[0035] A server, provided with a task upload module, a task queue scheduler, a multi-modal executor, and a data analysis module; the task upload module is used to receive tasks uploaded by the client and send the parsed tasks to the task queue scheduler. The task queue scheduler queues or dequeues the uploaded tasks according to the priority; the multi-modal executor is connected to the PCS and the PCS-VP (Plasma Control System Virtual Platform). The multi-modal executor obtains the head task of the task queue and, according to the task mode, starts the PCS process to execute the task or simultaneously starts the PCS and PCS-VP processes to execute the task. After the task execution is completed, the data analysis module obtains the original execution data and feeds it back to the client interactive visualization module after data processing.

[0036] Specifically, the task submission module supports two task upload methods. The first method is single-task form configuration, where users fill in relevant task parameters such as test type and test mode through the client interface. The second method is batch task file upload. Users select a file containing multiple task configurations, and the client will perform format verification on the file to ensure that the file content conforms to the specified format. If the file format meets the requirements, the client will parse it into JSON format and upload it to the task upload module on the server side.

[0037] The interactive visualization module receives the data processed by the data analysis module and displays it graphically. The display content includes but is not limited to the execution status of test tasks, task results, etc. Users can view the task execution process and its results in real time in this module, and perform interactive operations such as zooming in, zooming out, or selecting specific data for detailed analysis.

[0038] The server side includes a task upload module, a task queue scheduler, a multi-modal executor, and a data analysis module.

[0039] The task upload module receives single tasks or batch task files uploaded from the client and parses them. For single-task form configuration, the task upload module parses the form content into JSON format data; for batch task file upload, the task upload module parses the file content and converts it into JSON format data, and then transmits the parsed task data to the task queue scheduler.

[0040] The task queue scheduler manages the task queue according to the priority of tasks. The priority of tasks is set by the client when uploading tasks and can be divided into high priority and normal priority. The task queue scheduler sorts the tasks according to the priority and decides whether to cut in line according to the priority of the tasks. If there is a high-priority task, the task queue scheduler will insert the task at the head of the queue; if it is a normal task, it will be inserted at the end of the queue.

[0041] The multi-modal executor is used to start and manage the task execution process of the plasma control system PCS and PCS-VP. The multi-modal executor obtains the task at the head of the task queue and decides whether to only start the PCS process to execute the task or start both the PCS process and the PCS-VP process to execute the task according to the task mode. If the task mode is scenario playback test, only the PCS process is started; if the task mode is simulation test, both the PCS process and the PCS-VP process are started. During the execution process, the coordination and data exchange between PCS and PCS-VP are completed through specific interfaces to ensure the synchronization and accuracy of task execution.

[0042] After the task execution is completed, the data analysis module retrieves the original execution data from the PCS or PCS-VP process. The data analysis module processes the retrieved data, including but not limited to data cleaning, computational analysis, result evaluation, etc. The processed data will be sent to the interactive visualization module of the client through the feedback mechanism for users to view and analyze.

[0043] With the above technical solution, the plasma control system test platform of the present invention can efficiently manage and execute the test tasks of the plasma control system by integrating multiple functions such as task submission, queue scheduling, task execution, and data analysis. Through clear functional division and close cooperation between the client and the server, users can conveniently submit tasks and view test results in real time. At the same time, the system can reasonably schedule according to the priority and type of tasks, thereby improving the test efficiency and resource utilization rate.

[0044] In an embodiment of the present application, inserting or queuing the uploaded tasks according to the priority includes:

[0045] The task queue scheduler obtains the task priority and determines whether the current priority is a privileged task. If it is a privileged task, it queries whether there are any unfinished privileged tasks in the current queue; if there are unfinished privileged tasks, the new privileged task is inserted after the nearest privileged task; if there are no unfinished privileged tasks, it is inserted at the head of the queue;

[0046] If it is not a privileged task, the task is inserted at the end of the queue. After the task execution is completed, the task at the head of the queue is deleted and the database status field is updated.

[0047] Specifically, the task queue scheduler first obtains the priority of the task, and the task priority is set by the client when submitting the task. The priorities are divided into two categories: "privileged tasks" and "non-privileged tasks". Privileged tasks represent urgent or high-priority test tasks, while non-privileged tasks are regular tasks with lower processing priorities.

[0048] The task queue scheduler determines whether the task is a privileged task based on the priority identifier of the task. If the current task is a privileged task, it enters the insertion process for privileged tasks; if the current task is not a privileged task, it enters the regular task processing process.

