A job method and system of a virtual simulation training platform

Abstract

This invention relates to the field of virtual device interaction, and provides a method and system for operating a virtual simulation training platform, comprising: if the server determines that the client is sending an interaction request for the first time, adding a row-level exclusive lock to the ownership record of the target operation step, the row-level exclusive lock indicates that no reading or modification operation can be performed on the ownership record; the server determines the ownership status of the target operation step, the ownership status indicating whether the target operation step has been claimed by other clients; if the server determines that the ownership status of the target operation step indicates that the target operation step has not been claimed by other clients, the server releases the row-level exclusive lock of the target operation step and locks the target client and the target operation step; if the server determines that the ownership status of the target operation step indicates that the target operation step has been claimed by other clients, the server rejects the interaction request, thereby ensuring that the same operation step is executed by only one client at any given time, eliminating the problem of ambiguous operation ownership in group assessments.

Description

Technical Field

[0001] This invention relates to the field of virtual device interaction technology, and in particular to a method and system for operating a virtual simulation training platform. Background Technology

[0002] In the practical training and job assessment system for power generation operation and maintenance personnel in the power industry, multi-person collaborative simulation training has become an important means of competency assessment. It usually requires multiple trainees to form an assessment group and complete collaborative tasks with strong time constraints and strong interactive coupling characteristics, such as switching operations, equipment start-up and shutdown, and fault diagnosis, in the same virtual equipment environment according to the preset standard operating procedures. This is to simulate the working mode of multi-role cooperation in a real power plant, and to make a comprehensive assessment of the operational capabilities of each group member. However, existing multi-user virtual simulation systems generally use a collaborative mechanism centered on device state synchronization at the underlying architecture level. The server only maintains the global state vector of the virtual device and synchronizes the state changes to each client in the form of broadcast. It lacks system-level management of the mapping relationship between operation steps and executors. On the one hand, the system does not set a unique execution ownership field for each operation step. Multiple students can concurrently initiate interaction requests to the virtual devices involved in the same operation step. The server processes them uniformly according to the order of message arrival and does not exclusively bind and lock the identity of the operation initiator. On the other hand, when recording device state changes, the server usually only records the state values ​​before and after the change and the time of occurrence, without strongly associating the state change with the specific operation subject, resulting in the loss of operation behavior ownership information at the data level. Summary of the Invention

[0003] To address the aforementioned problems in the prior art, this invention provides a method and system for operating a virtual simulation training platform, which solves the problem of unclear attribution of operation behaviors when multiple trainees simultaneously interact with the same equipment involved in the same operation step.

[0004] According to one aspect of the present invention, a method for operating a virtual simulation training platform is provided, comprising: The server obtains the operation tasks in the rendered virtual space; wherein, the operation tasks include multiple operation steps; Multiple clients send interaction requests; wherein, the interaction request indicates a request to interact with the virtual device corresponding to the target operation step among the multiple operation steps; The server receives the interaction request and determines whether the client is sending the interaction request for the first time. If the server determines that the client is sending the interaction request for the first time, it adds a row-level exclusive lock to the record to which the target operation step belongs; wherein, the row-level exclusive lock means that no read or modification operation can be performed on the record to which the target operation step belongs; The server determines the ownership status of the target operation step; wherein, the ownership status indicates whether the target operation step has been claimed by other clients; If the server determines that the ownership status of the target operation step indicates that the target operation step has not been claimed by other clients, it releases the row-level exclusive lock of the target operation step and locks the target client with the target operation step. If the server determines that the ownership status of the target operation step indicates that the target operation step has been claimed by another client, then it will reject the interaction request.

[0005] According to another aspect of the present invention, a working system for a virtual simulation training platform is provided, comprising: a server for acquiring an operation task in a rendered virtual space; wherein the operation task includes multiple operation steps; receiving an interaction request and determining whether a client is sending the interaction request for the first time; if it is determined that the client is sending the interaction request for the first time, then adding a row-level exclusive lock to the ownership record of the target operation step; wherein the row-level exclusive lock indicates that no read or modification operation can be performed on the ownership record; determining the ownership status of the target operation step; wherein the ownership status indicates whether the target operation step has been claimed by other clients; if it is determined that the ownership status of the target operation step indicates that the target operation step has not been claimed by other clients, then releasing the row-level exclusive lock of the target operation step and locking the target client with the target operation step; if it is determined that the ownership status of the target operation step indicates that the target operation step has been claimed by other clients, then rejecting the interaction request; and multiple clients for sending interaction requests; wherein the interaction request indicates a request to interact with the virtual device corresponding to the target operation step among the multiple operation steps.

[0006] The beneficial effects of this invention are reflected in the fact that it provides a method and system for operating a virtual simulation training platform, including a server acquiring an operation task in a rendered virtual space, wherein the operation task includes multiple operation steps, multiple clients sending interaction requests, wherein the interaction request indicates a request to interact with the virtual device corresponding to the target operation step among the multiple operation steps, the server receiving the interaction request and determining whether the client is sending the interaction request for the first time, if the server determines that the client is sending the interaction request for the first time, then adding a row-level exclusive lock to the ownership record of the target operation step; wherein the row-level exclusive lock indicates that no read or modification operation can be performed on the ownership record, the server determining the ownership status of the target operation step, wherein the ownership status indicates whether the target operation step has been claimed by other clients, if the server determines that the ownership status of the target operation step indicates that the target operation step has not been claimed by other clients, then releasing the row-level exclusive lock of the target operation step and locking the target client and the target operation step, if the server determines that the ownership status of the target operation step indicates that the target operation step has been claimed by other clients, then rejecting the interaction request. At any given time, the same operation step can be led and executed by at most one trainee. The operation behavior corresponds one-to-one with the trainee's identity, eliminating the problem of unclear operation attribution in group assessments and enabling the assessment results to objectively reflect each group member's actual contribution to the collaborative task. Attached Figure Description

[0007] Figure 1 This is a flowchart illustrating the operating method of the virtual simulation training platform provided by the present invention.

