A multi-resource collaborative exhibition hall navigation control system based on timeline arrangement

Abstract

This invention relates to the field of control system technology, specifically disclosing a multi-resource collaborative exhibition hall guide control system based on timeline orchestration. The system includes: a timeline synchronization engine, a hierarchical resource track manager, a user intent perception module, dynamic triggers, a status monitoring and feedback module, a precise jump controller, and a closed-loop collaborative optimization module. Through the timeline synchronization engine and hierarchical resource management, precise coordination and intelligent conflict resolution of multiple resources such as guide audio, media videos, and lighting are achieved. Combined with precise jump control and rapid scene switching functions, the convenience and flow of guide operations are improved. The user intent perception module adapts to personalized needs, and the status monitoring and feedback mechanism ensures stable system operation. Furthermore, the closed-loop collaborative optimization module enables dynamic parameter updates and continuous performance improvement, comprehensively meeting the core requirements of exhibition hall guides for accuracy, smoothness, personalization, and stability.

Description

Technical Field

[0001] This invention relates to the field of control system technology, specifically to a multi-resource collaborative exhibition hall navigation control system based on time axis arrangement. Background Technology

[0002] In exhibition hall guided tour scenarios, the presentation of the guided tour relies on the coordinated use of various resources such as guided audio, media videos, and lighting control. The core requirement is to achieve precise, chronological linkage of multiple resources while adapting to the browsing needs and scene switching requirements of different users. However, existing exhibition hall guided tour control systems have many shortcomings: First, the system suffers from poor multi-resource coordination, with audio, video, and lighting resources frequently experiencing time synchronization issues, leading to a fragmented guided tour experience. Second, resource scheduling lacks flexibility, with different types and levels of resources prone to time conflicts, and the system cannot dynamically adjust priorities based on user behavior. Third, the navigation function lacks accuracy, and the recovery of resource states after navigation is not smooth, affecting the continuity of the guided tour. Fourth, it lacks the ability to perceive and adapt to user intent, making it difficult to provide personalized guided tour services based on user interests. Fifth, status monitoring is incomplete, with weak capabilities in detecting and repairing resource execution anomalies and system connection problems, and a lack of continuous optimization mechanisms, leading to a decline in system performance after long-term use. These problems make it difficult for the existing system to meet the high requirements of exhibition hall guided tours for accuracy, smoothness, personalization, and stability. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a multi-resource collaborative exhibition hall navigation control system based on timeline arrangement, which solves the problems mentioned in the background technology.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a multi-resource collaborative exhibition hall navigation control system based on timeline arrangement, comprising: The timeline synchronization engine is used to generate a unified timeline signal based on a global timestamp and drive multiple resource tracks to execute in chronological order. The hierarchical resource track manager is used to manage at least three types of hierarchical resource tracks, including a guide audio track, a media video track, a lighting control track, and an optional interactive device track. Each track arranges resource trigger instructions according to time intervals and supports dynamic adjustment of inter-layer dependencies and priorities. The user intent perception module collects user behavior data through sensors to generate user intent vectors, providing dynamic input for resource scheduling. Dynamic triggers are used to automatically trigger corresponding control commands when the timeline reaches the start time point of each resource track, based on user intent and resource status. The status monitoring and feedback module is used to monitor the execution status of each resource, the system connection status, and the user interaction status in real time, and provides visual status feedback. The precise jump controller supports second-level jumps on the timeline based on the input target time point, synchronously updates the status of all track resources, and compensates for execution delays. The closed-loop collaborative optimization module dynamically adjusts the timeline orchestration scheme and resource scheduling strategy based on resource execution data, user feedback data, and synchronization consistency indicators.

[0005] Preferably, the timeline synchronization engine calculates the global time synchronization deviation and performs real-time calibration using the following formula: ; in, For time synchronization deviation, This represents the total number of resource orbitals in the system. For the first Local timestamps for each track For global master timestamp; when hour, The preset threshold is set between 10ms and 100ms. The system automatically triggers the time axis calibration mechanism, and the calibration parameters are updated in real time through the closed-loop collaborative optimization module. The calibration process includes the following steps: S11, All in the periodic acquisition system Local timestamp of each resource track and global master timestamp The acquisition period is set to 50ms to 200ms; S12, Collect the data and Substituting into the above formula for calculating time synchronization deviation, we obtain the current... ; S13, Judgment With preset threshold Size relationship; S14, if Record the current synchronization status as normal, and return to S11 to continue periodic data collection; S15, if This triggers the calibration process: S151. Obtain the latest calibration parameters from the closed-loop collaborative optimization module; S152. Adjust the local timestamps of each resource track according to the calibration parameters, with the adjustment amount being... , The calibration coefficient ranges from 0.8 to 1.2. S153. After adjustment, re-collect the local timestamps and global master timestamps of each resource track, and calculate the adjusted values. S54, Judgment Is it ≤ Otherwise, repeat S152 to S153 until the condition is met; S16. Feed back the calibration results to the status monitoring and feedback module and record the calibration log.

[0006] Preferably, the hierarchical resource track manager allocates independent time slice intervals for each resource type and level, and determines the execution priority of each track during conflicting time periods through a resource scheduling weight function that incorporates user intent: ; in, For the first The scheduling weight of each resource track. The duration of this resource orbit. Total guided tour duration, This is the priority coefficient for this resource type. The highest priority coefficient, This represents the historical number of times this resource has been triggered. This represents the average number of system triggers. This represents the user's intent weight for the resource, with a value ranging from 0 to 1, and is output by the user intent perception module. Let be the weighting adjustment coefficient, and satisfy . ; Time slice allocation and priority conflict resolution include the following steps: S21. Get the total guided tour duration and the duration of each resource orbit The time slice intervals of each track are initially divided according to the hierarchical order of the base layer, enhancement layer, and interaction layer to ensure that the time slices of the base layer resources do not overlap. S22. Collect the output of the user intent perception module. Combined with historical trigger data and resource type priority coefficient Substitute into the weighting function to calculate each orbital ; S23. Detect whether there are overlapping conflicts in the time slices of each track. If there are no conflicts, determine the final time slice allocation scheme. S24. If a conflict exists, extract all orbitals within the conflict time period. ; S25, according to Sort from largest to smallest Larger tracks retain their original time slots, while lower-priority tracks have their time slots adjusted to outside of conflicting time periods, with the adjustment not exceeding their own time slot. 20%; S26. Verify whether there are still conflicts after the adjustment. If there are, repeat S24 to S25 until there are no conflicts, and output the final time slice allocation scheme.

[0007] Preferably, the dynamic trigger detects that the time axis has reached the resource trigger point. At that time, a trigger determination function that incorporates user intent is used to determine whether to execute the resource instruction: ; in, This is the current timeline position. To trigger the tolerance time window, the value ranges from 50ms to 500ms. This is the current state of the resource. The threshold for triggering user intent is set, with a value ranging from 0.3 to 0.7; if If so, the corresponding resource control command will be executed immediately; if it is an enhancement layer or interaction layer resource, and If so, the triggering of the resource will be delayed or skipped; The execution process includes the following steps: S31. Real-time monitoring of the current position of the timeline. Compare the trigger points of each resource track ; S32, when When this occurs, a pre-check process is triggered to obtain the current status of the corresponding resource. and the user's intent weight for the resource ; S33, will and Substitute into the trigger determination function for calculation ; S34, if Generate resource control instructions and send them to the corresponding execution unit, while recording the trigger time and resource identifier; S5, if Determine the resource type: S351. If it is a basic layer resource, output a trigger exception prompt to the status monitoring and feedback module, and still execute the basic control commands; S352. If it is an enhancement layer resource, delay the trigger point to , The delay duration is 1 to 3 seconds. After the delay, S31 to S33 are re-executed. S353: If it is an interactive layer resource, skip the trigger point directly and record the reason for skipping and the user intent data. S36. Receive instruction execution feedback from the execution unit, confirm the execution result, and synchronize it to the status monitoring and feedback module.

[0008] Preferably, the precise jump controller receives the target jump time. Then, the state recovery time of each resource track is calculated using a jump compensation algorithm that integrates the hierarchical resource states: ; in, For the first Recovery time of each resource track after a jump This is the length of the pending instruction queue for this resource track at the jump point. The system load rate for this resource type. This represents the hierarchical coefficient of the resource orbit. This is the compensation coefficient, with a value ranging from 0.5 to 2.0. This is the hierarchical weight coefficient, with a value ranging from 0.1 to 0.5, used to adjust the smoothness of jumps between different levels of resources; The jump execution and state synchronization process includes the following steps: S41. Receive the target jump time input by the user. ,verify Is it within the total guided tour duration? If the range is exceeded, an invalid message will be output. S42, if Effective. Pause execution of all current resource tracks and record the current length of the pending instruction queue for each track. and system load rate ; S43, will , and corresponding level coefficients Substitute the jump compensation algorithm to calculate the jump compensation for each track. ; S44. Quickly jump the timeline to... The system locates the position and simultaneously sends status reset commands to each resource orbit, requiring each orbit to reset its position. Internal recovery to The corresponding preset state; S45. Real-time monitoring of the recovery progress of each track status. If a certain track is in... If the recovery process is not yet complete, adjust the compensation coefficient for that orbit. Increase by 10%, recalculate And extend the recovery time; S46. Once all tracks have completed their status recovery, restart the timeline operation and send the jump completion information and the recovery status of each track back to the status monitoring and feedback module.