[0049] When the task queue scheduler determines that the current task is a privileged task, it first needs to query whether there are any unfinished privileged tasks in the task queue. Each task in the task queue records its execution status (such as "pending execution", "executing", "completed", etc.). Unfinished privileged tasks refer to tasks whose status is not "completed".

[0050] If there are unfinished privileged tasks in the task queue, the task queue scheduler will insert the newly submitted privileged tasks after the last unfinished privileged task in the sequential queue. This operation ensures that privileged tasks are executed in the order of submission, avoiding conflicts between different privileged tasks. If there are no unfinished privileged tasks in the queue, the new privileged task will be directly inserted at the head of the queue to ensure its priority over other tasks.

[0051] If the current task is a non-privileged task, the task queue scheduler inserts it at the end of the task queue. This ensures that regular tasks are executed in the order of submission but do not take precedence over privileged tasks. When the task at the head of the queue is completed, the task queue scheduler deletes it from the queue and updates the task status field in the database. The update operation of the status field includes marking the status of the task as "completed" and recording the execution time or other relevant information of the task. The operation of updating the database ensures the synchronization of the task execution history and the system state.

[0052] Adopting the above technical solution, the task queue scheduler can reasonably schedule tasks according to the priority of tasks. Privileged tasks can be executed first, and regular tasks are queued for execution. This scheduling method can not only ensure that urgent tasks are processed in a timely manner but also reasonably allocate system resources, improving the efficiency of task execution and the response ability of the system. Through the update of task status and reasonable queue management, the system can maintain a good task execution order and transparency of the running state, providing reliable support for subsequent data analysis and system optimization.

[0053] In an embodiment of the present application, the task mode includes: a scenario playback mode and a simulation test mode;

[0054] When the multimodal actuator judgment mode is the scenario playback mode, modify the PCS configuration file to mode, start the PCS process, call the simulated central control system module to generate a hardware trigger signal, and the signal outputs a high-level pulse through the NII / O card;

[0055] When the multimodal actuator judgment mode is the simulation test mode, modify the PCS configuration file to mode, start the PCS and PCS-VP processes; transmit the control signal to the PCS-VP through shared memory, and receive the simulated acquisition signal and transmit it back to the PCS to form a closed loop.

[0056] Specifically, the multimodal actuator determines the execution mode of the task according to the mode of the task at the head of the task queue obtained from the task queue. The task mode includes a scenario playback mode and a simulation test mode.

[0057] After the multimodal actuator obtains a task, it first identifies the mode of the task and selects different execution methods according to the mode.

[0058] If it is determined that the task mode is the scenario playback mode, the configuration file of the PCS needs to be modified first. The process of modifying the configuration file includes setting the operating mode of the PCS to " ", in this mode, the PCS will run historical data playback without real-time control. After the configuration file modification is completed, the multi-modal actuator starts the PCS process. In this mode, the PCS processes the playback historical data to simulate the process of historical experiments. To simulate the dynamic changes of the historical scenario, the multi-modal actuator calls the simulated central control system module to generate a hardware trigger signal. This hardware trigger signal is a high-level pulse signal output by the card. In this way, the control signal can interact correctly between the PCS and the hardware.

[0059] If the task mode is the simulation test mode, the multi-modal actuator will first modify the configuration file of the PCS and set its operating mode to " ". In this mode, the PCS and the PCS-VP will work together for simulation control and analog acquisition. After the configuration file modification is completed, the multi-modal actuator starts the PCS process and the PCS-VP process simultaneously. The PCS is responsible for generating control signals, while the PCS-VP is responsible for simulating the response of the plasma. To achieve real-time transmission of the control signal, the multi-modal actuator transfers the control signal generated by the PCS process to the PCS-VP through shared memory. The shared memory mechanism ensures timely and stable data exchange between the PCS and the PCS-VP. After the PCS-VP simulates and acquires signals, the simulated acquisition signals are transmitted back to the PCS to form a control closed-loop. In this way, the PCS can adjust the system according to the feedback signal of the PCS-VP to achieve closed-loop control of the test task.

[0060] It can be imagined that:

[0061] During the actual operation of the system, it will be affected by various disturbance factors. Random noise, offset error, and drift error are set to deal with the disturbance factors that appear in the system. It is used to study the anti-interference ability of the control system and the tolerance to signal errors.