[0008] Figure 2 This is a flowchart illustrating the assessment enable vector display method provided by the present invention.

[0009] Figure 3 This is a flowchart illustrating the virtual space rendering method provided by the present invention.

[0010] Figure 4 This is a flowchart illustrating the operation method of a virtual simulation training platform provided in another embodiment of the present invention.

[0011] Figure 5 This is a schematic diagram of the operating system of the virtual simulation training platform provided by the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0013] The virtual simulation training platform is a multi-user online virtual simulation training platform for the industrial operation and maintenance field (especially power generation equipment operation and maintenance). It supports smooth operation on mobile devices and enables instructors to conduct training demonstrations and collaborative assessments for trainees. The platform uses virtual rooms to facilitate communication between instructors and trainees. Instructors can create collaborative and assessment spaces according to their teaching needs. Collaborative spaces are divided into live streaming rooms and collaborative rooms. Assessment spaces are divided into individual assessment rooms and group assessment rooms. In live streaming rooms, the host's operation screen is synchronized one-way in real time to other members in the room, who are in a read-only viewing state. In collaborative rooms, all members' operations on the virtual device and the resulting changes in device status are synchronized bi-directionally in real time to all online members' local scenes, enabling multi-user system operation within the same virtual device environment. In individual assessment rooms, each trainee answers questions in an independent virtual scene copy, without affecting others. In group assessment rooms, all members share the same virtual device status. Operations completed or in progress by one member are locked and inoperable in other members' interfaces to prevent duplicate execution and ensure the uniqueness of each group member's operation.

[0014] The room maintains order and a secure assessment environment through mechanisms such as a six-digit password, access control for assessments, reconnection after disconnection, and forced removal. In terms of training content, the platform covers core power generation equipment such as speed controllers, circuit breakers, transformers, generator main shafts, and wind turbines, as well as general industrial components such as switches, valves, instruments, cabinet doors, and tools. For multi-person collaboration, the platform supports real-time voice communication with fine-grained microphone access control, text recording of scheduling reports and parameter readings, bidirectional synchronization of device status across terminals, and parallel monitoring of multiple trainees' 3D screens by the instructor. The assessment process supports one-click switching between training and assessment modes. Furthermore, the platform offers on-demand downloads, incremental updates, and one-click cache clearing functions for each course.

[0015] Figure 1 This is a flowchart illustrating the operating method of the virtual simulation training platform provided by the present invention. Figure 1 As shown, the operational methods of the virtual simulation training platform include: S110: The server obtains the operation tasks in the rendered virtual space, where the operation tasks include multiple operation steps.

[0016] In this embodiment of the invention, when there is multi-user collaboration, such as in a group assessment room, all group members share the same device state vector and need to collaborate to complete the same operation task. However, because multiple students simultaneously initiate interactions with the devices involved in the same operation step, the attribution of the operation behavior is unclear. The server cannot determine which student led the completion of a certain device state change, which in turn leads to inaccurate group assessment results and an inability to objectively evaluate.

[0017] The virtual space is constructed from three-dimensional virtual scenes.

[0018] S120: Multiple clients send interaction requests, where each interaction request represents a request to interact with the virtual device corresponding to the target operation step in multiple operation steps.

[0019] In this embodiment of the invention, when a student performs an operation on a virtual device in a virtual space, the client encapsulates the student's operation intention into an interaction request and sends it to the server, thereby clarifying the target operation step requested by the student and the virtual device corresponding to the target operation step in multiple operation steps.

[0020] S130: The server receives the interaction request and determines whether the client is sending the interaction request for the first time.

[0021] In this embodiment of the invention, the server maintains an independent ownership status for each operation step and sets the initial status of each operation step to unclaimed. When a student attempts to interact with the device involved in the target operation step for the first time, the server and the ownership of that step are automatically assigned to the student. Subsequent interaction requests from other students to the device involved in the same step are silently intercepted by the server, and the corresponding operation area on the client interface is locked and cannot be operated.

[0022] S140: If the server determines that the client is sending an interaction request for the first time, it adds a row-level exclusive lock to the record to which the target operation step belongs. The row-level exclusive lock means that no read or modification operation can be performed on the record to which the target operation step belongs.

[0023] In this embodiment of the invention, the core function of row-level exclusive locks is to ensure that only one client can process a specific data record at any given time when multiple users operate on the same data simultaneously. This avoids problems such as data conflicts, state confusion, or unclear operation attribution caused by multiple users writing simultaneously. Furthermore, once a client adds a row-level exclusive lock to a data record, no other client can read or modify that record until the lock is released; they must wait in a queue. The row-level characteristic of the lock means that it only locks this specific record and does not affect other records in the same data table. Therefore, it ensures the consistency of critical data without excessively blocking the overall concurrency performance of the system. In multi-user collaborative virtual simulation assessment scenarios, this locking mechanism ensures that only one student can lead the execution of the same operation step at any given time.