[0009] Preferably, the status monitoring and feedback module detects the consistency of multi-track resource execution through a status consistency check function and outputs collaborative quality indicators: ; in, As a system synchronization consistency indicator, This represents the number of currently active resource orbitals. For the first The actual execution time of each resource For the expected execution time, The synchronization error tolerance threshold is set between 20ms and 200ms. Score the quality of resource execution. This is the quality threshold, with a value ranging from 0.7 to 0.9; if , The consistency threshold is set between 0.8 and 0.95. The system automatically triggers the synchronization repair process and transmits the abnormal data to the closed-loop collaborative optimization module. Status monitoring and anomaly handling include the following steps: S51, Real-time collection of currently active data. Actual execution time of each resource orbit And execution quality data, calculate each track ; S52. Obtain the expected execution time for each track. Substitute into the consistency test function to calculate and ; S53, will With consistency threshold In comparison, the system connection status and user interaction status are monitored simultaneously. S54, if Furthermore, the system connection is normal, there are no abnormal user interactions, a normal status report is generated and displayed visually; S55, if There may be system connection abnormalities or user interaction abnormalities, triggering the synchronization repair process: S551, locate the abnormal track and the abnormality type; S552, perform time calibration or command retransmission on the abnormal track; S553, re-acquire the abnormal track. and Calculate the repaired S554, if If the condition is met, the repair is complete, and a repair log is recorded. If the condition is still not met, the abnormal data is packaged and sent to the closed-loop collaborative optimization module to wait for optimization parameters. S6. Real-time updates to the visual status feedback interface.

[0010] Preferably, the multi-resource collaborative exhibition hall navigation control system based on timeline orchestration supports rapid scene switching, dynamically loads different timeline orchestration schemes through a scene configuration matrix, and integrates user intent vectors to achieve adaptive scene adjustment. ; in, Indicates the first In the first scenario The activation status of each resource track, 1 for active and 0 for disabled. This is a user intent vector, with dimensions matching the number of resource tracks, and element values... , For Hadama accumulation, Configure the scene matrix for adaptive adjustment; the system indexes the current scene. Automatically switch to the corresponding configuration based on the user intent vector; The scene switching process includes the following steps: S61. Receive scene switching command and obtain target scene index. ; S62. Load the original scene configuration matrix corresponding to the target scene from the system scene library. ; S63. Obtain the user intent vector output by the user intent awareness module. ; S64, Calculation and The Hadamard product is used to obtain the adaptively adjusted product. Tracks with element values ​​less than 0.3 are marked as to be disabled; S65, Analysis Determine the list of resource tracks to be activated and the corresponding timeline arrangement scheme; S66. Pause the execution of all resource tracks in the current scenario, and save the current timeline position and the status of each track; S67. Load the timeline arrangement scheme for the target scene and activate it. The resource tracks marked as active are disabled, and the tracks to be disabled are disabled. S68. Position the timeline to the initial time point of the target scene or the currently saved timeline position, and start running the timeline; S69. Monitor the initial execution status of each track in the target scenario. If there are tracks that fail to activate, re-execute S67 to S68. S610. After the switch is completed, a message indicating successful scene switching is displayed on the visual interface, and a switch log is recorded.

[0011] Preferably, the hierarchical resource track manager uses a resource dependency graph. Managing dependencies between tracks, where vertices Divided by level , , ,side This indicates time dependency or logical dependency; during resource updates, the system automatically checks for intra-layer and inter-layer conflicts based on the dependency graph and adjusts the resource time interval through a conflict resolution function. ; in, The adjusted resource time range The original time interval, This is the conflict adjustment coefficient, with a value ranging from 0.2 to 0.8. , Resource orbits With dependent orbit Trigger time; Resource dependency management and conflict resolving include the following steps: S71. Constructing a resource dependency graph Clearly define each vertex Hierarchical affiliation and boundaries Corresponding dependency type; S72. When a resource track needs to be updated, obtain the original time interval of that track. and trigger time ; S73, Based on Dependency Graph Find all orbitals that are dependent on this orbital. Get its trigger time ; S74. Calculate the relationship between this orbit and each dependent orbit. Trigger time difference Check for any overlapping or conflicting time intervals; S75. If there is no conflict, save the updated resource track information directly; S76. If a conflict exists, substitute the conflict resolution function to calculate the adjusted time interval. ; S77, according to Update the time interval and trigger time of this orbit. ; S78. After verification and update, this track is compatible with all dependent tracks. Do conflicts still exist? If so, adjust accordingly. The value is increased or decreased by 0.1, and S76 to S77 are repeated until there is no conflict. S79. Update the relevant information of this track in the resource dependency graph and synchronize it to the timeline synchronization engine and dynamic triggers.

[0012] Preferably, the user intent perception module generates a user intent vector through user behavior feature extraction and intent classification models. The specific calculations are as follows: ; in, For user behavior feature vectors, The feature weight matrix, For bias vectors, For the activation function, ensure Furthermore, the sum of the user intent weights for all resource tracks is 1; The intent vector generation process includes the following steps: S81. Real-time collection of user behavior data through sensors deployed in the exhibition hall, including dwell time, gaze direction and focus time, number and type of interactive device operations; S82. Preprocess the collected behavioral data: S821. Normalize the dwell time to the [0,1] interval to obtain the normalized dwell time value; S822. Calculate the proportion of gaze focus time to the total dwell time to obtain the gaze focus time proportion; S823. Count the number of interactive operations per unit time, and use the standardized value as the interactive operation feature value. S83. Combine the preprocessed feature values ​​to construct a user behavior feature vector. ; S84, Load the pre-trained feature weight matrix and bias vector ,Will Substitute the values ​​into the calculation formula to obtain the original intent weights; S85, Perform weighting on the original intent. Activation operation to obtain the corresponding resource tracks To form the user intent vector ; S86, Verification Check if the sum of all elements is 1. If the deviation exceeds 0.01, re-execute S85 until the condition is met. S87, will generate Output to the hierarchical resource track manager, dynamic triggers, and scene switching module.

[0013] Preferably, the closed-loop collaborative optimization module dynamically updates the resource scheduling weight coefficient and time axis calibration parameters using the following formula to achieve end-to-end collaborative optimization: ; ; in, This is the updated weight adjustment coefficient. This is the historical average user intent vector. To optimize the step size, a value of 0.1 to 0.3 is used. The updated time synchronization threshold. This is the threshold adjustment coefficient, with a value ranging from 0.2 to 0.5; the system automatically performs an optimization update every 100 to 500 collaborative data entries. The parameter update process includes the following steps: S91. Real-time collection of collaborative data during system operation, including resource scheduling weight coefficients, time synchronization thresholds, , Resource execution status data; S92. Count the number of accumulated collaborative data entries and determine whether the update condition of 100-500 entries has been met; S93. If the target is not met, continue accumulating data; if the target is met, extract the historical average user intent vector. and the current , , ; S94. Substitute the above parameters into the weight adjustment coefficient update formula to calculate... Ensure the sum of the updated coefficients is still 1; if not, adjust proportionally. S5. Substitute into the time synchronization threshold update formula to calculate... ,make sure Within the range of 10ms to 100ms; S96. Send the updated weight adjustment coefficient to the hierarchical resource track manager and the updated time synchronization threshold to the timeline synchronization engine. S97, Verify the updated system Whether to upgrade; if so, save the updated parameters as the new baseline parameters; if not, discard the update results and retain the original parameters. S98. Clear the accumulated collaborative data and restart the data accumulation process.

[0014] This invention provides a multi-resource collaborative exhibition hall navigation control system based on timeline arrangement, which has the following beneficial effects: 1. Improve the accuracy of multi-resource collaboration: Achieve global time uniform calibration through the timeline synchronization engine, and combine the priority dynamic adjustment mechanism of the hierarchical resource track manager to effectively solve the synchronization conflict problem of multiple resources such as audio, video, and lighting, and ensure the smooth and coordinated tour process.

[0015] 2. Optimize navigation operation and navigation experience: The precise navigation controller supports precise navigation within seconds. At the same time, the state compensation algorithm ensures that all resources are quickly restored to the target state after navigation. Combined with the preset scene quick switching function, the convenience and flow of navigation operation are greatly improved.

[0016] 3. Achieve personalized tour guide adaptation: The user intent perception module can capture user behavior characteristics and generate intent vectors. The system dynamically adjusts resource scheduling and scene configuration based on these vectors to make the tour guide service more in line with user interests and improve user immersion and satisfaction.

[0017] 4. Ensure stable system operation: The status monitoring and feedback module monitors the resource execution status, system connection status, and user interaction status in real time, promptly detects and repairs anomalies, reduces the probability of failures during the tour, and improves system reliability.

[0018] 5. Continuous optimization capability: The closed-loop collaborative optimization module dynamically adjusts system parameters and orchestration schemes based on resource execution data and user feedback, enabling the system to continuously optimize performance as usage scenarios change and data accumulates, extending the system lifecycle and adapting to more application scenarios.

[0019] 6. Enhance resource scheduling flexibility: By managing the dependencies between tracks through a resource dependency graph and combining the conditional triggering mechanism of dynamic triggers, flexible resource scheduling and intelligent conflict resolution are achieved, adapting to the tour guide process design requirements of different exhibition halls.

[0020] In summary, the system achieves precise coordination and intelligent conflict resolution of multiple resources such as audio guides, media videos, and lighting through a timeline synchronization engine and hierarchical resource management. Combined with precise jump control and rapid scene switching functions, it enhances the convenience and flow of the tour operation. The user intent perception module adapts to personalized needs, and the status monitoring and feedback mechanism ensures stable system operation. Furthermore, the closed-loop collaborative optimization module enables dynamic parameter updates and continuous performance improvement, comprehensively meeting the core requirements of exhibition hall tours for accuracy, smoothness, personalization, and stability. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of a multi-resource collaborative exhibition hall navigation control system based on time axis arrangement, as described in this invention.