[0062] In this application, an error injection module is designed. The error injection module includes three error types to simulate the operation of the PCS under different abnormal working conditions, namely random noise, offset error, and drift error. Among them, random noise corresponds to the unpredictable random fluctuations in the measurement signal, usually caused by thermal noise or environmental electromagnetic interference. Offset error corresponds to the constant deviation between the measured value and the true value, often caused by inaccurate sensor calibration or hardware aging. Drift error corresponds to the gradual deviation of the measurement system from the true value over time, usually caused by temperature changes or component aging.

[0063] In the random noise test, high-frequency random fluctuations are simulated by multiplying the signal by a random gain coefficient. The expression of the signal is:

[0064] ;

[0065] in, is the original signal, is the amplitude range of the random gain coefficient, is a random noise signal with a mean of 0 and a variance of 1. Random noise testing is used to evaluate the stability and anti-interference ability of a system under high-frequency random fluctuations.

[0066] The offset error test is performed by adding a fixed offset value to the signal. , its signal model is:

[0067]

[0068] in, is the original signal, A constant offset value is used to simulate measurement deviations caused by sensor calibration errors or hardware aging. This test verifies the system's control accuracy and error compensation capabilities under constant offset conditions.

[0069] The drift error test is performed by introducing an offset that gradually increases over time. , the mathematical model of the signal is:

[0070] ;

[0071] in, is the original signal, is a function that grows linearly with time and is usually expressed as:

[0072] ;

[0073] in, Drift error testing simulates the gradual deviation of a measurement system from its true value over long-term operation. The focus is on examining the system's ability to track gradual signal changes, as well as its adaptability and robustness over long-term operation. By testing these three error types, we can evaluate the performance of the Lingshu PCS under various abnormal operating conditions, providing a basis for optimized system design.

[0074] The error injection test is mainly divided into two modes: anti-interference ability test and control ability test.

[0075] During the anti-interference capability test, the PCS-VP reads historical discharge data from the database and converts it into 288 electromagnetic measurement signals through the ToRFM module. After adding disturbances through the error injection module, the signals are transmitted back to the RT node in the PCS through the RFMout module. The user can set a threshold and determine the algorithm's anti-interference capability by comparing the difference between the algorithm output after injection and before injection.

[0076] Control capability test: a closed control loop is formed between PCS and PCS-VP, the actual plasma is replaced by a simulation model, a control signal is input to PCS-VP, and PCS-VP outputs a simulated acquisition signal. The error injection module is used to modify the error injection signal and its intensity, which can help users test the control capability of Lingshu PCS under abnormal conditions.

[0077] During actual operation, plasma can be affected by a variety of abnormal factors, which can lead to plasma rupture. This rupture can cause serious consequences, such as device damage. Therefore, an exception handling mechanism is introduced into the PCS. To facilitate exception handling, an exception test unit is set up.

[0078] Definition of PCS abnormal time:

[0079] The plasma current anomaly is when the deviation between the measured plasma current and the target current exceeds a threshold.

[0080] A vertical displacement event occurs when the displacement growth rate of the plasma in the vertical direction exceeds a threshold.

[0081] Fragmentation prediction means that the system predicts that the plasma has a high possibility of fragmentation.

[0082] The poloidal field coil current anomaly is when the poloidal field coil current deviates from the target current beyond the allowable range.

[0083] Poloidal field coil overcurrent means that the current in the poloidal field coil exceeds the set maximum allowable value.

[0084] Power failure is to detect whether there is power abnormality or interruption through power signal.

[0085] The electromagnetic diagnostic signal is abnormal. The signal from the electromagnetic self-diagnosis system is abnormal.

[0086] The abnormal current of the RMP coil is caused by a large deviation between the actual current of the RMP coil and the target current.

[0087] RMP coil overcurrent means the current in the RMP coil exceeds the maximum set limit.

[0088] The first wall hot spot warning is when the surface temperature of the first wall component exceeds the warning threshold.

[0089] The first wall hot spot fault means that the surface temperature of the first wall component exceeds the fault threshold.

[0090] When the signal anomaly test unit in this application works, its principle is the same as that of the anti-interference ability test and the control ability test. Only need to replace the error injection module in the anti-interference ability test and the control ability test with the signal anomaly test module to work. Since it has been described above, it will not be elaborated here one by one.