[0024] Furthermore, upon receiving an interaction request from the client, the server first determines whether the request is the first one arriving under the current operation step. If the server confirms that the client is the first student to initiate an interaction for the target operation step (i.e., no group member has previously claimed this step), it immediately adds a row-level exclusive lock to the ownership record corresponding to the target operation step in the database. Through this mechanism, the server can lock the ownership determination process of the step the moment concurrent requests flood in, ensuring that only the student who first initiated the request can successfully complete the step claiming or binding. Subsequent requests from other members for the same step will be naturally blocked because they cannot access the ownership record, thus fundamentally preventing conflicts caused by multiple people operating the same step simultaneously, clarifying the unique ownership of the step, and ensuring the orderliness and traceability of operational behavior during the group assessment process.

[0025] In this context, "claim" means that the server exclusively assigns the target operation steps to the client that first initiates the interaction request.

[0026] S150: The server determines the ownership status of the target operation step, where ownership status indicates whether the target operation step has been claimed by other clients.

[0027] S160: If the server determines that the ownership status of the target operation step indicates that the target operation step has not been claimed by other clients, it releases the row-level exclusive lock of the target operation step and locks the target client and the target operation step.

[0028] S170: If the server determines that the ownership status of the target operation step indicates that the target operation step has been claimed by another client, then it will reject the interaction request.

[0029] In this embodiment of the invention, if the ownership status of the target operation step indicates that the target operation step has been claimed by another client, a conflict notification is returned to the client. Before the transaction executed on the server side, which adds a row-level exclusive lock to the ownership record and writes the claim information, is committed, all other concurrent requests for the same step will be blocked and queued. Only after the transaction is committed and the row-level exclusive lock is released will these requests be allowed to enter the ownership determination process in sequence.

[0030] Furthermore, to ensure that the virtual scene state observed by all members in a room remains synchronized in a multi-user collaborative scenario, multi-device synchronization is achieved through a server-side mechanism. The server centrally maintains the state vectors of multiple virtual devices in the virtual scene, recording the current operating state of each device, including switch positions, valve angles, and instrument readings. Each client only stores a local copy of the state to drive 3D scene rendering and does not have the ability to independently change the global state. Any operation initiated by a client on any virtual device must be verified by the server before it takes effect. The server sends this state change as an incremental message to all members in the room. This incremental message is constructed from the number of virtual devices that have changed, their new operating state, and a monotonically increasing global logical timestamp assigned by the server for each accepted operation. Each client updates its local rendering state sequentially according to the global logical timestamp order, thus ensuring eventual consistency of device states across multiple devices even under conditions of concurrency and network fluctuations.

[0031] In this process, once the client successfully claims the target operation step and obtains exclusive operation rights, the server first formally accepts the operation on its own side, semantically changing the maintained device state, such as flipping a switch from open to closed, or adjusting the valve opening from a smaller setting to a larger setting, and writing these changes into the global state vector. The state vector is the complete set of the latest operating states of all virtual devices in the room, centrally maintained by the server. Each write corresponds to a deterministic change in the global state of the virtual scene. After writing, the server does not reissue the entire state list, but only sends the minimum set of changes caused by this operation, i.e., an incremental message with a global logical timestamp, to all members. The client partially refreshes the local 3D scene and device state according to the global logical timestamp order, so that the local state copies of all members in the room converge to the same values ​​as the server in a very short time, thus providing a stable state foundation for subsequent step completion judgment, log solidification, and assessment attribution traceability.

[0032] Figure 2 This is a flowchart illustrating the assessment enable vector display method provided by the present invention. Figure 2 As shown, the operational methods of the virtual simulation training platform may also include: S180: The server obtains the mapping function corresponding to the virtual space of the group assessment type. The mapping function has an assessment enable vector. Each display enable bit in the assessment enable vector indicates whether the guidance information is activated and displayed when the current frame is rendered in the virtual space. The guidance information is visual prompts that help trainees locate the operation target in the virtual space.

[0033] In this embodiment of the invention, since the three-dimensional operation scenario in the virtual space may be simultaneously applied to two drastically different application scenarios—teaching and training, and competency assessment—multi-dimensional operation guidance is needed for teaching and training to reduce cognitive bias and accelerate skill acquisition. For competency assessment, all external guidance must be shielded to objectively evaluate the trainee's true mastery of power generation operation and maintenance procedures. If the two scenarios share the same guidance information display logic, the assessment process will lose objectivity due to residual prompts.

[0034] Therefore, this application constructs a mapping function to manage the mapping relationship between the operation mode (corresponding to the type of virtual space) and the display state of the boot information. When switching the operation mode, the client updates the target configuration with all enable bits at the beginning of the next frame. All boot elements synchronously read the new values ​​in the same frame without generating intermediate states.

[0035] Furthermore, the mapping function has an assessment enable vector, where the value of each display enable bit indicates whether the guidance information is activated and displayed during the rendering of the current frame in virtual space. The display enable bits in the assessment enable vector represent different types of guidance information, including device space highlighting prompts corresponding to the current target operation step, tool selection prompts, text descriptions of the target operation step, and a banner at the top when the target operation step is completed.

[0036] S190: The client will sequentially activate the guidance information corresponding to the assessment enable vector during the rendering of the current frame, so as to display it in virtual space.

[0037] In this embodiment of the invention, during the frame-by-frame rendering of the scene in the virtual space by the client, at the rendering time of the current frame, various types of guidance information are activated sequentially according to the value of each bit in the assessment enable vector. The activated guidance information is drawn and superimposed on the corresponding position in the virtual space and presented to the user. The inactive guidance information does not participate in the drawing of this frame. By uniformly activating the guidance information according to the vector value at the beginning of the frame, the display status of various types of guidance information can be switched and presented synchronously within the same frame, avoiding intermediate transitional screens where some guidance is displayed and some is not, thereby ensuring that the guidance information presented in the virtual space is always strictly consistent with the current assessment mode.

