Large-screen visual dynamic effect splitting solution and system
The number of motion effect executions is predicted through the ARIMA algorithm and combined with the two-stage submission method and the time server unified clock, the problem of large-screen dynamic effect splitting is solved, the integrity and stability of the motion effect on the large-screen, and the audience experience is improved.
Patent Information
- Application Number
- CN202411922637.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-12-20
- Filing Date
- 2024-12-25
- Publication Date
- 2025-04-25
AI Technical Summary
In high-resolution large-area large-screen visualization applications, when the dynamic effect crosses the domain, when each logic screen corresponds to the output of a graphic workstation, the dynamic effect content time is out of synchronization, resulting in the display content time and accumulation error on the large screen, which in turn causes dynamic effect splitting, affecting the visual perception.
The ARIMA algorithm is used to predict the number of dynamic efficiency executions, and the unified clock of the time server is achieved through the two-stage submission method and the time server, and the concurrency of transactions are achieved, transactions are eliminated and accumulated time errors are accumulated, and the integrity of dynamic efficiency on the large screen is ensured.
It effectively solves the phenomenon of splitting and misalignment of Web animation on the large screen, improves the viewing experience and application value of large screens, and ensures the integrity and stability of large screen display.
Smart Images
Figure CN119883163A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a large-screen visualization display method, and in particular to a large-screen visualization dynamic effect splitting solution and system. Background Art
[0002] In high-resolution, large-area, large-screen visualization applications, a single graphics workstation is not capable of performing complete display and output tasks. Multiple graphics workstations are required to form a cluster. Each graphics workstation corresponds to a large-screen logical split-screen area, outputs the corresponding image, and splices it into a complete image on the large screen for display. When the motion effect spans two or more large-screen logical screens, since each logical screen corresponds to the output of a graphics workstation, when the output content of multiple graphics workstations corresponding to the motion effect is not synchronized in time, the corresponding display content on the large screen will be out of sync in time. When the time difference accumulates to a certain extent, the large-screen spliced image will have visible horizontal or vertical misalignment, resulting in the large-screen motion effect fragmentation phenomenon, which seriously affects the viewing experience of the large screen. Summary of the invention
[0003] Purpose of the invention: In view of the above problems, the present invention proposes a large-screen visual animation fragmentation solution and system, which can eliminate the transaction asynchrony and accumulated time error phenomena in the animation execution process, effectively solve the fragmentation and dislocation phenomenon of Web animation on the large screen, and enhance the large-screen viewing experience and application value.
[0004] Technical solution: The technical solution adopted by the present invention is a large-screen visual dynamic effect splitting solution, including:
[0005] Taking the execution time and execution count of each animation as input, the ARIMA algorithm is used to predict the execution count of the animation from the start of the animation to the point when the transaction generates animation split, which is recorded as the zero prediction value; the zero prediction value and page refresh request are sent to each graphics workstation;
[0006] Calculate the concurrent time of the current transaction of the graphics workstation according to the reception time of the page refresh request of the graphics workstation;
[0007] The graphics workstation starts the current transaction according to the agreed concurrent time, realizing transaction concurrency on each graphics workstation;
[0008] When all requests from the graphics workstations are not received within the specified time, a rollback operation is initiated;
[0009] When the animation starts to execute, the number of executions of the animation is calculated. When the number of executions of the animation is close to the predicted value of zeroing, a zeroing request is sent;
[0010] Based on the reset request, agree on the next concurrent time for the animation.
[0011] Furthermore, the ARIMA algorithm uses a model based on first-order differences, and the calculation formula is:
[0012] y(t)=0.003+0.939*y(t-1)-0.869*ε(t-1)
[0013] Where y(t) is the cumulative error value at the current time point, y(t-1) is the cumulative error value at the previous time point, and ε(t-1) is the interference noise value.