[0091] Adopting the above technical solution, the multi-modal actuator can flexibly switch the execution mode according to different task modes. In the scenario playback mode, the system can accurately simulate the historical experimental process and trigger hardware signals to ensure the accuracy of the playback; in the simulation test mode, the PCS and PCS-VP can work together to simulate the response process of the plasma through the interaction of control signals and analog signals, ensuring the comprehensiveness and reliability of the test. Through the precise control of the task mode, the platform can effectively support various complex test requirements and improve the efficiency and accuracy of the test.

[0092] In an embodiment of this application, the data is processed and then fed back to the client interactive visualization module, including:

[0093] The data analysis module reads the experimental gun data and the historical reference gun data from the database storing the original execution data, identifies the flat top segments in the experimental gun data and the historical reference gun data, performs flat top segment waveform analysis and boundary offset Euclidean distance calculation, stores the intermediate results as target format files, records the paths in the database test gun table, and feeds back the analysis results to the interactive visualization module of the client, and the interactive visualization module displays the visualization charts.

[0094] Specifically, the data analysis module first reads the experimental gun data and the historical reference gun data from the database storing the original execution data. The experimental gun data refers to the data in the currently executed test task, while the historical reference gun data is the data used for comparison and evaluation. The data analysis module selects the relevant data from the test gun table in the database as needed.

[0095] After reading the experimental gun data and the historical reference gun data, the data analysis module will process these data and identify the flat top segments. The flat top segment is a specific time period in the plasma discharge process. The data analysis module detects and marks these data through specific algorithms (such as signal analysis, waveform matching, etc.) to identify the start time and end time of the flat top segment.

[0096] After identifying the flat top section, the data analysis module performs waveform analysis on the data of the flat top section. The purpose of this step is to compare the waveform characteristics in the experimental gun data and the historical reference gun data, and analyze the performance of both in the flat top section. The data analysis module calculates the shape, amplitude change, and other relevant characteristics of the waveform, quantifies these characteristics, and forms the analysis result.

[0097] The data analysis module also calculates the boundary offset between the experimental gun and the historical reference gun. To this end, the data analysis module uses the Euclidean distance algorithm to calculate the offset between them by comparing the boundary positions of the experimental gun data and the historical reference gun data in the flat top section. The calculation result is used to evaluate the deviation of the experimental data and the stability of the test task.

[0098] The analysis method is as follows:

[0099] Identify the start time of the flat top section and the end time , which mark the flat top section of the plasma. This part of the data will be used for subsequent offset calculation.

[0100] Process each time slice of the flat top section , and the range of the time slice is from to . Each time slice corresponds to a set of boundary point data.

[0101] For each time slice , extract the reference boundary point set and the boundary point set after error introduction .

[0102] : represents the set of boundary points of the reference boundary on the time slice , where each boundary point is represented by the coordinate , is the number of reference boundary points. represents the coordinate value of the reference boundary point on the x-axis, represents the coordinate value of the reference boundary point on the y-axis.

[0103] : represents the set of boundary points of the boundary with error on the time slice , where each boundary point is represented by the coordinate , is the number of error boundary points. represents the coordinate value of the error boundary point on the x-axis, represents the coordinate value of the error boundary point on the y-axis.

[0104] For the reference boundary and the boundary after error introduction, calculate the Euclidean distance between each pair of points.

[0105] For the reference boundary points , select the closest point from the set of boundary points after error introduction .

[0106] For each pair of matching reference boundary points and error boundary points, calculate the Euclidean distance between them , :

[0107]

[0108] where is the reference point and the error point between the Euclidean distance.

[0109] For a time slice , calculate the average offset after matching all reference boundary points and error boundary points :

[0110] ;

[0111] where is the average offset in the time slice , is the set of reference boundary points the number of boundary points.

[0112] After completing the flat-top segment waveform analysis and the calculation of the Euclidean distance of the boundary offset, the data analysis module stores the intermediate calculation results in the form of a target format file. These results include waveform analysis data, boundary offset data, and other relevant data. The storage path will be recorded in the test gun table of the database for subsequent query and access.

[0113] After completing the data processing, the data analysis module transmits the analysis results to the interactive visualization module on the client side through a feedback mechanism. The feedback data includes waveform analysis results, boundary offset analysis results, etc. The client generates corresponding visualization charts based on these data. The interactive visualization module can dynamically display these charts and support users to perform interactive operations on the data (such as zooming in, zooming out, viewing detailed data, etc.).