[0022] Figure 2 This is a diagram illustrating a multi-resource collaborative exhibition hall navigation control system based on timeline arrangement, as described in this invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] like Figures 1-2As shown, this invention provides a technical solution: a multi-resource collaborative exhibition hall navigation control system based on timeline orchestration, comprising: a timeline synchronization engine, a hierarchical resource track manager, a user intent perception module, a dynamic trigger, a status monitoring and feedback module, a precise jump controller, and a closed-loop collaborative optimization module; the timeline synchronization engine is used to generate a unified timeline signal based on a global timestamp and drive multiple resource tracks to execute in chronological order; the hierarchical resource track manager is used to manage at least three types of hierarchical resource tracks (basic layer, enhanced layer, and interactive layer), including a navigation audio track, a media video track, a lighting control track, and an optional interactive device track, with resource trigger instructions arranged according to time intervals for each track, supporting inter-layer dependency association and dynamic priority adjustment; the user intent perception module, dynamic trigger, status monitoring and feedback module, precise jump controller, and closed-loop collaborative optimization module; the timeline synchronization engine is used to generate a unified timeline signal based on a global timestamp and drive multiple resource tracks to execute in chronological order; the hierarchical resource track manager is used to manage at least three types of hierarchical resource tracks (basic layer, enhanced layer, and interactive layer), including a navigation audio track, a media video track, a lighting control track, and an optional interactive device track, with resource trigger instructions arranged in ... The image perception module collects user behavior data (dwell time, gaze direction, and interactive operations) through sensors to generate user intent vectors, providing dynamic input for resource scheduling. Dynamic triggers automatically activate corresponding control commands based on user intent and resource status when the timeline reaches the start time point of each resource track. The status monitoring and feedback module monitors the execution status of each resource, system connection status, and user interaction status in real time, providing visualized status feedback. The precise jump controller supports second-level jumps on the timeline based on the input target time point, synchronously updating the status of all track resources and compensating for execution delays. The closed-loop collaborative optimization module dynamically adjusts the timeline orchestration scheme and resource scheduling strategy based on resource execution data, user feedback data, and synchronization consistency indicators.

[0025] More specifically, the timeline synchronization engine calculates the global time synchronization deviation and performs real-time calibration using the following formula: ; in, For time synchronization deviation, This represents the total number of resource orbitals in the system. For the first Local timestamps for each track For global master timestamp; when hour, The preset threshold is set between 10ms and 100ms. The system automatically triggers the time axis calibration mechanism, and the calibration parameters are updated in real time through the closed-loop collaborative optimization module. The calibration process includes the following steps: S11, All in the periodic acquisition system Local timestamp of each resource track and global master timestamp The acquisition period is set to 50ms to 200ms; S12, Collect the data and Substituting into the above formula for calculating time synchronization deviation, we obtain the current... ; S13, Judgment With preset threshold Size relationship; S14, if Record the current synchronization status as normal, and return to S11 to continue periodic data collection; S15, if This triggers the calibration process: S151. Obtain the latest calibration parameters from the closed-loop collaborative optimization module; S152. Adjust the local timestamps of each resource track according to the calibration parameters, with the adjustment amount being... , The calibration coefficient ranges from 0.8 to 1.2. S153. After adjustment, re-collect the local timestamps and global master timestamps of each resource track, and calculate the adjusted values. S54, Judgment Is it ≤ Otherwise, repeat S152 to S153 until the condition is met; S16. Feed back the calibration results to the status monitoring and feedback module and record the calibration log.

[0026] The deviation calculation and calibration scheme of the timeline synchronization engine uses the average deviation between the local timestamps of all resource tracks and the global master timestamp. As a basis for determining synchronization status, it can comprehensively reflect the overall time synchronization level of multiple resource tracks in the system, avoiding local synchronization imbalance caused by time deviation of a single track; a reasonable threshold of 10ms to 100ms is preset. It not only meets the high-precision synchronization requirements of multiple resources such as audio, video, and lighting in exhibition hall tour scenarios, but also avoids the loss of resources due to frequent calibration caused by excessively strict thresholds. At the same time, the periodic acquisition cycle of 50ms to 200ms can realize real-time monitoring of the synchronization status, ensuring that deviations are captured in a timely manner.

[0027] The calibration process incorporates dynamic calibration parameters from the closed-loop collaborative optimization module, combined with an adjustable calibration coefficient of 0.8–1.2. Adjustment amount It can adapt to the response characteristics of different resource orbits and changes in the system operating environment, and then through an iterative calibration mechanism until... By meeting the threshold requirements, the calibration accuracy and stability were effectively guaranteed. Ultimately, the time synchronization deviation of multiple resource tracks was strictly controlled within the allowable range, which significantly reduced the probability of problems such as multi-resource triggering misalignment and lag during the guided tour. This provided users with a coherent and accurate immersive guided tour experience. At the same time, the calibration result feedback and log recording functions also provided reliable data support for system maintenance and performance optimization.

[0028] In this embodiment, taking the tour guide control system of an interactive exhibition hall in a science and technology museum as an example, the system includes four resource tracks: an audio narration track, a video playback track, a lighting adjustment track, and an interactive touch feedback track. ), preset time synchronization threshold The acquisition period is set to 100ms, and the calibration coefficient is... The initial value is 1.0.

[0029] During system operation, the timeline synchronization engine collects the local timestamps of the four tracks every 100ms according to step S11. and global master timestamp Substituting into the formula in step S12 Calculate the current ,because The system triggers the calibration process; step S151 obtains the latest calibration parameters from the closed-loop collaborative optimization module and determines the calibration coefficients. Step S152 according to the adjustment amount The local timestamps of each track are compensated and adjusted; step S153 re-acquires the timestamps of each track and calculates the results. Step S154 judgment The calibration process meets the synchronization requirements. Finally, the calibration results are fed back to the status monitoring and feedback module and logged in step S16. The entire calibration process takes 200ms. After calibration, the tracks of each resource run synchronously and stably. The audio narration, video images and lighting changes are precisely matched. The interactive touch feedback is lag-free, which fully meets the multi-resource collaboration needs of the exhibition hall tour.

[0030] More specifically, the hierarchical resource track manager allocates independent time slice intervals for each resource type and level, and determines the execution priority of each track during conflicting time periods through a resource scheduling weight function that incorporates user intent: ; in, For the first The scheduling weight of each resource track. The duration of this resource orbit. Total guided tour duration, This is the priority coefficient for this resource type. The highest priority coefficient, This represents the historical number of times this resource has been triggered. This represents the average number of system triggers. This represents the user's intent weight for the resource, with a value ranging from 0 to 1, and is output by the user intent perception module. Let be the weighting adjustment coefficient, and satisfy . ; Time slice allocation and priority conflict resolution include the following steps: S21. Get the total guided tour duration and the duration of each resource orbit The time slice intervals of each track are initially divided according to the hierarchical order of the base layer, enhancement layer, and interaction layer to ensure that the time slices of the base layer resources do not overlap. S22. Collect the output of the user intent perception module. Combined with historical trigger data and resource type priority coefficient Substitute into the weighting function to calculate each orbital ; S23. Detect whether there are overlapping conflicts in the time slices of each track. If there are no conflicts, determine the final time slice allocation scheme. S24. If a conflict exists, extract all orbitals within the conflict time period. ; S25, according to Sort from largest to smallest Larger tracks retain their original time slots, while lower-priority tracks have their time slots adjusted to outside of conflicting time periods, with the adjustment not exceeding their own time slot. 20%; S26. Verify whether there are still conflicts after the adjustment. If there are, repeat S24 to S25 until there are no conflicts, and output the final time slice allocation scheme.

[0031] This hierarchical resource track manager achieves refined and personalized scheduling of resource tracks through independent time slice interval allocation and multi-dimensional scheduling weight function design that integrates user intent. It effectively solves the technical pain points of frequent time conflicts of multiple types and multiple levels of resources, fixed scheduling priorities, and inability to adapt to personalized user needs in exhibition hall tour scenarios.

[0032] The scheduling weight function is as follows: ; It integrates four core dimensions: resource duration percentage, inherent type priority, historical trigger popularity, and real-time user intent, and through... The constraints ensure a reasonable weight allocation across all dimensions, guaranteeing the scheduling priority of core resources at the foundational layer (such as audio narration) while dynamically adapting the resource execution order based on user intent, thus enhancing the personalized experience of the guided tour. Time slice allocation follows the core principle of prioritizing the foundational layer without overlap, reducing the probability of key resource conflicts from the source. For conflict scenarios, a resolution mechanism of weighted sorting and limited adjustment is adopted, limiting the adjustment range to no more than 20% of its own duration. This avoids the disruption of the guided tour logic caused by excessive resource scheduling offset. At the same time, iterative verification ensures no residual conflicts, guaranteeing the stability and feasibility of the scheduling scheme. The entire design takes into account the orderliness, flexibility, and personalization of scheduling, providing reliable track management support for multi-resource collaborative guided tours.

[0033] In this embodiment, taking a museum history exhibition hall guided tour control system as an example, the system includes a basic layer audio narration track (k1), an enhanced layer dynamic lighting track (k2) and video playback track (k3), and an interactive layer touch-screen Q&A track (k4), with a total guided tour duration of [missing information]. The durations of each resource orbit are as follows: , , , The priority coefficients for resource types are respectively , , , Maximum priority coefficient The historical trigger counts are respectively , , , Average number of system triggers The weighting adjustment coefficient is set to , , , ,satisfy .

[0034] During system operation, the hierarchical resource track manager initially divides the time slices according to step S21, ensuring that the audio narration track (k1) time slices do not overlap; and in step S22, it collects the intent weights of each track output by the user intent perception module. , , , Substituting into the weighting function, we get: ; ; ; ; Detection S23 revealed an overlap and conflict between the dynamic lighting track (k2) and the video playback track (k3) during the 300s-500s time period; track weights during the conflict period were extracted using S24. , After sorting by S25 The original time slice of the video playback track (k3) is retained, and the time slice of the dynamic light track (k2) is adjusted to 500s-700s, with an adjustment range of 100s, which does not exceed 20% (160s) of its duration of 800s. After verification by S26, there is no time overlap or conflict among the tracks after the adjustment, and the final time slice allocation scheme is output.