[0038] When the display enable bit is 1, the guidance information is activated and displayed during the rendering of the current frame. When the display enable bit is 0, the guidance information is not displayed during the rendering of the current frame.

[0039] In one embodiment, S110 can be specifically implemented as follows: obtaining a global mode state variable; wherein the global mode state variable represents the current operation mode; determining the operation mode of the virtual space based on the global mode state variable; wherein the operation mode corresponds to the type of the virtual space; and obtaining the operation task in the virtual space based on the operation mode of the virtual space.

[0040] In this embodiment of the invention, during the actual operation of the virtual simulation training platform, the same virtual scene often needs to serve multiple purposes such as training demonstrations, collaborative exercises, individual assessments, and group assessments. The operational tasks that the system needs to load, the guidance and prompting strategies that should be enabled, and the operational constraints that should be applied vary significantly depending on the purpose. If the client cannot clearly understand the current operating mode when entering the virtual space, problems such as incorrect task loading, mismatch between guidance information and training requirements, and misalignment of mutual exclusion control strategies can easily occur, thus affecting the trainees' training experience and the effectiveness of the assessment results.

[0041] Therefore, this invention uses a global mode state variable to represent the current operation mode. The client determines the current operation mode of the virtual space based on the value of this global mode state variable. A one-to-one correspondence is pre-established between the operation mode and the type of virtual space, so the operation mode can be directly mapped to the corresponding virtual space type (such as training type, collaborative type, individual assessment type, or group assessment type). Based on the determined operation mode, the client obtains the operation task matching the mode from the corresponding task resources and loads it into the virtual space for trainees to execute.

[0042] Meanwhile, before the introduction of global mode state variables, the rendering layer typically faced two prominent problems: First, if a certain type of guidance information needed to be exceptionally retained in the assessment mode, the exception logic had to be hard-coded separately in the rendering code of that element. Each additional differentiated requirement required an additional exception code, and the system maintenance cost increased linearly with the number of guidance types. Second, global state variables could be modified at any time during the rendering process. If a frame had already been rendered halfway through when the mode switching instruction arrived, different guidance elements within that frame would read state values ​​from different moments, resulting in intervening frame ghosting where some elements had switched and others had not.

[0043] Therefore, this invention abstracts each type of guidance information into an independent display enable bit in the enable vector, and uses a mapping function to centrally express the correspondence between modes and enable vectors. The rendering code only needs to uniformly read the enable vector and activate the corresponding guidance element bit by bit, without needing to be aware of any specific mode semantics. When adding or adjusting the differentiated requirements of a certain type of guidance information, it is only necessary to extend the corresponding bit in the enable vector and add or modify a rule in the mapping function. The rendering code does not need to be modified, thus converging the hard-coded exceptions that were originally scattered in various rendering elements into a centralized and configurable rule table, making the system maintenance cost change from linearly increasing with the number of guidance types to being basically constant. Furthermore, the effective time of mode switching is strictly constrained to the start of the frame. Even if the global mode state variable is modified in the middle of rendering a frame, the rendering process will not immediately apply the change, but will delay the new value to the start of the next frame. The mapping function atomically generates a new enable vector and uniformly applies it to the activation judgment of all guidance information in the current frame. Since all guidance information within the same frame is always rendered based on the same enable vector, there will be no intermediate state where some elements have been switched and some have not, thus fundamentally eliminating the ghosting phenomenon at the moment of mode switching.

[0044] For example, let the global mode state variable be... Its range is Where M=0 represents the training mode and M=1 represents the assessment mode. The system defines the following: Class guidance information, denoted as The device space highlighted in the current step. (Red flashing rendering), tool selection prompts (Toolbox flashing yellow) Step-by-step instructions (Text highlighted in the operation list), banner at the top indicating operation completion. (Step completion notification), and task completion pop-up. Each type of guidance information Has an independent display enable bit , This indicates that this type of information is activated for display during the rendering of the current frame. This indicates masking. The platform will display all enable bits forming the enable vector. Define the mapping function from the mode to the enable vector. : ; When M=0 (training mode), That is, all All guidance information is activated; when M=1 (assessment mode), That is, only the task completion pop-up is retained. (Triggered when assessment time expires or when trainees submit their work voluntarily), all others are disabled.

[0045] In one embodiment, the operation method of the virtual simulation training platform can be specifically implemented as follows: the server obtains the dwell time corresponding to the current target operation step of the client; the server determines the display frequency of the displayed target guidance information based on the operation mode and dwell time of the virtual space; wherein, the current target operation step and the displayed target guidance information have a corresponding relationship.

[0046] In this embodiment of the invention, due to the strict sequential and standardized operation procedures of power generation operation and maintenance, a single course (such as waterless governor commissioning) covers ten operational steps, and designs the cross-operation of multiple devices and the collaborative use of various tools. To enable trainees to clearly grasp the operation progress and receive effective guidance in complex three-dimensional scenarios, the platform organizes the training tasks using a two-level hierarchical structure of major items and minor items.

[0047] The major items are upper-level task units formed by dividing the operation task into stages according to the operation process, and are used to provide a general description of the operation objectives to be completed in that stage. The minor items are lower-level task units further subdivided under the major items, corresponding to the smallest executable operation action in the operation task (such as performing a click, switch, adjustment, or parameter input on a specific device), and are the smallest operation unit for the system to perform operation judgment, trigger guidance prompts, and record assessment scores. Each target operation step corresponds to a minor item.