[0114] In the interactive visualization module on the client side, users will be able to view the charts generated based on the data analysis results. The charts can include flat-top segment waveform charts, boundary offset charts, etc., showing information such as the time variation of the data, feature comparison, and boundary changes. Through interactive operations, users can select different experimental gun and reference gun data for comparative analysis, facilitating real-time evaluation of the experimental results and making adjustments.

[0115] Using this technical solution, the data analysis module automatically processes and analyzes experimental shot data and historical reference shot data, accurately identifying flat-top segments and performing waveform analysis and boundary offset calculations. By storing and feeding these analysis results back to the client's interactive visualization module, users can view charts and data analysis results in real time. This visual display not only enhances the test platform's data analysis capabilities but also improves the user's intuitive understanding of data changes during mission execution, helping to further optimize test processes and experimental design.

[0116] In one embodiment of the present application, the client is further provided with a document generation module, which obtains the path of the interactive visualization module displaying the visualization chart, the analysis result summary, and the custom annotations, calls the general large language model to integrate the visualization chart, the analysis result summary, and the custom annotations to generate a structured document.

[0117] Specifically, the document generation module first obtains the path to the visualization charts displayed by the interactive visualization module. This path points to the chart files that users view on the client. These charts display the analysis results of the test task, such as the flat-top waveform chart and the boundary offset chart. The document generation module obtains the path information to these charts by interacting with the interactive visualization module's interface.

[0118] After obtaining the path to the visualization chart, the document generation module also needs to obtain the corresponding analysis result summary. This summary is a concise summary of the data analysis module's output, including an overview of the data analysis results, evaluation conclusions, and key findings. The analysis result summary provides essential context for the document's content, allowing users to quickly understand the key points of the analysis results.

[0119] In addition to the analysis result summary and chart paths, the documentation module also needs to capture user-defined annotations on the client side. These annotations can include user explanations of the test process, personal insights into the data analysis, or special points of interest regarding the results. Customized annotations provide personalized explanations and context, making the final documentation more comprehensive and tailored to user needs.

[0120] The document generation module integrates the acquired content using a general-purpose large language model (such as the DeepSeek series). The large language model first converts the visualization chart path, analysis result summary, and custom annotations into structured content. It then logically integrates this content into a structured document. The document generation process includes a detailed description of the analysis results, explanations of the charts, and the organization and integration of user-defined annotations.

[0121] The content integrated by the general large language model will be converted into a structured document. The structured document includes components such as titles, chapters, charts, text descriptions, and annotations, with a standardized format that is convenient for users to understand and further use. The document generation module ensures that the document content is well-organized and in a unified format, and embeds the charts and analysis results into the document in an appropriate manner for easy viewing and archiving by users.

[0122] The generated structured document will be stored on the server, and the client can download the document through the corresponding download interface. The document can be saved in PDF, Word, or other common formats, and users can download, store, or print it according to their needs.

[0123] Adopting the above technical solution, the document generation module can automatically integrate the charts displayed by the interactive visualization module, the summary of data analysis results, and user-defined annotations into a structured document. This process not only reduces the workload of manual document writing but also improves the efficiency and accuracy of document generation. Through the powerful integration ability of the general large language model, the generated document content has high quality and can meet the needs of users in aspects such as test result archiving and report generation. This automated document generation method enhances the user experience, enabling the test platform to not only provide real-time data analysis but also offer efficient document output support for users.

[0124] In an embodiment of the present application, the server is further provided with a status monitor module. The status monitor module subscribes to the PCS health information published by the open-source distributed control system framework, integrates the running status of the platform itself, and pushes the integrated status data to the client through WebSocket to trigger real-time refreshing of the client interface.

[0125] Specifically, the status monitor module first subscribes to and obtains the health information of the PCS (Plasma Control System) through the open-source distributed control system framework. The open-source distributed control system framework will regularly publish the health status data of the PCS, including but not limited to the running status of the system, the health of equipment, and the connection status. The status monitor module will receive the health data of the PCS through the framework interface to ensure real-time acquisition of the latest system health status.

[0126] In addition to subscribing to the PCS health information, the status monitor module also needs to integrate the running status of the platform itself. These status data include but not limited to the CPU usage rate, memory occupancy, disk space, and network connection status of the server. The status monitor module will regularly check and collect the running data of the platform itself and integrate it with the PCS health information. The integrated data can comprehensively reflect the health status of the system and the platform, ensuring the comprehensiveness and accuracy of monitoring.