[0035] After implementation, each resource track operated in an orderly manner according to the allocation plan. The core audio guide was uninterrupted throughout the entire process, and the video and lighting coordinated to match the exhibition theme. There were no resource conflicts or delays when users triggered touch-based Q&A sessions. The personalized tour experience and scheduling stability both met the design requirements.

[0036] More specifically, the dynamic trigger detects that the timeline has reached the resource trigger point. At that time, a trigger determination function that incorporates user intent is used to determine whether to execute the resource instruction: ; in, This is the current timeline position. To trigger the tolerance time window, the value ranges from 50ms to 500ms. This is the current state of the resource. The threshold for triggering user intent is set, with a value ranging from 0.3 to 0.7; if If so, the corresponding resource control command will be executed immediately; if it is an enhancement layer or interaction layer resource, and If so, the triggering of the resource will be delayed or skipped; The execution process includes the following steps: S31. Real-time monitoring of the current position of the timeline. Compare the trigger points of each resource track ; S32, when When this occurs, a pre-check process is triggered to obtain the current status of the corresponding resource. and the user's intent weight for the resource ; S33, will and Substitute into the trigger determination function for calculation ; S34, if Generate resource control instructions and send them to the corresponding execution unit, while recording the trigger time and resource identifier; S5, if Determine the resource type: S351. If it is a basic layer resource, output a trigger exception prompt to the status monitoring and feedback module, and still execute the basic control commands; S352. If it is an enhancement layer resource, delay the trigger point to , The delay duration is 1 to 3 seconds. After the delay, S31 to S33 are re-executed. S353: If it is an interactive layer resource, skip the trigger point directly and record the reason for skipping and the user intent data. S36. Receive instruction execution feedback from the execution unit, confirm the execution result, and synchronize it to the status monitoring and feedback module.

[0037] This dynamic trigger uses a three-dimensional triggering decision function that integrates time synchronization conditions, resource status, and user intent. The design, along with the hierarchical resource differentiation mechanism, effectively solves the technical pain points of traditional triggers, such as rigid trigger timing, neglect of user needs, and disconnection of resource coordination.

[0038] Among them, the trigger determination function is ( The value (ranging from 50ms to 500ms) ensures accurate time synchronization, avoiding false triggering due to minor time deviations while reserving reasonable fault tolerance for resource response; Ensure the feasibility of instruction execution and prevent invalid triggers from consuming system resources; ( (Values ​​range from 0.3 to 0.7) to achieve precise matching between user intent and resource triggers, making the tour guide service tailored to user interests.

[0039] For resources at different levels, basic-level resources are enforced and exceptions are reported to ensure the core navigation flow remains uninterrupted; enhanced-level resources are re-verified after a 1-3 second delay to balance user experience continuity with personalized adaptation; interactive-level resources skip low-intent trigger points directly, reducing redundant execution and improving system efficiency. The entire trigger execution process, through a closed-loop design of pre-checking, layered judgment, and result feedback, achieves both the accuracy and orderliness of multi-resource collaborative triggering and personalized navigation adaptation through user intent integration. Meanwhile, logging and status synchronization provide data support for system maintenance and optimization.

[0040] In this embodiment, taking an art gallery tour guide control system as an example, the system includes a basic layer audio track for exhibit explanation (k1), an enhanced layer ambient lighting adjustment track (k2), and an interactive layer exhibit detail touch query track (k3), with a set trigger tolerance time window. User intent trigger threshold Enhancement layer delay duration .

[0041] During system operation, the dynamic trigger monitors the time axis in real time according to step S31. When the current time axis position is detected... , and audio track trigger point satisfy At that time, perform the S32 pre-check to obtain the audio resource status. User intent weighting for audio Calculated by S33 The audio playback command is generated and sent according to S34, and the trigger time and resource identifier are recorded; then the timeline runs to... , and the trigger point of the light track When the time condition is met, S32 obtains the status of the lighting resources. User intent weight S33 calculation Since k2 is an enhancement layer resource, the trigger point is delayed until S352. After a delay, S31 to S33 are re-executed. At this time, the user intent weight remains unchanged and is still satisfied. After another delay, users' attention focused on the exhibits. Rising to 0.6 This triggers a lighting adjustment command; when the timeline reaches... , and the touch track trigger point If the time condition is met, S32 obtains the user intent weight. S33 calculation Since k3 is an interactive layer resource, press S353 to skip the trigger point and record the reason; after all instructions are executed, S36 receives feedback from the execution unit and synchronizes it to the status monitoring and feedback module. Throughout the process, the triggers of each resource are precisely adapted to the user's behavior, the core audio is uninterrupted, the lighting adjustment is tailored to the user's interests, redundant touch triggers are filtered out, the system runs efficiently and the personalized experience is outstanding.

[0042] More specifically, the precise jump controller receives the target jump time. Then, the state recovery time of each resource track is calculated using a jump compensation algorithm that integrates the hierarchical resource states: ; in, For the first Recovery time of each resource track after a jump This is the length of the pending instruction queue for this resource track at the jump point. The system load rate for this resource type. This is the hierarchical coefficient for resource orbitals (0.1 for the basic layer, 0.3 for the enhancement layer, and 0.6 for the interaction layer). This is the compensation coefficient, with a value ranging from 0.5 to 2.0. This is the hierarchical weight coefficient, with a value ranging from 0.1 to 0.5, used to adjust the smoothness of jumps between different levels of resources; The jump execution and state synchronization process includes the following steps: S41. Receive the target jump time input by the user. ,verify Is it within the total guided tour duration? If the range is exceeded, an invalid message will be output. S42, if Effective. Pause execution of all current resource tracks and record the current length of the pending instruction queue for each track. and system load rate ; S43, will , and corresponding level coefficients Substitute the jump compensation algorithm to calculate the jump compensation for each track. ; S44. Quickly jump the timeline to... The system locates the position and simultaneously sends status reset commands to each resource orbit, requiring each orbit to reset its position. Internal recovery to The corresponding preset state; S45. Real-time monitoring of the recovery progress of each track status. If a certain track is in... If the recovery process is not yet complete, adjust the compensation coefficient for that orbit. Increase by 10%, recalculate And extend the recovery time; S46. Once all tracks have completed their status recovery, restart the timeline operation and send the jump completion information and the recovery status of each track back to the status monitoring and feedback module.

[0043] This precise jump controller effectively solves the technical pain points of traditional exhibition hall navigation systems, such as asynchronous resource status recovery, significant jump delay, and poor adaptability of hierarchical resources, by integrating a jump compensation algorithm based on hierarchical resource status and a closed-loop jump execution process design. It achieves precise and smooth jumps in the navigation process.

[0044] Among them, the jump compensation algorithm Innovatively incorporates the length of the pending instruction queue System load rate and hierarchical coefficient (Base layer 0.1, Enhancement layer 0.3, Interaction layer 0.6) Three core state parameters, combined with an adjustable compensation coefficient of 0.5 to 2.0. With hierarchical weight coefficients of 0.1 to 0.5 It can dynamically calculate differentiated state recovery time based on the real-time operating load and hierarchical attributes of different resource tracks, thereby fundamentally avoiding synchronization imbalance problems caused by differences in resource states after jumps.

[0045] The jump execution process follows a closed-loop logic of target verification, pause recording, precise calculation, jump reset, real-time monitoring, iterative adjustment, and restart feedback. Invalid jumps are prevented by verifying the validity of the target time; the accuracy of basic jump data is ensured through full-track pause and status recording; and incomplete recovery tracks are addressed through... The 10% incremental adjustment mechanism ensures that all resources can be stably restored to the target state, ultimately achieving precise jumps within seconds while ensuring the continuity and stability of the navigation process. In addition, the synchronization of jump completion information and recovery status feedback provides accurate data input for the status monitoring and closed-loop collaborative optimization modules, helping to continuously improve the overall system performance.

[0046] In this embodiment, taking a digital exhibition hall guide control system of a history museum as an example, the system includes a basic layer guide audio track (k1), an enhanced layer artifact video playback track (k2), and an interactive layer AR track (k3), with a total guide duration of [missing information]. Preset compensation coefficient Hierarchical weight coefficient Users input the target redirection time through the navigation terminal. The system verifies in step S41 that the time falls within the total tour duration of 0-1800 seconds and determines it as a valid jump instruction; it then executes step S42 to pause the execution of all current resource tracks and synchronously records key status data for each track: the length of the pending instruction queue for the tour audio track (k1). System load rate The artifact video playback track (k2) , AR interactive track (k3) , Meanwhile, the coefficients for each orbital level were determined as follows: (Basic layer) (Enhancement layer) (Interaction Layer); S43 substitutes the above parameters into the jump compensation algorithm to calculate the recovery time of each track: ; ; ; S44 rapidly jumps the system timeline to the 600s position and simultaneously sends status reset commands to the three resource tracks, requiring each track to reset its corresponding status. The system restores the system to the preset state corresponding to 600 seconds (audio playback of the corresponding artifact introduction clip, video display of the corresponding artifact details, and AR interaction ready); S45 monitors the restoration progress of each track in real time and finds that the artifact video playback track (k2) has not completed the state restoration at 607.23 seconds, and then adjusts its compensation coefficient. Increase by 10% to 1.1 and recalculate the recovery time: ; Extend the recovery time; S46 continues to monitor and restarts the time axis to normal operation after all tracks have completed their status recovery within the adjusted recovery time. It also feeds back the jump completion information, the recovery time of each track, and the status data to the status monitoring and feedback module.

[0047] The implementation verification shows that the entire jump took only 0.8 seconds. After the jump, the audio, cultural relic images, and AR interactive resources ran synchronously without any lag or misalignment. The tour guide process was smooth and seamless, fully meeting the accurate jump requirements in the exhibition hall tour scenario.