[0048] Furthermore, at the spatial guidance level of the 3D scene, visual cues need to be applied to the target object corresponding to the currently active item. If the target object is a virtual device in the scene, a red flashing highlight is overlaid on the device model. If the current step requires selecting a specific tool from the toolbox, a yellow flashing cue is overlaid on the toolbox icon. These two types of cues are rendered independently in space, without interfering with each other, allowing trainees to simultaneously perceive both the device target and the tool selection requirement.

[0049] To prevent static, fixed-intensity prompts from losing their guiding effect when users remain inactive for extended periods, the platform introduces an adaptive prompt intensity mechanism based on dwell time. By obtaining the dwell time corresponding to the client's current target operation step, the platform adaptively adjusts the display frequency of the displayed guidance information. In turn, the client adjusts the display frequency of the virtual device corresponding to the displayed target guidance information based on this frequency.

[0050] In one embodiment, the operation method of the virtual simulation training platform can be specifically implemented as follows: based on the guidance information corresponding to the current client's operation steps, obtain the initial flashing frequency; and adjust the display frequency of the displayed guidance information according to the initial flashing frequency, the upper limit flashing frequency, and the dwell time.

[0051] In this embodiment of the invention, when a student enters an active item, the system first reads the initial flashing frequency based on the guidance information corresponding to that item, which serves as the reference frequency for the guidance information when the student has just entered the item and no operational lag has yet occurred. The system pre-sets an upper limit flashing frequency as the highest allowed flashing frequency for this type of guidance information under any circumstances, and uses a time constant to control the rate at which the flashing frequency approaches the upper limit flashing frequency. Furthermore, the system records the student's dwell time on the current active item in real time to characterize their degree of lag on that item.

[0052] Furthermore, according to the exponentially increasing function Where t is the duration of stay, The initial flashing frequency, This is the upper limit of the flashing frequency. The flashing frequency that should be used for the guidance information at this moment is continuously calculated and the display frequency is adjusted in real time: when t=0, The prompt flashes lightly at a low frequency, giving trainees time to independently determine their operational goals. As the dwell time t accumulates, It increases monotonically in an exponential manner and gradually approaches [the desired value]. The visual intensity of the prompts gradually increases, providing more significant spatial guidance for trainees who have been unable to progress for a long time. Because... The entire process is continuous and bounded, with smooth changes in cue intensity throughout, preventing abrupt visual disturbances caused by jumps in values. Once the current sub-item is completed and the program moves to the next sub-item, the dwell timer t for the new sub-item is reset to zero, and the cue frequency returns to normal. And based on the guidance information corresponding to the new sub-item, it re-enters the above adaptive adjustment process.

[0053] Therefore, when trainees first enter a sub-item, the prompts are presented at a lower frequency to avoid interfering with their independent thinking and judgment, thus preserving ample space for independent exploration. When trainees experience operational delays, the flashing frequency increases smoothly in an exponential manner, and the intensity of the prompts gradually transitions from mild reminders to significant guidance, helping trainees quickly relocate the target equipment or tool.

[0054] Further, determine the operation type corresponding to the current target operation step; if the operation type is a binary operation, obtain the device status value after the target operation step is completed; if the device status value matches the preset status value, then determine that the target operation step has been completed; if the operation type is a parameter adjustment operation, obtain the parameter vector after the target operation step is completed; if the tolerance between each parameter in the parameter vector and its corresponding preset parameter is less than the preset tolerance, then determine that the target operation step has been completed.

[0055] Binary operations are operations where the result has only two discrete states and does not involve intermediate values. After the operation, the equipment can only be in one of the defined states, such as the closing and opening of a switch, or the opening and closing of a cabinet door. Parameter adjustment operations are operations where the result is one or a set of continuously adjustable values, and there is no unique precise value. After the operation, the equipment state is characterized by a set of parameters, such as adjusting valve opening, setting voltage or current, or setting instrument parameters.

[0056] Figure 3 This is a flowchart illustrating the virtual space rendering method provided by the present invention. Figure 3 As shown, the rendering methods for the rendered virtual space include: S210: The server obtains the processor's overall benchmark score, graphics processor's overall benchmark score, and currently available running memory for the virtual space to be rendered.

[0057] In this embodiment of the invention, existing mobile rendering adaptive solutions typically rely on server-side device database queries to determine the initial performance level and trigger rendering level switching based on the original frame rate threshold. However, the virtual simulation scenario for power generation operation and maintenance has two special characteristics: First, when trainees operate the equipment and trigger animations, there will be instantaneous frame rate drops unrelated to equipment performance, and directly judging based on the original frame rate will frequently lead to false triggering of switching; Second, trainees must move to the side of the equipment in the 3D scene to perform operations, and the perceptual matching degree between movement speed and frame rate directly affects the smoothness of operation, while existing solutions only adjust visual rendering parameters and do not process motion parameters.

[0058] Therefore, this invention renders the virtual space based on the processor's overall benchmark score, the graphics processor's overall benchmark score, and the currently available running memory.

[0059] The processor benchmark score is a score in points, obtained through the built-in lightweight CPU benchmark series test, representing the overall performance of the central processing unit. The graphics processor benchmark score is a score in points, obtained through the OpenGL ES rendering stress test, representing the overall performance of the graphics processor. The currently available RAM is the amount of RAM currently available for allocation on the device, read in real-time via the system ActivityManager interface, in GB.