[0014] According to the reception time of the page refresh request of the graphics workstation, the concurrent time of the current transaction of the graphics workstation is calculated as Max|ts|+Δt, where ts is the reception time of the page refresh request of the graphics workstation and Δt is the setup time.
[0015] According to the reset request, the next concurrent time of the animation is agreed upon, which also includes: modifying the animation status mark to mark the status of the animation as no longer executed; after the animation reads the "no longer executed" status mark, it stops executing and waits for the next animation execution notification.
[0016] This method uses a time server to unify the clock.
[0017] The present invention also proposes a large-screen visualization dynamic effect splitting solution system, comprising a background server, a transaction manager and a transaction coordinator, wherein the transaction manager is arranged in a graphics workstation;
[0018] The backend server includes a prediction module, a page request module and a rollback operation module. The prediction module is used to: take the execution time and execution times of each animation as input, adopt the ARIMA algorithm, and predict the execution times of the animation from the beginning of the animation to the point when the transaction generates the animation split, which is recorded as the zero prediction value; the page request module sends the page refresh request and the zero prediction value to the transaction manager of each graphics workstation; the rollback operation module is used to execute the rollback task;
[0019] The transaction manager includes a negotiation module and a submission module; the transaction manager receives a page refresh request and a zero prediction value from a background server, and the negotiation module is used to forward the zero prediction value to the submission module, and forward the page refresh request to the transaction coordinator; the negotiation module is also used to send a zero message to the transaction coordinator; after receiving the agreed concurrent time, the submission module starts the timer, and when the timer time reaches the agreed concurrent time, the current transaction is started to realize the transaction concurrency on each graphics workstation; the submission module is also used to calculate the number of executions of the animation when the animation starts to execute, and when the number of executions of the animation is close to the zero prediction value, send a zero request to the negotiation module of the transaction manager;
[0020] The transaction coordinator includes a negotiation module, an agreement module and a rollback module; the transaction coordinator receives page refresh requests forwarded by various graphics workstations, and the negotiation module of the transaction coordinator forwards the reception time of the request received by the transaction manager to the agreement module; the agreement module calculates the concurrent time of the current transaction of the graphics workstation according to the reception time, and returns the agreed concurrent time to the submission module of the transaction manager; the agreement module executes the agreement according to the zeroing message sent by the negotiation module, and agrees on the next round of concurrent time of the animation; the rollback module is used to determine whether the page refresh requests of all graphics workstations are received within the specified time, and if not, the rollback request is sent to the background service to execute the rollback.
[0021] The transaction manager also includes a zeroing module, which is used to: receive a zeroing message sent by the negotiation module of the transaction manager, modify the animation status mark, and mark the status of the animation as no longer executed.
[0022] The system uses NTP time server to unify the system time of transaction manager, transaction coordinator and background server.
[0023] The rollback task is to notify the backend server that the current transaction page request has failed and needs to be re-initiated.
[0024] Beneficial effects: Compared with the prior art, the present invention has the following advantages: Aiming at the fragmentation of large-screen animation, the present invention takes the core transaction of large-screen animation execution as the key point to solve the problem, proposes a method based on the combination of two-stage submission and ARIMA prediction algorithm, eliminates the transaction asynchrony and cumulative time error in the animation execution process, effectively solves the fragmentation and dislocation of Web animation on the large screen, improves the large-screen viewing experience and application value, and ensures the integrity and stability of the large-screen display. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1 It is a schematic diagram of the large-screen visual dynamic effect splitting solution system module and flow chart of the present invention. DETAILED DESCRIPTION
[0026] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.
[0027] Example 1
[0028] The large-screen visualization dynamic effect splitting solution described in the present invention takes the core transaction of large-screen dynamic effect execution as the key point to solve the problem, and proposes a method based on the combination of two-stage submission and ARIMA prediction algorithm.
[0029] Protocol clock synchronization method: By using the time server, the NTP time server in the local area network is used to unify the system time of graphics workstations, application servers and other devices in the large-screen visualization system, ensuring that all device times are from the same source and synchronized, thereby reducing system time errors.