[0048] More specifically, the status monitoring and feedback module detects the consistency of multi-track resource execution through a status consistency check function and outputs collaborative quality indicators: ; in, As a system synchronization consistency indicator, This represents the number of currently active resource orbitals. For the first The actual execution time of each resource For the expected execution time, The synchronization error tolerance threshold is set between 20ms and 200ms. The resource execution quality score is calculated based on a combination of audio clarity, video smoothness, and lighting accuracy. This is the quality threshold, with a value ranging from 0.7 to 0.9; if , The consistency threshold is set between 0.8 and 0.95. The system automatically triggers the synchronization repair process and transmits the abnormal data to the closed-loop collaborative optimization module. Status monitoring and anomaly handling include the following steps: S51, Real-time collection of currently active data. Actual execution time of each resource orbit And execution quality data, calculate each track ; S52. Obtain the expected execution time for each track. Substitute into the consistency test function to calculate and ; S53, will With consistency threshold In comparison, the system connection status and user interaction status are monitored simultaneously. S54, if Furthermore, the system connection is normal, there are no abnormal user interactions, a normal status report is generated and displayed visually; S55, if There may be system connection abnormalities or user interaction abnormalities, triggering the synchronization repair process: S551, locate the abnormal track and the abnormality type; S552, perform time calibration or command retransmission on the abnormal track; S553, re-acquire the abnormal track. and Calculate the repaired S554, if If the condition is met, the repair is complete, and a repair log is recorded. If the condition is still not met, the abnormal data is packaged and sent to the closed-loop collaborative optimization module to wait for optimization parameters. S6. Real-time updates to the visual status feedback interface.

[0049] The status monitoring and feedback module, through the design of a status consistency verification function that integrates time synchronization accuracy and execution quality, combined with a closed-loop anomaly handling process, constructs a full-link, high-precision system status control system, effectively solving the technical pain points of traditional exhibition hall guide systems such as one-sided status monitoring, ambiguous anomaly location, and delayed feedback.

[0050] Among them, the consistency test function and The combined design, through ( (Values ​​ranging from 20ms to 200ms) ensure the time synchronization accuracy of multi-resource execution, and also through... ( (Values ​​range from 0.7 to 0.9) This balances the quality of resource execution, achieving dual verification of "time + quality". With consistency threshold The comparison and judgment mechanism of (0.8~0.95) can accurately identify system synchronization imbalance and various abnormal states. The triggered synchronization repair process quickly resolves minor abnormalities through a closed-loop logic of location, processing and re-inspection. For complex abnormalities, it links the closed-loop collaborative optimization module to obtain optimization parameters, forming a full-link control of monitoring, repair and optimization. At the same time, the real-time visualization status feedback and log recording function not only provide operation and maintenance personnel with an intuitive reference for the system operation status, but also provide accurate abnormal data support for the closed-loop collaborative optimization module, which greatly improves the stability and maintainability of the system operation and builds a solid protection line for the immersive tour experience.

[0051] In this embodiment, taking a science and technology exhibition hall guide control system as an example, the system currently activates three resource tracks: the basic layer audio narration track (k1), the enhanced layer dynamic projection track (k2), and the interactive layer touch demonstration track (k3). Preset synchronization error tolerance threshold Execution quality threshold Consistency threshold .

[0052] During system operation, the status monitoring and feedback module collects the actual execution time and quality data of each track in real time according to step S51: audio narration track (k1) , Dynamic projection track (k2) , Touchscreen demo track (k3) , S52 obtains the expected execution time for each track. Substitute into the consistency test function to calculate: orbital k1 and Therefore K2 orbital Therefore K3 track and Therefore , and then calculate S53 comparison This triggers the S55 anomaly handling process; S551 identifies the abnormal track as k2 (dynamic projection track), with the anomaly type being excessive time synchronization deviation; S552 performs a time calibration operation on track k2, with a calibration increment of -150ms; S553 reacquires data from track k2. , Calculated After repair S554 determines that the repair is complete and records the repair log; S56 updates the visualization interface in real time, showing that the status of each track is normal and the synchronization consistency index is 1.0.

[0053] After implementation, the system can accurately monitor the operating status of multiple resources, with an anomaly response time of ≤300ms and a repair success rate of over 95%. The visual interface clearly presents the synchronization status and quality score of each track, significantly improving the convenience of operation and maintenance and the stability of system operation, providing a reliable guarantee for the continuous and stable operation of the exhibition hall tour.

[0054] More specifically, the multi-resource collaborative exhibition hall navigation control system based on timeline orchestration supports rapid scene switching. It dynamically loads different timeline orchestration schemes through a scene configuration matrix and integrates user intent vectors to achieve adaptive scene adjustment. ; in, Indicates the first In the first scenario The activation status of each resource track, 1 for active and 0 for disabled. This is a user intent vector, with dimensions matching the number of resource tracks, and element values... , For Hadama accumulation, Configure the scene matrix for adaptive adjustment; the system indexes the current scene. Automatically switch to the corresponding configuration based on the user intent vector; The scene switching process includes the following steps: S61. Receive scene switching command and obtain target scene index. ; S62. Load the original scene configuration matrix corresponding to the target scene from the system scene library. ; S63. Obtain the user intent vector output by the user intent awareness module. ; S64, Calculation and The Hadamard product is used to obtain the adaptively adjusted product. Tracks with element values ​​less than 0.3 are marked as to be disabled; S65, Analysis Determine the list of resource tracks to be activated and the corresponding timeline arrangement scheme; S66. Pause the execution of all resource tracks in the current scenario, and save the current timeline position and the status of each track; S67. Load the timeline arrangement scheme for the target scene and activate it. The resource tracks marked as active are disabled, and the tracks to be disabled are disabled. S68. Position the timeline to the initial time point of the target scene or the currently saved timeline position (select according to the switching command) and start the timeline running; S69. Monitor the initial execution status of each track in the target scenario. If there are tracks that fail to activate, re-execute S67 to S68. S610. After the switch is completed, a message indicating successful scene switching is displayed on the visual interface, and a switch log is recorded.

[0055] The scenario-based rapid switching function of the multi-resource collaborative exhibition hall navigation control system based on time axis arrangement effectively solves the technical pain points of traditional exhibition hall navigation systems, such as long scene switching time, rigid resource adaptation, and inability to match users' personalized needs, through the fusion design of scene configuration matrix and user intent vector, combined with standardized switching execution process, and achieves rapid, accurate and adaptive scene switching.

[0056] Among them, the scene configuration matrix By clearly defining the activation status of resource tracks in various scenarios using binary elements, structured data support is provided for rapid loading in different scenarios; the Hadamard product operation is innovatively introduced. , user intent vector Deeply integrated with the original scene configuration matrix, the system adaptively activates and disables resource tracks by filtering element values ​​(values ​​less than 0.3 are marked as to be disabled), ensuring that scene configurations accurately match user interests and preferences. The switching process follows a standardized closed-loop logic of command reception, matrix loading, intent fusion, state saving, scheme loading, startup monitoring, and feedback recording. It ensures switching continuity by pausing the current scene and saving its state, and guarantees switching reliability through target scene initial state monitoring and retry mechanisms. It also supports two positioning modes: initial time point or current location, improving switching flexibility. The entire design achieves millisecond-level rapid scene switching and enhances the personalized experience of the navigation through adaptive adjustment of user intent. Switching logs and visual feedback also provide accurate data support for system operation and maintenance and scene optimization, significantly improving the system's scene adaptability and user experience satisfaction.

[0057] In this embodiment, taking a comprehensive museum exhibition hall guide control system as an example, the system includes three scenes: "Ancient Civilization," "Modern Development," and "Modern Technology." It covers four resource tracks: audio explanation (k1), dynamic lighting (k2), video playback (k3), and AR interaction (k4). The original scene configuration matrix stored in the scene library is The first row corresponds to the ancient civilization scenario (activate k1, k2, k3, disable k4), the second row corresponds to the modern development scenario (activate k1, k3, k4, disable k2), and the third row corresponds to the modern technology scenario (activate k1, k2, k4, disable k3).

[0058] The user is currently in an ancient civilization scenario, triggering a command to switch to a modern technology scenario (target scenario index). The system receives and obtains instructions according to step S61. S62 loads the original scene configuration matrix corresponding to the modern technology scene from the scene library. (Single-row matrix, corresponding to the third row); S63 obtains the user intent vector output by the user intent awareness module. (The elements correspond to k1-k4 respectively) ); S64 Calculate the Hadamard product: ; Track k2 has an element value of 0.2 < 0.3, so it is marked as to be disabled; S65 resolves that the tracks to be activated are k1 and k4, and the corresponding modern technology scene timeline arrangement scheme is audio explanation and AR interactive collaboration; S66 pauses the execution of all tracks in the current ancient civilization scene and saves the current timeline position. S67 Loads the modern technology scene timeline arrangement scheme, activates tracks k1 and k4, and disables tracks k2 and k3; S68 Selects to continue based on the current saved position according to the switching instruction, sets the timeline to 450s and starts running; S69 Monitoring shows that track k4 is activated normally, and the initial execution state of all tracks is stable; S610 Outputs a successful switching prompt to the visualization interface and records the switching log (switching time 0.5s, tracks k1 and k4 activated, tracks k2 and k3 disabled).

[0059] After implementation, scene switching was fast and smooth without any lag or interruption. The activated audio narration and AR interaction accurately matched the user's interest in modern technology. The disabled low-intent tracks (k2, k3) did not occupy system resources. After switching, all tracks ran synchronously and stably. Both personalized adaptation and switching efficiency met the design requirements.