[0127] After integrating the PCS health information and the platform running status, the status monitor module pushes the integrated status data to the client through WebSocket technology. WebSocket is a two-way communication protocol that allows the server and the client to establish a persistent connection, enabling the server to instantly push data to the client when the server status changes. The status monitor module uses WebSocket to push real-time data such as the health information and running status of the platform and PCS to the client, ensuring that users can see the running status of the system in real time on the client interface.

[0128] After the client receives the pushed status data through WebSocket, it triggers the real-time refresh of the interface. The interactive interface of the client updates the display content according to the latest status data received. For example, the client may display different warning or prompt messages according to the system health status, or dynamically display the usage of system resources. This enables users to instantly understand the running status of the system, discover potential system problems in a timely manner, and improve the maintainability and responsiveness of the system.

[0129] Adopting the above technical solution, the status monitor module can obtain the PCS health information and the running status of the platform itself in real time, and efficiently push this data to the client through WebSocket technology, triggering the real-time refresh of the interface. This real-time status monitoring method not only improves the monitoring ability of the platform, but also effectively helps users understand the system status in a timely manner, discover problems in advance and handle them. Through the real-time feedback of system health data, users can maintain full control of the platform status during the testing process, improving the reliability and maintenance efficiency of the system.

[0130] In an embodiment of the present application, the client is provided with an identity verification module, and the server is provided with an identity verification module. The identity verification module receives the identity data of the identity verification module. When the identity data is for new user registration, it stores the identity data and feedbacks the registration result; when the identity data is historical data, it compares the identity data with its own database and distributes corresponding permissions.

[0131] Specifically, the client is provided with an identity verification module. When a user conducts identity verification on the client, the user confirms the identity by inputting identity data such as a username and password. The identity verification module first checks the credential data input by the user. When the identity data input by the user has not been registered in the system, the identity verification module determines that the data is for new user registration. The system will guide the user into the registration process and require the user to provide necessary personal information (such as a username, password, email, etc.). When the identity data input by the user matches an existing user in the system, the identity verification module determines that the identity data is historical data and continues to process the login verification.

[0132] The server is equipped with an identity verification module, which receives the identity data transmitted from the client's identity authentication module. The identity verification module performs corresponding operations based on different identity data:

[0133] If the identity data transmitted by the identity authentication module is for new user registration, the identity verification module will receive the identity data and store it in the server's database. All user information is stored using encryption technology to ensure data security. Once the user registration is successful, at this time, the identity verification module will create the account information of the new user, store relevant data such as the username and password, and ensure the uniqueness and security of the user's identity in the system. After the storage process is completed, the identity verification module will feedback the registration result to the client, including whether the registration is successful and relevant prompts. If the registration fails, the client will prompt the user to check the input information and resubmit.

[0134] If the identity data transmitted by the identity authentication module is historical user data, the identity verification module will compare the transmitted identity data with the existing user data in its own database. The comparison content includes whether the username and password match, etc. If the comparison is successful, it means that the identity authentication is passed. The identity verification module will distribute corresponding permissions according to the user's role and permission information. These permissions determine the operation scope of the user in the system (such as administrator, core developer, ordinary developer, etc.). After the permission is assigned, the identity verification module will transmit the corresponding permission information back to the client, and the client controls the user's operation interface and function access according to the permission.

[0135] Adopting the above technical solution, the identity authentication and identity verification modules can effectively manage the identities and permissions of users. When a new user registers, the system can securely and conveniently store user data and feedback the registration result; when a historical user logs in, the system quickly verifies the user's identity by comparing the identity data with the database and assigns corresponding permissions according to the role. This mechanism not only improves the security and user experience of the system, but also ensures the rationality of permission allocation and the stable operation of the system. At the same time, the management and verification process of user identities conforms to the best practices of data protection, ensuring the security of user information.

[0136] In an embodiment of the present application, the permissions include: administrator, core developer, and ordinary developer.

[0137] Specifically, the administrator can add, delete, and modify user information, assign or adjust user roles. The administrator can create new user accounts, modify the permissions of existing users, or delete user accounts that are no longer needed. The administrator can view, modify, and delete all test tasks. The administrator can manually start, pause, or terminate test tasks and view the execution results of all tasks. The administrator can adjust the system configuration, such as platform parameter settings, test mode definitions, etc. These operations ensure that the platform can be flexibly configured according to actual needs.

[0138] Core developers can create, modify, and delete test tasks related to themselves. Core developers can execute and schedule test tasks, view the execution status and test results of tasks. Core developers have the privilege of jumping the queue in the task queue and can insert the test tasks they submit to the priority position in the task queue to ensure that urgent tasks can be processed first.