[0030] Two-phase commit method: When the page refresh request command is sent, since the console needs to send the page refresh request to multiple graphics workstations, there is an asynchronous delay in sending the request command, the network transmission time is different, and the time points when each graphics workstation receives the request command are inconsistent. The two-phase commit method is used to eliminate the initial time difference. The background service initiates a transaction request to the graphics workstation cluster. After receiving the request, the graphics workstation notifies the transaction coordinator of the receipt of the transaction request and requires the transaction coordinator to make a concurrent agreement. When the transaction coordinator receives the submissions from all graphics workstations, it calculates the agreed concurrent time and returns it to each graphics workstation. After the graphics workstation receives the returned agreed concurrent time, it starts the timer and submits the execution transaction at the same time at the agreed concurrent time.
[0031] ARIMA prediction algorithm model: The cumulative time difference when the animation is executed belongs to the time series. Since the observed values of the time series have certain dependencies or correlations, they are mainly manifested as certain periodicity, trend and stability. Therefore, the model can be described by a certain quantity, so that future values can be predicted based on historical values. In the present invention, a prediction model is designed with the ARIMA algorithm as the core. The model uses the time series of a single animation execution as input to obtain a model formula for calculating the animation split. The model formula is introduced into the system to accurately predict the time difference when the split occurs. The following is the model formula based on the first-order difference:
[0032] y(t)=0.003+0.939*y(t-1)-0.869*ε(t-1)
[0033] like Figure 1 As shown, the large-screen visualization dynamic effect splitting solution of the present invention includes:
[0034] Taking the execution time and execution count of each animation as input, the ARIMA algorithm is used to predict the execution count of the animation from the start of the animation to the point when the transaction generates animation split, which is recorded as the zero prediction value; the zero prediction value and page refresh request are sent to each graphics workstation;
[0035] The page refresh request is forwarded, and the concurrent time of the current transaction of the graphics workstation is calculated according to the receiving time of the page refresh request; for example, the concurrent time is set to Max|receiving time|+10ms.
[0036] The graphics workstation starts the current transaction according to the agreed concurrent time, realizing transaction concurrency on each graphics workstation;
[0037] When the requests of all the graphic workstations are not received within the specified time, it is determined that at least one of the graphic workstations is abnormal, and a rollback operation is started to execute the rollback task;
[0038] When the animation starts to execute, the number of executions of the animation is calculated. When the number of executions of the animation is close to the predicted value of zeroing, a zeroing request is sent;
[0039] According to the zeroing request, the next round of concurrent time of the animation is agreed upon, and the animation status mark is modified at the same time, and the status of the animation is marked as no longer executed. After the animation reads the status mark, it stops executing and waits for the next animation execution notification from the submission module.
[0040] Figure 1 In the example, transactions include: transaction one (T1) is the page refresh request command sending, transaction two (T2) is the browser request and rendering stage, and transaction three (T3) is the animation execution stage.