[0060] More specifically, the hierarchical resource track manager uses a resource dependency graph. Managing dependencies between tracks, where vertices Divided by level (Basic layer) (Enhancement layer) (Interaction layer), edge This indicates time dependency or logical dependency; during resource updates, the system automatically checks for intra-layer and inter-layer conflicts based on the dependency graph and adjusts the resource time interval through a conflict resolution function. ; in, The adjusted resource time range The original time interval, This is the conflict adjustment coefficient, with a value ranging from 0.2 to 0.8. , Resource orbits With dependent orbit Trigger time; Resource dependency management and conflict resolving include the following steps: S71. Constructing a resource dependency graph Clearly define each vertex Hierarchical affiliation and boundaries Corresponding dependency types (time dependency / logical dependency); S72. When a resource track needs to be updated, obtain the original time interval of that track. and trigger time ; S73, Based on Dependency Graph Find all orbitals that are dependent on this orbital. Get its trigger time ; S74. Calculate the relationship between this orbit and each dependent orbit. Trigger time difference Check for any overlapping or conflicting time intervals; S75. If there is no conflict, save the updated resource track information directly; S76. If a conflict exists, substitute the conflict resolution function to calculate the adjusted time interval. ; S77, according to Update the time interval and trigger time of this orbit. ; S78. After verification and update, this track is compatible with all dependent tracks. Do conflicts still exist? If so, adjust accordingly. The value is increased or decreased by 0.1, and S76 to S77 are repeated until there is no conflict. S79. Update the relevant information of this track in the resource dependency graph and synchronize it to the timeline synchronization engine and dynamic triggers.

[0061] This hierarchical resource track manager's resource dependency management scheme, through the collaborative design of hierarchical resource dependency graph construction and dynamic conflict resolution functions, builds a visualized and traceable resource dependency control system. It effectively solves the technical pain points of traditional exhibition hall navigation systems, such as ambiguous dependencies between resource tracks, low efficiency in investigating conflicts within and between layers, and strong blindness in time interval adjustments, and further enhances the orderliness and reliability of multi-resource scheduling.

[0062] Among them, the resource dependency graph Innovatively, the vertex By base layer reinforcement layer Interaction layer Perform hierarchical classification, through edges Clearly marking the time dependencies (such as light and shadow being triggered according to the rhythm of the explanation) or logical dependencies (such as VR experiences needing to be executed after the explanation) between tracks makes the complex multi-track relationships structured and visualized, providing core support for the accurate location and rapid investigation of conflicts.

[0063] Conflict resolution function It deeply integrates the orbital trigger time difference with a conflict adjustment coefficient adjustable from 0.2 to 0.8. It can dynamically calculate the adjustment amount over the time interval based on the tightness of dependencies, ensuring effective conflict resolution while avoiding over-adjustment that could break the navigation logic. The entire resource dependency management process follows a closed-loop logic of dependency building, update detection, conflict investigation, dynamic adjustment, iterative verification, and information synchronization. It achieves rapid conflict location through dependency graphs. Iterative adjustments to the coefficients ensure accuracy, ultimately guaranteeing stable dependencies between tracks and no overlapping or conflicting time intervals after resource updates. Simultaneously, the updated track information is synchronized to the timeline synchronization engine and dynamic triggers, ensuring overall system consistency and laying a solid foundation for dependency management for efficient multi-resource collaborative navigation.

[0064] In this embodiment, taking the "Space Exploration" themed exhibition hall guide control system of a science and technology museum as an example, the system includes a basic layer of aerospace knowledge audio explanation track (k1), an enhanced layer of aerospace video playback track (k2) and simulated lighting track (k3), and an interactive layer of space capsule VR experience track (k4), constructing a resource dependency graph. In the middle, the vertex level belongs to , , ,side The corresponding dependencies are as follows: k2 and k1 are time-dependent (video playback and narration are triggered synchronously), k3 and k1 are time-dependent (lighting changes match the narration rhythm), and k4 and k1 are logically dependent (the VR experience needs to be triggered after the "spacecraft structure" narration segment ends). A preset conflict adjustment coefficient is also provided. .

[0065] During system operation, the material content of the space image playback track (k2) needs to be updated, including its original time interval. Original trigger time ; Follow step S71 to confirm that the hierarchical affiliation and dependency information of the resource dependency graph are correct; S72 to obtain the original time interval of k2. and trigger time S73 uses a dependency graph to find track l, which is k1, that has a dependency relationship with k2, and obtains the trigger time of the explanatory segment corresponding to k1. S74 calculates the trigger time difference. Meanwhile, the inspection revealed an overlap and conflict between the original time interval [500s, 700s] of k2 and the time interval [550s, 650s] of the simulated light track k3 in the same enhancement layer; S76 substituted the conflict resolution function to calculate the adjusted time interval: S77 updates the time interval of k2 to [502.5s, 702.5s], triggering a time synchronization update. S78 verification revealed that the updated k2 time interval still partially overlaps and conflicts with k3. Adjusted to 0.7 (increase / decrease by 0.1), recalculated. Further verification revealed a small amount of overlap; [further details needed] Adjusted to 0.8, the calculation yields... Verification showed that k2, k3, and other dependent tracks had no time overlap or conflict; S79 updated the time information of k2 in the resource dependency graph and synchronized it to the time axis synchronization engine and dynamic triggers.

[0066] After implementation, the K2 track operated stably according to the updated time interval, the playback of aerospace images matched the rhythm of the K1 explanation precisely, there was no conflict or interference with the K3 simulated light track, and the K4 VR experience track could also be triggered normally after the corresponding explanation segment ended. The resource dependency relationship was stable and controllable, the overall system operated smoothly, and it fully met the multi-resource scheduling requirements after the resource update.

[0067] More specifically, the user intent perception module generates a user intent vector through user behavior feature extraction and intent classification models. The specific calculations are as follows: ; in, This is a user behavior feature vector (including normalized dwell time, percentage of gaze focus time, number of interaction operations, etc.). The feature weight matrix, For bias vectors, For the activation function, ensure Furthermore, the sum of the user intent weights for all resource tracks is 1; The intent vector generation process includes the following steps: S81. Real-time collection of user behavior data through sensors deployed in the exhibition hall, including dwell time, gaze direction and focus time, number and type of interactive device operations; S82. Preprocess the collected behavioral data: S821. Normalize the dwell time to the [0,1] interval to obtain the normalized dwell time value; S822. Calculate the proportion of gaze focus time to the total dwell time to obtain the gaze focus time proportion; S823. Count the number of interactive operations per unit time, and use the standardized value as the interactive operation feature value. S83. Combine the preprocessed feature values ​​to construct a user behavior feature vector. ; S84, Load the pre-trained feature weight matrix and bias vector ,Will Substitute the values ​​into the calculation formula to obtain the original intent weights; S85, Perform weighting on the original intent. Activation operation to obtain the corresponding resource tracks To form the user intent vector ; S86, Verification Check if the sum of all elements is 1. If the deviation exceeds 0.01, re-execute S85 until the condition is met. S87, will generate Output to the hierarchical resource track manager, dynamic triggers, and scene switching module.

[0068] This user intent perception module constructs an accurate and reliable user intent perception system through the fusion of multi-dimensional user behavior features and the Softmax normalization intent vector generation scheme. It effectively solves the technical pain points of traditional exhibition hall guidance systems, such as one-sided user intent perception, unreasonable weight allocation, and inability to accurately match the needs of multiple resource adaptation, and provides core support for personalized guidance collaboration.

[0069] Among them, key behavioral features such as the normalized value of dwell time, the proportion of eye focus time, and the number of interactive operations per unit time were innovatively selected to construct the system. It comprehensively captures users' interests in different exhibits and resources, avoiding misjudgments of intent caused by single behavioral features; through a pre-trained feature weight matrix. With bias vector Achieve precise mapping from features to intent weights, and combine this with the Softmax activation function to ensure the accuracy of each resource track. The presence of features in the [0,1] interval with a sum of 1 ensures the rationality and interpretability of the intent vector. The entire generation process follows a closed-loop logic of real-time acquisition, standardized preprocessing, feature vector construction, intent calculation, normalization verification, and multi-module output. Data preprocessing improves feature quality, and iterative verification ensures the accuracy of the intent vector, ultimately generating... The system synchronously outputs to the hierarchical resource track manager, dynamic triggers, and scene switching module, achieving a deep integration of user intent with resource scheduling, trigger control, and scene adaptation. This transforms the tour guide service from fixed push notifications to precise matching, significantly enhancing the user's immersive experience and the personalization of the tour guide service.

[0070] In this embodiment, taking an art gallery tour guide control system as an example, the system includes three resource tracks: audio narration (k1), video introduction (k2), and AR detail viewing (k3). The user intent perception module presets a feature weight matrix. (Rows correspond to resource tracks, columns correspond to normalized dwell time, percentage of gaze focus time, and standardized number of interactions per unit time), bias vector .

[0071] During system operation, step S81 collects user behavior data in front of a certain oil painting exhibit using infrared sensors, cameras, and touch-screen interactive terminals deployed in the exhibition hall: dwell time 180s (the maximum dwell time for this exhibit is preset to 300s), total time the user's gaze is focused on the oil painting and the corresponding video screen 126s, and 3 touch query operations within a unit time (60s) (the system's preset interaction threshold is 5 times); Step S82 performs preprocessing: S821 Normalized dwell time = 180s / 300s = 0.6, S822 Percentage of gaze focused time = 126s / 180s = 0.7, S823 Standardized interaction operation count = 3 / 5 = 0.6, constructing a user behavior feature vector. S84 loading and Calculate the original intent weight: The original weight of k1 = 0.4 × 0.6 + 0.3 × 0.7 + 0.3 × 0.6 + 0.05 = 0.24 + 0.21 + 0.18 + 0.05 = 0.68; The original weight of k2 = 0.3 × 0.6 + 0.4 × 0.7 + 0.3 × 0.6 + 0.05 = 0.18 + 0.28 + 0.18 + 0.05 = 0.69; The original weight of k3 = 0.2 × 0.6 + 0.2 × 0.7 + 0.6 × 0.6 + 0.05 = 0.12 + 0.14 + 0.36 + 0.05 = 0.67; S85 performs the Softmax activation operation: after exponentiation, k1 = e 0 · 68 ≈1.974, k2=e 0 · 69 ≈1.994, k3=e 0 · 67 ≈1.955, total ≈5.923; After normalization , , ,composition ; S86 verifies that the sum of the elements = 0.333 + 0.337 + 0.330 = 1.0, the deviation is 0, and the requirement is met; S87 will... The output is sent to the hierarchical resource track manager, dynamic triggers, and scene switching module. After implementation, the system prioritizes scheduling the k2 (video introduction) resource based on this intent vector, and synchronously adapts the triggering timing of k1 (audio explanation) and k3 (AR viewing), accurately matching the user's need for in-depth understanding of the oil painting exhibits, resulting in a significant personalized tour adaptation effect.