[0060] Furthermore, since the three types of indicators have different dimensions, they need to be normalized separately. Taking the processor as an example, the calculation process for the normalized processor performance value is as follows: subtract the lower reference limit from the collected processor comprehensive score, and then divide by the reference interval span, which is the difference between the upper and lower reference limits, thus obtaining a linear ratio value. Take the minimum value between this ratio value and 1 to prevent exceeding the upper limit. Then take the larger value between 0 and the minimum value to prevent falling below the lower limit, thus truncating the result to the [0,1] interval to obtain the normalized processor performance value. Similarly, normalize the graphics processor comprehensive score and the currently available running memory to obtain the normalized graphics processor performance value and the normalized currently available running memory.

[0061] S220: The server calculates a weighted overall performance score based on the processor's overall benchmark score, the graphics processor's overall benchmark score, and the currently available running memory.

[0062] In this embodiment of the invention, a first weight corresponding to the normalized processor performance value, a second weight corresponding to the normalized graphics processor performance value, and a third weight corresponding to the normalized current available running memory are obtained. Then, a first product of the normalized processor performance value and its corresponding first weight, a second product of the normalized graphics processor performance value and its corresponding second weight, and a third product of the normalized current available running memory and its corresponding third weight are calculated. The sum of the first, second, and third products is then calculated as the weighted comprehensive performance score.

[0063] S230: The client adjusts the character movement speed and visual rotation speed in the virtual space to be rendered based on the weighted overall performance score.

[0064] In this embodiment of the invention, since students must first move to the target device in the virtual space before they can trigger subsequent operations, and the performance of the terminal hardware used by different students varies greatly, if a fixed character movement speed and view rotation speed are used, the visual update frequency of low-performance devices will lag significantly behind the actual movement response of the character due to the low rendering frame rate, which may even lead to dizziness, misaligned clicks and misoperations in severe cases.

[0065] This invention first obtains the baseline values ​​for character movement speed and visual rotation speed. For the character movement speed, it calculates the product between the proportional scaling factor corresponding to the character movement speed and the weighted comprehensive performance score, calculates the sum of the product and 1, and multiplies the sum by the baseline value of the character movement speed to obtain the adjusted character movement speed. Similarly, the adjusted visual rotation speed can be obtained.

[0066] S240: The client renders the virtual space to be rendered based on the adjusted character movement speed and the adjusted visual rotation speed to obtain the rendered virtual space.

[0067] In one embodiment, the operation method of the virtual simulation training platform can be specifically implemented as follows: the client obtains the instantaneous frame rate of the rendered virtual space; the client smooths the instantaneous frame rate to obtain a smoothed frame rate; the client adjusts the rendering quality level of the rendered virtual space based on the smoothed frame rate; wherein, the rendering quality level is constructed by a weighted comprehensive performance score.

[0068] In this embodiment of the invention, the hardware evaluation during the startup phase is only a one-time static grading, which cannot keep up with the dynamic fluctuations in scene rendering load due to the progress of operations during operation, thus causing the actual frame rate to deviate from the expected range. In particular, at the moment when the student clicks the device or triggers an operation animation, the rendering load increases sharply for a short period of time, which may cause a drop in frame rate for a single frame or several consecutive frames. This drop is caused by the operation event itself rather than by insufficient device performance. If the degradation judgment is directly triggered based on the original frame rate, frequent false switching will occur, which is highly correlated with the operation node and seriously interferes with the training process. Therefore, this invention continuously monitors the real-time frame rate during the platform's operation phase and uses an exponential moving average method to smooth this sequence, adjusting the rendering quality level of the rendered virtual space based on the smoothed frame rate.

[0069] Specifically, the smoothing coefficient is determined, the fourth product of the current instantaneous frame rate and the smoothing coefficient is calculated, then the difference between 1 and the smoothing coefficient is calculated, and the fifth product of the difference and the instantaneous frame rate of the previous frame is calculated. The sum of the fourth and fifth products is the smoothed frame rate.

[0070] The rendering quality level is determined by a weighted overall performance score. A first preset threshold and a second preset threshold are set, with the first preset threshold being greater than the second. When the weighted overall performance score is less than the first preset threshold, the rendering quality level is determined to be Level 1. When the weighted overall performance score is greater than or equal to the first preset threshold but less than the second preset threshold, the rendering quality level is determined to be Level 2. When the weighted overall performance score is greater than or equal to the second preset threshold, the rendering quality level is determined to be Level 3, where Level 1 is lower than Level 2, and Level 2 is lower than Level 3.

[0071] Furthermore, the first level represents low image quality, the second level represents medium image quality, and the third level represents high image quality. The first preset threshold can be 0.35, and the second preset threshold can be 0.65. The server sends the corresponding texture resolution, shadow precision, and particle count limit to the client based on the above rendering quality levels.

[0072] Furthermore, to avoid frequent oscillations and switching of rendering quality levels near frame rate thresholds, resulting in repeated jumps in image quality, the server incorporates a set of cooling timers as a time-locking mechanism for quality level adjustments. The initial value is set to zero, indicating that the system is initially in an adjustable state. Whenever the system performs a rendering quality level upgrade or downgrade operation based on the comparison between the smoothed frame rate and the uplink and downlink thresholds, the cooling timer is immediately reset to a preset cooling duration (e.g., five seconds). This timer decrements frame by frame according to the system clock. While the cooling timer is not yet zero, regardless of whether the smoothed frame rate reaches the uplink or downlink threshold again, the server blocks the quality level adjustment path, forcibly maintaining the current rendering quality level. The lock is only released when the cooling timer decrements to zero, allowing the next quality level adjustment decision. This preserves adaptive responsiveness while providing a stable observation window for rendering parameter switching, effectively suppressing frequent jumps in image quality caused by instantaneous frame rate fluctuations.