[0041] Example 2
[0042] The large-screen visualization dynamic effect splitting solution system described in the present invention includes a background server, a transaction manager and a transaction coordinator, and the transaction manager is arranged in a graphics workstation. The transaction coordinator is a component that calculates and coordinates the concurrent time of the current transaction of the graphics workstation and determines the execution of the rollback task. After the transaction coordinator receives the request forwarded by each graphics workstation, the agreement module calculates the concurrent time of the current transaction of the graphics workstation according to the receiving time of each request, and forwards the agreed concurrent time to the submission module of the transaction manager. If the request of all graphics workstations is not received, the rollback operation is started. When the dynamic effect starts to execute, the zeroing message is received and the next round of concurrent time of the agreed dynamic effect is executed. The transaction manager is a component that stores the zeroing prediction value and manages the status mark, and executes and starts the concurrent transaction on the graphics workstation. After the transaction manager receives the background request, the negotiation module forwards the request to the submission module and the transaction coordinator, receives the concurrent time of the current transaction calculated by the transaction coordinator agreement module, starts the timer, and when the timer time reaches the agreed concurrent time, starts the current transaction to realize the transaction concurrency on each graphics workstation. When the animation starts to execute, the negotiation module receives the reset request, sends the reset request to the reset module, sends the reset message to the transaction coordinator and modifies the status mark. Figure 1 As shown:
[0043] The backend server includes a prediction module, a page request module, and a rollback operation module. The prediction module uses the ARIMA algorithm to predict the number of times a transaction will generate an animation split after the animation starts to execute, which is recorded as the zero prediction value. The page request module sends the page refresh request and the zero prediction value to the transaction manager of each graphics workstation. The rollback operation module is used to execute the rollback task. The rollback task is to notify the backend server that the current transaction page request has failed and the page request needs to be re-initiated.
[0044] The transaction manager includes a negotiation module, a submission module and a zeroing module; the transaction manager receives a page refresh request and a zeroing prediction value from a background server, the negotiation module is used to forward the zeroing prediction value to the submission module, and the page refresh request is forwarded to the transaction coordinator; after receiving the agreed concurrent time, the submission module starts the timer, and when the timer time reaches the agreed concurrent time, the current transaction is started to realize the transaction concurrency on each graphic workstation; when the animation starts to execute, the number of executions of the animation is calculated, and when the number of executions of the animation is close to the zeroing prediction value, a zeroing request is sent to the negotiation module of the transaction manager; the negotiation module sends the zeroing request to the zeroing module, and at the same time sends a zeroing message to the transaction coordinator; after receiving the zeroing message, the zeroing module modifies the status mark and marks the status of the animation as no longer executed. After the animation reads the status mark, it stops executing and waits for the next execution notification.
[0045] The transaction coordinator includes a negotiation module, an agreement module and a rollback module; the transaction coordinator receives page refresh requests forwarded by each graphic workstation, and the negotiation module forwards the reception time of the request received by the transaction manager to the agreement module; the agreement module calculates the concurrent time of the current transaction of the graphic workstation according to the reception time, and returns the agreed concurrent time to the submission module of the transaction manager; the agreement module executes the agreement according to the zeroing message, and agrees on the next round of concurrent time of the animation; the rollback module is used to determine whether the page refresh requests of all graphic workstations are received within the specified time. If not, it is determined that at least one graphic workstation is abnormal, and the rollback operation is started, and the rollback request is sent to the background service.
[0046] The NTP time server is used to unify the system time of devices such as graphic workstations, transaction coordinators, and background servers in the large-screen visualization system.
[0047] The present invention improves the currently used visualization display platform, adds transaction coordination, transaction management and service, and rollback operation functional units, and effectively solves the problem of fragmentation and dislocation of Web page dynamic effects on the large screen.
Claims
1. A solution to the splitting of large-screen visual dynamic effects, characterized in that: include: Taking the execution time and execution count of each animation as input, the ARIMA algorithm is used to predict the number of animation executions from the start of the animation to the point when the transaction is split, which is recorded as the zero prediction value. Send zero prediction values and page refresh requests to each graphics workstation; Calculate the concurrent time of the current transaction of the graphics workstation according to the reception time of the page refresh request of the graphics workstation; The graphics workstation starts the current transaction according to the agreed concurrent time, realizing transaction concurrency on each graphics workstation; When all requests from the graphics workstations are not received within the specified time, a rollback operation is initiated; When the animation starts to execute, the number of executions of the animation is calculated. When the number of executions of the animation is close to the predicted value of zeroing, a zeroing request is sent; Based on the reset request, agree on the next concurrent time for the animation.