[0072] More specifically, the closed-loop collaborative optimization module dynamically updates the resource scheduling weight coefficients and time axis calibration parameters using the following formula to achieve end-to-end collaborative optimization: ; ; in, This is the updated weight adjustment coefficient. This is the historical average user intent vector. To optimize the step size, a value of 0.1 to 0.3 is used. The updated time synchronization threshold. This is the threshold adjustment coefficient, with a value ranging from 0.2 to 0.5; the system automatically performs an optimization update every 100 to 500 collaborative data entries. The parameter update process includes the following steps: S91. Real-time collection of collaborative data during system operation, including resource scheduling weight coefficients, time synchronization thresholds, , Resource execution status data; S92. Count the number of accumulated collaborative data entries and determine whether the update condition of 100-500 entries has been met; S93. If the target is not met, continue accumulating data; if the target is met, extract the historical average user intent vector. and the current , , ; S94. Substitute the above parameters into the weight adjustment coefficient update formula to calculate... Ensure the sum of the updated coefficients is still 1; if not, adjust proportionally. S5. Substitute into the time synchronization threshold update formula to calculate... ,make sure Within the range of 10ms to 100ms; S96. Send the updated weight adjustment coefficient to the hierarchical resource track manager and the updated time synchronization threshold to the timeline synchronization engine. S97, Verify the updated system Whether to upgrade; if so, save the updated parameters as the new baseline parameters; if not, discard the update results and retain the original parameters. S98. Clear the accumulated collaborative data and restart the data accumulation process.

[0073] The closed-loop collaborative optimization module, by integrating synchronization consistency indicators and dynamic parameter update schemes based on user intent deviations, combined with data accumulation triggering mechanisms and closed-loop verification processes, constructs an end-to-end system collaborative optimization system. This effectively solves the technical pain point of traditional exhibition hall guidance systems where core parameters such as resource scheduling weights and time synchronization thresholds are rigid and inflexible, unable to adapt to changes in system operating status and differences in user intent, leading to a decline in collaborative effects after long-term operation. This enables continuous iterative improvement of system performance.

[0074] Among them, the formula for updating the weight adjustment coefficient. It innovatively links synchronization consistency indicators. Deviation from user intent The optimization step size can be adjusted from 0.1 to 0.3. This allows the weighting coefficients to be dynamically adjusted based on the user's real-time intent and the system's synchronization status, ensuring that resource scheduling always matches user needs and the optimal system operating state; the time synchronization threshold update formula... based on The feedback dynamically optimizes the threshold, using a threshold adjustment coefficient of 0.2 to 0.5. The threshold is flexibly adapted to ensure synchronization accuracy while avoiding frequent calibration losses caused by overly strict thresholds. The accumulation triggering mechanism of 100 to 500 collaborative data points ensures the statistical reliability of the optimization parameters and avoids incorrect parameter updates caused by single or a small number of abnormal data.

[0075] The entire process follows a closed-loop logic of data accumulation, condition determination, parameter calculation, constraint verification, module deployment, effect verification, and data clearing. Parameter rationality is ensured through constraint verification that the sum of coefficients is 1 and the threshold range is met. Enhancing the effectiveness of the optimization process ensures its effectiveness, ultimately enabling precise iteration of resource scheduling and time synchronization parameters. This allows the system's collaborative performance to be continuously optimized as operational data accumulates, providing core technological support for a long-term, stable, and accurate immersive guided tour experience.

[0076] 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 multi-resource collaborative exhibition hall navigation control system based on timeline arrangement, characterized in that, include: The timeline synchronization engine is used to generate a unified timeline signal based on a global timestamp and drive multiple resource tracks to execute in chronological order. The hierarchical resource track manager is used to manage at least three types of hierarchical resource tracks, including a guide audio track, a media video track, a lighting control track, and an optional interactive device track. Each track arranges resource trigger instructions according to time intervals and supports dynamic adjustment of inter-layer dependencies and priorities. The user intent perception module collects user behavior data through sensors to generate user intent vectors, providing dynamic input for resource scheduling. Dynamic triggers are used to automatically trigger corresponding control commands when the timeline reaches the start time point of each resource track, based on user intent and resource status. The status monitoring and feedback module is used to monitor the execution status of each resource, the system connection status, and the user interaction status in real time, and provides visual status feedback. The precise jump controller supports second-level jumps on the timeline based on the input target time point, synchronously updates the status of all track resources, and compensates for execution delays. The closed-loop collaborative optimization module dynamically adjusts the timeline orchestration scheme and resource scheduling strategy based on resource execution data, user feedback data, and synchronization consistency indicators.

2. The multi-resource collaborative exhibition hall navigation control system based on time axis arrangement according to claim 1, characterized in that, The timeline synchronization engine calculates the global time synchronization deviation and performs real-time calibration using the following formula: ； in, For time synchronization deviation, This represents the total number of resource orbitals in the system. For the first Local timestamps for each track For global master timestamp; when hour, The preset threshold is set between 10ms and 100ms. The system automatically triggers the time axis calibration mechanism, and the calibration parameters are updated in real time through the closed-loop collaborative optimization module. The calibration process includes the following steps: S11, All in the periodic acquisition system Local timestamp of each resource track and global master timestamp The acquisition period is set to 50ms to 200ms; S12, Collect the data and Substituting into the above formula for calculating time synchronization deviation, we obtain the current... ; S13, Judgment With preset threshold Size relationship; S14, if Record the current synchronization status as normal, and return to S11 to continue periodic data collection; S15, if This triggers the calibration process: S151. Obtain the latest calibration parameters from the closed-loop collaborative optimization module; S152. Adjust the local timestamps of each resource track according to the calibration parameters, with the adjustment amount being... , The calibration coefficient ranges from 0.8 to 1.

2. S153. After adjustment, re-collect the local timestamps and global master timestamps of each resource track, and calculate the adjusted values. S54, Judgment Is it ≤ Otherwise, repeat S152 to S153 until the condition is met; S16. Feed back the calibration results to the status monitoring and feedback module and record the calibration log.

3. The multi-resource collaborative exhibition hall navigation control system based on time axis arrangement according to claim 2, characterized in that, The hierarchical resource track manager allocates independent time slice intervals for each resource type and level, and determines the execution priority of each track during conflicting time periods through a resource scheduling weight function that incorporates user intent. ； in, For the first The scheduling weight of each resource track. The duration of this resource orbit. Total guided tour duration, This is the priority coefficient for this resource type. The highest priority coefficient, This represents the historical number of times this resource has been triggered. This represents the average number of system triggers. This represents the user's intent weight for the resource, with a value ranging from 0 to 1, and is output by the user intent perception module. Let be the weighting adjustment coefficient, and satisfy . ; Time slice allocation and priority conflict resolution include the following steps: S21. Get the total guided tour duration and the duration of each resource orbit The time slice intervals of each track are initially divided according to the hierarchical order of the base layer, enhancement layer, and interaction layer to ensure that the time slices of the base layer resources do not overlap. S22. Collect the output of the user intent perception module. Combined with historical trigger data and resource type priority coefficient Substitute into the weighting function to calculate each orbital ; S23. Detect whether there are overlapping conflicts in the time slices of each track. If there are no conflicts, determine the final time slice allocation scheme. S24. If a conflict exists, extract all orbitals within the conflict time period. ; S25, according to Sort from largest to smallest Larger tracks retain their original time slots, while lower-priority tracks have their time slots adjusted to outside of conflicting time periods, with the adjustment not exceeding their own time slot. 20%; S26. Verify whether there are still conflicts after the adjustment. If there are, repeat S24 to S25 until there are no conflicts, and output the final time slice allocation scheme.

4. The multi-resource collaborative exhibition hall navigation control system based on time axis arrangement according to claim 3, characterized in that, The dynamic trigger detects that the timeline has reached the resource trigger point. At that time, a trigger determination function that incorporates user intent is used to determine whether to execute the resource instruction: ； in, This is the current timeline position. To trigger the tolerance time window, the value ranges from 50ms to 500ms. This is the current state of the resource. The threshold for triggering user intent is set, with a value ranging from 0.3 to 0.7; if If so, the corresponding resource control command will be executed immediately; if it is an enhancement layer or interaction layer resource, and If so, the triggering of the resource will be delayed or skipped; The execution process includes the following steps: S31. Real-time monitoring of the current position of the timeline. Compare the trigger points of each resource track ; S32, when When this occurs, a pre-check process is triggered to obtain the current status of the corresponding resource. and the user's intent weight for the resource ; S33, will and Substitute into the trigger determination function for calculation ; S34, if Generate resource control instructions and send them to the corresponding execution unit, while recording the trigger time and resource identifier; S5, if Determine the resource type: S351. If it is a basic layer resource, output a trigger exception prompt to the status monitoring and feedback module, and still execute the basic control commands; S352. If it is an enhancement layer resource, delay the trigger point to , The delay duration is 1 to 3 seconds. After the delay, S31 to S33 are re-executed. S353: If it is an interactive layer resource, skip the trigger point directly and record the reason for skipping and the user intent data. S36. Receive instruction execution feedback from the execution unit, confirm the execution result, and synchronize it to the status monitoring and feedback module.