[0073] Specifically, if the smoothed frame rate is less than the downlink threshold and the current rendering quality level is greater than the first level, the current rendering quality level is reduced. If the smoothed frame rate is greater than the uplink threshold and the current rendering quality level is less than the third level, the current rendering quality level is increased. There is a hysteresis interval between the uplink and downlink thresholds, meaning that the uplink and downlink thresholds maintain a distance of no less than 20 frames per second, fundamentally avoiding frequent level switching caused by frame rate fluctuations near the critical point. After the quality level is adjusted, the rendering engine reloads the corresponding parameter configuration in the background without interrupting the current simulation operation. Users can also manually override the automatic configuration in the settings interface. Manual setting will pause the automatic adjustment mechanism until the user actively resumes or restarts the client.

[0074] Figure 4 This is a flowchart illustrating the operation method of a virtual simulation training platform provided in another embodiment of the present invention. Figure 4 As shown, the operational methods of the virtual simulation training platform may also include: S410: The server obtains the number of device interaction events corresponding to the execution of the target operation step.

[0075] In this embodiment of the invention, the platform categorizes operational events into two types: device interaction events and text input events. Device interaction events are triggered by user actions such as clicking, switching, and adjusting virtual device models. The system captures an event triple when the event occurs. The event triple includes the index of the currently active target operation step, the number of the operated device, and a global logical timestamp assigned by the server. Text input events are triggered by users submitting text content through an input box. The system captures an event triple, which includes the text string submitted by the user, the index of the currently active target operation step, and a global logical timestamp assigned by the server. Both types of events are uniformly arranged in ascending order of the global logical timestamp on the server, forming a time-series operation log. The log is presented on the client as a list of operation records, with the two types of records distinguished by different colors: device interaction records are rendered in white text and are not editable to ensure the originality of the operation records. Text input records are rendered in green text and allow users to correct them afterward in training mode (green records can also be corrected before the end of the assessment in assessment mode to allow trainees to supplement and improve the scheduling report content). The correction operation generates a new text input event, which is appended to the end of the log. The original record is retained, and the correction version is based on the latest timestamp event, ensuring that the correction action itself can also be traced.

[0076] For scoring assistance, the server groups and aggregates logs according to the target operation step index, generating a step-level operation summary. The operation summary of the target operation step is defined as a triple, which consists of the number of device interaction events, the number of text input events, and the time taken for the target operation step to be executed.

[0077] Furthermore, for each target operation step in the standard operating procedure, it is expected that only one device interaction is required. The deviation of the number of device interaction events from 1 directly reflects the degree of hesitation or accidental triggering by the trainee.

[0078] S420: The server obtains the squared deviation of the number of device interactions corresponding to the target operation step from the standard number, based on the number of device interaction events.

[0079] In this embodiment of the invention, the difference between the number of device interaction events and the standard number (which can be 1) is calculated, and the difference is squared to obtain the squared deviation of the number of device interactions corresponding to the target operation step from the standard number.

[0080] S430: The server obtains the squared value of the relative time deviation when the target operation step is executed.

[0081] In this embodiment of the invention, the ratio between the time taken for the target operation step and the average time taken for all clients in the same virtual space to complete the target operation step is calculated. Based on the ratio and the standard number of times, the standard deviation is obtained. The standard deviation is squared to obtain the squared value of the relative time deviation when the target operation step is executed.

[0082] S440: The server determines the client's weighted operation deviation score based on the squared deviation of the number of device interactions corresponding to the target operation step from the standard number of interactions and the squared deviation of the relative time consumption.

[0083] In this embodiment of the invention, a tradeoff coefficient and an importance weight are obtained for the time consumption deviation relative to the operation number deviation. The importance weight represents the degree of importance to the target operation step. A sixth product is calculated between the tradeoff coefficient and the squared value of the relative time consumption deviation. The sum of the sixth product and the squared value of the deviation is then calculated. Finally, a seventh product is calculated between this sum and the importance weight to obtain a weighted operation deviation score. When there are multiple target operation steps, the seventh product is accumulated to obtain the final accumulated weighted operation deviation score.

[0084] In one embodiment, after S160, the operation method of the virtual simulation training platform can be specifically implemented as follows: each target client performs the operation on its locked target operation steps; the server obtains the assessment start time and the completion time of the final target operation step; the server obtains the cumulative time for completing multiple target operation steps based on the assessment start time and the completion time of the final target operation step.

[0085] In this embodiment of the invention, the platform introduces a completion time recording mechanism for each target operation step in the assessment mode. It obtains the start time of the assessment mode and the completion time of the final target operation step, thereby obtaining the cumulative time taken to complete multiple target operation steps.

[0086] Figure 5 This is a schematic diagram of the operating system of the virtual simulation training platform provided by the present invention. Figure 5 As shown, the operating system of the virtual simulation training platform includes: The server-side is used to obtain operation tasks in the rendered virtual space; each operation task includes multiple operation steps; it receives interaction requests and determines whether the client is sending an interaction request for the first time; if it is determined that the client is sending an interaction request for the first time, it adds a row-level exclusive lock to the ownership record of the target operation step; the row-level exclusive lock indicates that no read or modification operations can be performed on the ownership record; it determines the ownership status of the target operation step; the ownership status indicates whether the target operation step has been claimed by other clients; if it is determined that the ownership status of the target operation step indicates that the target operation step has not been claimed by other clients, it releases the row-level exclusive lock of the target operation step and locks the target client and the target operation step; if it is determined that the ownership status of the target operation step indicates that the target operation step has been claimed by other clients, it rejects the interaction request; multiple clients are used to send interaction requests; each interaction request indicates a request to interact with the virtual device corresponding to the target operation step among multiple operation steps.