2. The large-screen visualization dynamic effect splitting solution according to claim 1 is characterized by: The ARIMA algorithm uses a model based on first-order differences, and the calculation formula is: y(t)=0.003+0.939*y(t-1)-0.869*ε(t-1) Where y(t) is the cumulative error value at the current time point, y(t-1) is the cumulative error value at the previous time point, and ε(t-1) is the interference noise.
3. The large-screen visualization dynamic effect splitting solution according to claim 1 is characterized by: According to the reception time of the page refresh request of the graphics workstation, the concurrent time of the current transaction of the graphics workstation is calculated as Max|ts|+Δt, where ts is the reception time of the page refresh request of the graphics workstation, and Δt is the setup time.
4. The large-screen visualization dynamic effect splitting solution according to claim 1 is characterized by: According to the reset request, the next concurrent time of the animation is agreed upon, which also includes: modifying the animation status mark to mark the status of the animation as no longer executed; after the animation reads the "no longer executed" status mark, it stops executing and waits for the next animation execution notification.
5. The large-screen visualization dynamic effect splitting solution according to claim 1 is characterized by: This method uses a time server to unify the clock.
6. A large-screen visual dynamic effect fragmentation solution system, characterized in that: It includes a background server, a transaction manager and a transaction coordinator, wherein the transaction manager is located in a graphic workstation; The backend server includes a prediction module, a page request module and a rollback operation module. The prediction module is used to: take the execution time and execution times of each animation as input, use the ARIMA algorithm to predict the execution times of the animation from the start of the animation to the point when the transaction generates the animation split, and record it as the zero prediction value; The page request module sends the page refresh request and the zero prediction value to the transaction manager of each graphics workstation; the rollback operation module is used to execute the rollback task; Transaction manager, including negotiation module and submission module; The transaction manager receives the page refresh request and the zero prediction value from the background server, and the negotiation module is used to forward the zero prediction value to the submission module and forward the page refresh request to the transaction coordinator; The negotiation module is further used to: send a zeroing message to the transaction coordinator; After receiving the agreed concurrent time, the submission module starts the timer. When the timer reaches the agreed concurrent time, the current transaction is started to realize the transaction concurrency on each graphics workstation. The submission module is also used to: when the animation starts to execute, calculate the number of executions of the animation, and when the number of executions of the animation is close to the zero prediction value, send a zero request to the negotiation module of the transaction manager; Transaction coordinator, including negotiation module, agreement module and rollback module; The transaction coordinator receives the page refresh request forwarded by each graphic workstation, and the negotiation module of the transaction coordinator forwards the receiving time of the request received by the transaction manager to the agreement module; The agreement module calculates the concurrent time of the current transaction of the graphics workstation according to the receiving time of the request received by the transaction manager, and returns the agreed concurrent time to the commit module of the transaction manager; The agreement module executes the agreement according to the zeroing message sent by the negotiation module and agrees on the next round of concurrent time of the animation effect; the rollback module is used to determine whether the page refresh request of all graphics workstations is received within the specified time. If not, the rollback request is sent to the background service to execute the rollback.
7. The large-screen visualization dynamic effect fragmentation solution system according to claim 6 is characterized by: The concurrent time of the current transaction of the graphics workstation is calculated according to the receiving time of the request received by the transaction manager. The calculation formula is: Max|ts|+Δt, where ts is the receiving time of the page refresh request of the graphics workstation, and Δt is the setting time.
8. The large-screen visualization dynamic effect fragmentation solution system according to claim 6 is characterized by: The transaction manager also includes a zeroing module, which is used to: receive a zeroing message sent by the negotiation module of the transaction manager, modify the animation status mark, and mark the status of the animation as no longer executed.
9. The large-screen visualization dynamic effect fragmentation solution system according to claim 6 is characterized by: The system uses NTP time server to unify the system time of transaction manager, transaction coordinator and background server.
10. The large-screen visual dynamic effect fragmentation solution system according to claim 6 is characterized by: The rollback task is to notify the backend server that the current transaction page request has failed and that the page request needs to be re-initiated.