5. A multi-resource collaborative exhibition hall navigation control system based on timeline arrangement according to claim 4, characterized in that, The precise jump controller receives the target jump time. Then, the state recovery time of each resource track is calculated using a jump compensation algorithm that integrates the hierarchical resource states: ； in, For the first Recovery time of each resource track after a jump This is the length of the pending instruction queue for this resource track at the jump point. The system load rate for this resource type. This represents the hierarchical coefficient of the resource orbit. This is the compensation coefficient, with a value ranging from 0.5 to 2.

0. This is the hierarchical weight coefficient, with a value ranging from 0.1 to 0.5, used to adjust the smoothness of jumps between different levels of resources; The jump execution and state synchronization process includes the following steps: S41. Receive the target jump time input by the user. ,verify Is it within the total guided tour duration? If the range is exceeded, an invalid message will be output. S42, if Effective. Pause execution of all current resource tracks and record the current length of the pending instruction queue for each track. and system load rate ; S43, will , and corresponding level coefficients Substitute the jump compensation algorithm to calculate the jump compensation for each track. ; S44. Quickly jump the timeline to... The system locates the position and simultaneously sends status reset commands to each resource orbit, requiring each orbit to reset its position. Internal recovery to The corresponding preset state; S45. Real-time monitoring of the recovery progress of each track status. If a certain track is in... If the recovery process is not yet complete, adjust the compensation coefficient for that orbit. Increase by 10%, recalculate And extend the recovery time; S46. Once all tracks have completed their status recovery, restart the timeline operation and send the jump completion information and the recovery status of each track back to the status monitoring and feedback module.

6. A multi-resource collaborative exhibition hall navigation control system based on timeline arrangement according to claim 5, characterized in that, The status monitoring and feedback module detects the consistency of multi-track resource execution through a status consistency check function and outputs collaborative quality indicators: ； in, As a system synchronization consistency indicator, This represents the number of currently active resource orbitals. For the first The actual execution time of each resource For the expected execution time, The synchronization error tolerance threshold is set between 20ms and 200ms. Score the quality of resource execution. This is the quality threshold, with a value ranging from 0.7 to 0.9; if , The consistency threshold is set between 0.8 and 0.

95. The system automatically triggers the synchronization repair process and transmits the abnormal data to the closed-loop collaborative optimization module. Status monitoring and anomaly handling include the following steps: S51, Real-time collection of currently active data. Actual execution time of each resource orbit And execution quality data, calculate each track ; S52. Obtain the expected execution time for each track. Substitute into the consistency test function to calculate and ; S53, will With consistency threshold In comparison, the system connection status and user interaction status are monitored simultaneously. S54, if Furthermore, the system connection is normal, there are no abnormal user interactions, a normal status report is generated and displayed visually; S55, if There may be system connection abnormalities or user interaction abnormalities, triggering the synchronization repair process: S551, locate the abnormal track and the abnormality type; S552, perform time calibration or command retransmission on the abnormal track; S553, re-acquire the abnormal track. and Calculate the repaired S554, if If the condition is met, the repair is complete, and a repair log is recorded. If the condition is still not met, the abnormal data is packaged and sent to the closed-loop collaborative optimization module to wait for optimization parameters. S6. Real-time updates to the visual status feedback interface.

7. A multi-resource collaborative exhibition hall navigation control system based on time axis arrangement according to claim 6, characterized in that, The multi-resource collaborative exhibition hall navigation control system based on timeline orchestration supports rapid scene switching. It dynamically loads different timeline orchestration schemes through a scene configuration matrix and integrates user intent vectors to achieve adaptive scene adjustment. ； in, Indicates the first In the first scenario The activation status of each resource track, 1 for active and 0 for disabled. This is a user intent vector, with dimensions matching the number of resource tracks, and element values... , For Hadama accumulation, Configure the scene matrix for adaptive adjustment; the system indexes the current scene. Automatically switch to the corresponding configuration based on the user intent vector; The scene switching process includes the following steps: S61. Receive scene switching command and obtain target scene index. ; S62. Load the original scene configuration matrix corresponding to the target scene from the system scene library. ; S63. Obtain the user intent vector output by the user intent awareness module. ; S64, Calculation and The Hadamard product is used to obtain the adaptively adjusted product. Tracks with element values ​​less than 0.3 are marked as to be disabled; S65, Analysis Determine the list of resource tracks to be activated and the corresponding timeline arrangement scheme; S66. Pause the execution of all resource tracks in the current scenario, and save the current timeline position and the status of each track; S67. Load the timeline arrangement scheme for the target scene and activate it. The resource tracks marked as active are disabled, and the tracks to be disabled are disabled. S68. Position the timeline to the initial time point of the target scene or the currently saved timeline position, and start running the timeline; S69. Monitor the initial execution status of each track in the target scenario. If there are tracks that fail to activate, re-execute S67 to S68. S610. After the switch is completed, a message indicating successful scene switching is displayed on the visual interface, and a switch log is recorded.

8. A multi-resource collaborative exhibition hall navigation control system based on time axis arrangement according to claim 7, characterized in that, The hierarchical resource track manager uses a resource dependency graph. Managing dependencies between tracks, where vertices Divided by level , , ,side This indicates time dependency or logical dependency; during resource updates, the system automatically checks for intra-layer and inter-layer conflicts based on the dependency graph and adjusts the resource time interval through a conflict resolution function. ； in, The adjusted resource time range The original time interval, This is the conflict adjustment coefficient, with a value ranging from 0.2 to 0.

8. , Resource orbits With dependent orbit Trigger time; Resource dependency management and conflict resolving include the following steps: S71. Constructing a resource dependency graph Clearly define each vertex Hierarchical affiliation and boundaries Corresponding dependency type; S72. When a resource track needs to be updated, obtain the original time interval of that track. and trigger time ; S73, Based on Dependency Graph Find all orbitals that are dependent on this orbital. Get its trigger time ; S74. Calculate the relationship between this orbit and each dependent orbit. Trigger time difference Check for any overlapping or conflicting time intervals; S75. If there is no conflict, save the updated resource track information directly; S76. If a conflict exists, substitute the conflict resolution function to calculate the adjusted time interval. ; S77, according to Update the time interval and trigger time of this orbit. ; S78. After verification and update, this track is compatible with all dependent tracks. Do conflicts still exist? If so, adjust accordingly. The value is increased or decreased by 0.1, and S76 to S77 are repeated until there is no conflict. S79. Update the relevant information of this track in the resource dependency graph and synchronize it to the timeline synchronization engine and dynamic triggers.

9. A multi-resource collaborative exhibition hall navigation control system based on timeline arrangement according to claim 8, characterized in that, The user intent perception module generates a user intent vector through a user behavior feature extraction and intent classification model. The specific calculations are as follows: ； in, For user behavior feature vectors, The feature weight matrix, For bias vectors, For the activation function, ensure Furthermore, the sum of the user intent weights for all resource tracks is 1; The intent vector generation process includes the following steps: S81. Real-time collection of user behavior data through sensors deployed in the exhibition hall, including dwell time, gaze direction and focus time, number and type of interactive device operations; S82. Preprocess the collected behavioral data: S821. Normalize the dwell time to the [0,1] interval to obtain the normalized dwell time value; S822. Calculate the proportion of gaze focus time to the total dwell time to obtain the gaze focus time proportion; S823. Count the number of interactive operations per unit time, and use the standardized value as the interactive operation feature value. S83. Combine the preprocessed feature values ​​to construct a user behavior feature vector. ; S84, Load the pre-trained feature weight matrix and bias vector ,Will Substitute the values ​​into the calculation formula to obtain the original intent weights; S85, Perform weighting on the original intent. Activation operation to obtain the corresponding resource tracks To form the user intent vector ; S86, Verification Check if the sum of all elements is 1. If the deviation exceeds 0.01, re-execute S85 until the condition is met. S87, will generate Output to the hierarchical resource track manager, dynamic triggers, and scene switching module.

10. A multi-resource collaborative exhibition hall navigation control system based on time axis arrangement according to claim 9, characterized in that, The closed-loop collaborative optimization module dynamically updates the resource scheduling weight coefficient and time axis calibration parameters using the following formula to achieve end-to-end collaborative optimization: ； ； in, This is the updated weight adjustment coefficient. This is the historical average user intent vector. To optimize the step size, a value of 0.1 to 0.3 is used. The updated time synchronization threshold. This is the threshold adjustment coefficient, with a value ranging from 0.2 to 0.5; the system automatically performs an optimization update every 100 to 500 collaborative data entries. The parameter update process includes the following steps: S91. Real-time collection of collaborative data during system operation, including resource scheduling weight coefficients, time synchronization thresholds, , Resource execution status data; S92. Count the number of accumulated collaborative data entries and determine whether the update condition of 100-500 entries has been met; S93. If the target is not met, continue accumulating data; if the target is met, extract the historical average user intent vector. and the current , , ; S94. Substitute the above parameters into the weight adjustment coefficient update formula to calculate... Ensure the sum of the updated coefficients is still 1; if not, adjust proportionally. S5. Substitute into the time synchronization threshold update formula to calculate... ,make sure Within the range of 10ms to 100ms; S96. Send the updated weight adjustment coefficient to the hierarchical resource track manager and the updated time synchronization threshold to the timeline synchronization engine. S97, Verify the updated system Whether to upgrade; if so, save the updated parameters as the new baseline parameters; if not, discard the update results and retain the original parameters. S98. Clear the accumulated collaborative data and restart the data accumulation process.

Images