[0087] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for operating a virtual simulation training platform, characterized in that, include: The server obtains the operation tasks in the rendered virtual space; wherein, the operation tasks include multiple operation steps; Multiple clients send interaction requests; wherein, the interaction request indicates a request to interact with the virtual device corresponding to the target operation step among the multiple operation steps; The server receives the interaction request and determines whether the client is sending the interaction request for the first time. If the server determines that the client is sending the interaction request for the first time, it adds a row-level exclusive lock to the record to which the target operation step belongs; wherein, the row-level exclusive lock means that no read or modification operation can be performed on the record to which the target operation step belongs; The server determines the ownership status of the target operation step; wherein, the ownership status indicates whether the target operation step has been claimed by other clients; If the server determines that the ownership status of the target operation step indicates that the target operation step has not been claimed by other clients, it releases the row-level exclusive lock of the target operation step and locks the target client with the target operation step. If the server determines that the ownership status of the target operation step indicates that the target operation step has been claimed by another client, then it will reject the interaction request.

2. The operating method of the virtual simulation training platform according to claim 1, characterized in that, Also includes: The server obtains the mapping function corresponding to the virtual space; wherein, the mapping function has an assessment enable vector, and each display enable bit in the assessment enable vector indicates whether the guidance information is activated and displayed when the current frame is rendered in the virtual space, and the guidance information is visual prompt information that assists the trainee in locating the operation target in the virtual space; The client sequentially activates the guidance information corresponding to the assessment enable vector during the rendering of the current frame, so as to display it in the virtual space.

3. The operating method of the virtual simulation training platform according to claim 2, characterized in that, The server-side acquisition of operation tasks in the virtual space includes: Obtain global mode state variables; wherein, the global mode state variables represent the current operation mode; Based on the global mode state variables, the operation mode of the virtual space is determined; wherein, the operation mode corresponds to the type of the virtual space. Based on the operation mode of the virtual space, the operation tasks in the virtual space are obtained.

4. The operating method of the virtual simulation training platform according to claim 3, characterized in that, Also includes: The server obtains the dwell time corresponding to the current target operation step of the client; The server determines the display frequency of the target guidance information based on the operation mode of the virtual space and the dwell time; wherein, the current target operation step and the displayed target guidance information have a corresponding relationship.

5. The operating method of the virtual simulation training platform according to claim 4, characterized in that, The server-side adaptive adjustment of the display frequency of the displayed guidance information includes: Based on the guidance information corresponding to the current client's operation steps, obtain the initial flashing frequency; The display frequency of the displayed guidance information is adjusted based on the initial flashing frequency, the upper limit flashing frequency, and the dwell time.

6. The operating method of the virtual simulation training platform according to claim 1, characterized in that, The rendering method for the virtual space includes: The server obtains the overall processor score, graphics processor score, and currently available memory for the virtual space to be rendered. The server calculates a weighted overall performance score based on the processor's overall benchmark score, the graphics processor's overall benchmark score, and the currently available running memory; The client adjusts the character movement speed and visual rotation speed within the virtual space to be rendered based on the weighted overall performance score. The client renders the virtual space to be rendered based on the adjusted character movement speed and the adjusted visual rotation speed to obtain the rendered virtual space.

7. The operating method of the virtual simulation training platform according to claim 6, characterized in that, Also includes; The client obtains the instantaneous frame rate of the rendered virtual space; The client performs smoothing processing on the instantaneous frame rate to obtain a smoothed frame rate; The client adjusts the rendering quality level of the rendered virtual space based on the smoothed frame rate; wherein the rendering quality level is constructed from the weighted comprehensive performance score.

8. The operating method of the virtual simulation training platform according to claim 1, characterized in that, Also includes: The server obtains the number of device interaction events corresponding to the execution of the target operation step; The server obtains the squared deviation of the number of device interactions corresponding to the target operation step from the standard number based on the number of device interaction events. The server obtains the squared value of the relative time consumption deviation when the target operation step is executed; The server determines the client's weighted operation deviation score based on the squared deviation of the number of device interactions corresponding to the target operation step from the standard number of interactions and the squared relative time consumption deviation.

9. The operating method of the virtual simulation training platform according to claim 1, characterized in that, After locking the target client with the target operation step, the method further includes: Each target client performs the operation steps for its locked target; The server obtains the start time of the assessment and the completion time of the final target operation steps; The server obtains the cumulative time taken to complete multiple target operation steps based on the start time of the assessment and the completion time of the final target operation step.

10. A working system for a virtual simulation training platform, characterized in that, include: The server-side is used to obtain operation tasks in the rendered virtual space; wherein the operation task includes multiple operation steps; receive an interaction request and determine whether the client is sending the interaction request for the first time; if it is determined that the client is sending the interaction request for the first time, then add a row-level exclusive lock to the ownership record of the target operation step; wherein the row-level exclusive lock indicates that no read or modification operation can be performed on the ownership record; determine the ownership status of the target operation step; wherein the ownership status indicates whether the target operation step has been claimed by other clients; if it is determined that the ownership status of the target operation step indicates that the target operation step has not been claimed by other clients, then release the row-level exclusive lock of the target operation step and lock the target client and the target operation step; if it is determined that the ownership status of the target operation step indicates that the target operation step has been claimed by other clients, then reject the interaction request; Multiple clients are used to send interaction requests; wherein the interaction request indicates a request to interact with the virtual device corresponding to the target operation step among the multiple operation steps.

Images