[0139] Ordinary developers can submit test tasks through the client, but cannot modify other tasks in the task queue. Ordinary developers can only submit and view tasks related to themselves, cannot jump the queue or modify the priority of tasks. Ordinary developers can view the execution status and results of the tasks they submit, but cannot view the test task data of other users.

[0140] With the above technical solutions, the permission management system can ensure that users of different roles obtain corresponding operation permissions according to their responsibilities. The administrator can comprehensively control the system. Core developers can flexibly manage test tasks and jump the queue for execution in case of emergencies, while ordinary developers can only perform basic operations such as task submission and status viewing. Through precise permission allocation and control, the system can effectively protect the security of user data, ensure the standardization of operations and the stable operation of the system. At the same time, this permission management mechanism enables the platform to flexibly adapt to different user needs, improving the maintainability and scalability of the system.

[0141] As Figure 2 shown, the present application also discloses a test method for a plasma control system, including the following steps:

[0142] S1. The user submits a single-task form configuration and / or a batch task file through the client;

[0143] S2. The server verifies the permissions and schedules the tasks to the queue;

[0144] S3. The multi-modal actuator dynamically adjusts the PCS operating parameters according to the task configuration;

[0145] S4. The status monitor real-time collects the PCS health data and pushes it to the client;

[0146] S5. The document generation module processes the test results and generates visual charts and automated reports.

[0147] Specifically, the user manually fills in the parameters of the test task through the graphical interface, such as the test type, test mode, required equipment, etc. After completion, the client packages these configuration parameters into a data file in JSON format and submits it to the server. The user can also choose to upload the task configuration file in batches, with each line in the file defining a complete set of test task configurations. The client will perform format verification on these files to ensure that the task format meets the requirements, and convert each task configuration item into JSON format and upload it to the server for processing.

[0148] After receiving the task data submitted by the client, the identity verification module on the server first verifies the identity of the submitting user. Based on the user's role in the system (such as administrator, core developer, or ordinary developer), the identity verification module determines whether the user has the permission to submit tasks and perform related operations. After successful identity verification, the task upload module will parse and transfer the task data to the task queue scheduler. The task queue scheduler will determine the insertion position of the task in the queue according to the priority of the task (set by the user when submitting). If the task is a privileged task, it may jump to the head of the queue, while regular tasks are queued according to the submission order.

[0149] The multimodal executor fetches the task to be executed from the task queue and reads the specific configuration of the task. According to the task configuration, the multimodal executor dynamically adjusts the operating parameters of the PCS (Plasma Control System). These parameters may include test mode, hardware configuration, simulation control parameters, etc. If the task configuration is in the scenario playback mode, the executor will adjust the configuration file of the PCS to mode, start the PCS process to playback historical experimental data; if the task configuration is in the simulation test mode, the executor will adjust the configuration file of the PCS to mode, and start both the PCS and PCS-VP processes for simulation testing and data collection.

[0150] The status monitor module on the server subscribes to the PCS health information published by the open-source distributed control system framework to obtain the health status of the PCS in real time. This data includes system operating conditions, hardware health conditions, connection status, etc. Once the health data is collected, the status monitor integrates this information and pushes it to the client through WebSocket. The client will receive the real-time updated PCS health data and refresh the interface. The user can view the system status in real time on the client interface to ensure the stability of the test.

[0151] After the test task is completed, the document generation module processes the generated test results, including obtaining experimental data and analysis result summaries from the database. The document generation module generates relevant visualization charts based on the analysis results, such as flat-top section waveform charts, boundary offset charts, etc. The charts will display the time variation of test data, key performance indicators, etc. By calling a general large language model, the document generation module automatically generates a test report according to the test results and analysis summaries. The report includes descriptions of the test process, result analysis, chart displays, etc., and the format of the report is a structured document, which is convenient for users to view, archive, and share.

[0152] Adopting the above technical solution, the plasma control system test method can comprehensively cover all aspects from task submission, permission verification, task execution, status monitoring to result processing. By dynamically adjusting the PCS operation parameters, the system can flexibly respond to different test requirements; through real-time health data collection and push, users can master the system status in real time to ensure the stability and reliability of the test; the document generation module greatly improves work efficiency and reduces manual operation errors by automatically generating test reports and visualization charts. The implementation of this test method makes the test process more efficient and accurate, providing strong tool support for developers.

[0153] This application also discloses a computer-readable storage medium storing a computer program, and the steps of the above method are implemented when the program is executed by a processor.

[0154] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structural transformation made by using the description and drawings of the present invention under the inventive concept of the present invention, or direct / indirect application in other related technical fields, is included in the patent protection scope of the present invention.

Claims

1. A plasma control system test platform, characterized in that Including: A client, equipped with a task submission module and an interactive visualization module. The task submission module is used to support single-task form configuration and / or batch task file upload; A server, equipped with a task upload module, a task queue scheduler, a multi-modal executor, and a data analysis module. The task upload module is used to receive tasks uploaded by the client, parse the tasks, and send them to the task queue scheduler. The task queue scheduler inserts or queues the uploaded tasks according to the priority; The multi-modal executor is connected to PCS and PCS-VP. The multi-modal executor obtains the task at the head of the task queue and, according to the task mode, starts the PCS process to execute the task or starts the PCS and PCS-VP processes simultaneously to execute the task; After the task execution is completed, the data analysis module obtains the original execution data, processes the data, and feeds it back to the client interactive visualization module. The task mode includes: scenario playback mode and simulation test mode; When the multimodal actuator judgment mode is the scenario playback mode, modify the PCS configuration file to mode, start the PCS process, call the simulated central control system module to generate a hardware trigger signal, and the signal outputs a high-level pulse through the NII / O card; when the multimodal actuator judgment mode is the simulation test mode, modify the PCS configuration file to mode, start the PCS and PCS-VP processes; transmit the control signal to the PCS-VP through shared memory, and receive the simulated acquisition signal and return it to the PCS to form a closed loop.

2. The plasma control system test platform according to claim 1, wherein Inserting or queuing the uploaded tasks according to the priority includes: The task queue scheduler obtains the task priority, determines whether the current priority is a privileged task. If it is a privileged task, it queries whether there is an unfinished privileged task in the current queue. If there is an unfinished privileged task, the new privileged task is inserted after the nearest privileged task. If there is no unfinished privileged task, it is inserted at the head of the queue; If it is not a privileged task, the task is inserted at the end of the queue. After the task execution is completed, the task at the head of the queue is deleted and the database status field is updated.

3. The plasma control system test platform according to claim 1, characterized in that, Feeding the processed data back to the client interactive visualization module includes: The data analysis module reads the experimental gun data and historical reference gun data from the database storing the original execution data, identifies the flat top segments in the experimental gun data and historical reference gun data, performs flat top segment waveform analysis and boundary offset Euclidean distance calculation, stores the intermediate results as target format files, records the paths in the database test gun table, and feeds the analysis results back to the interactive visualization module of the client. The interactive visualization module displays visualization charts.

4. The plasma control system test platform according to claim 3, characterized in that, The client also has a document generation module. The document generation module obtains the path of the visualization chart displayed by the interactive visualization module, the analysis result summary, and custom annotations, calls a general large language model to integrate the visualization chart, the analysis result summary, and the custom annotations, and generates a structured document.

5. The plasma control system test platform according to claim 1, wherein The server also has a status monitor module. The status monitor module subscribes to the PCS health information published by the open source distributed control system framework, integrates the running status of the platform itself, and pushes the integrated status data to the client through WebSocket to trigger real-time refresh of the client interface.

6. The plasma control system test platform according to claim 1, wherein, The client has an identity authentication module, and the server has an identity verification module. The identity verification module receives the identity data of the identity authentication module. When the identity data is for new user registration, it stores the identity data and feeds back the registration result. When the identity data is historical data, it compares the identity data with its own database and distributes corresponding permissions.

7. The plasma control system test platform according to claim 6, characterized in that, The said permissions include: administrator, core developer, and ordinary developer.

8. A method for testing a plasma control system, which is used for the plasma control system test platform according to any one of claims 1 to 7, characterized in that, It includes the following steps: S1. The user submits a single-task form configuration and / or a batch task file through the client. S2. The server verifies the permissions and schedules the tasks to the queue. S3. The multi-modal executor dynamically adjusts the PCS operating parameters according to the task configuration. S4. The status monitor collects the PCS health data in real time and pushes it to the client. S5. The document generation module processes the test results and generates visual charts and automated reports.

9. A computer-readable storage medium storing a computer program, characterized in that, When the said program is executed by the processor, it implements the steps of the method as described in claim 8.

Images