A virtual reality-based digital cloud showroom interactive experience method for a vehicle

Abstract

The application discloses a kind of based on virtual reality's car digital cloud exhibition hall interactive experience method, it is related to virtual reality three-dimensional graphics processing technical field, including, constructing interactive vehicle digital body, solidification movable component movement description record and interactive semantic annotation, generate interaction dependent sequence table;Based on interaction dependent sequence table, user experience intention is obtained and is converted into task list, necessary element set is generated for each task, the minimum necessary interaction field is selected from necessary element set, and interaction script is generated;According to interaction script, minimum necessary interaction field is loaded to enter virtual reality, and action is executed according to task list order to form candidate vehicle state, consistency deviation is calculated and candidate vehicle state with smaller consistency deviation is selected, when script determines that task is completed, the evidence bound with task type is output;Through evidence, interaction script, action and vehicle digital body solidification reproducible experience record, and according to reproducible experience record, script is re-arranged to task order.

Description

Technical Field

[0001] This invention relates to the field of virtual reality 3D graphics processing technology, and in particular to an interactive experience method for a car digital cloud showroom based on virtual reality. Background Technology

[0002] With the maturity of 3D graphics, real-time rendering, cloud streaming, and head-mounted display devices, virtual reality is widely used for immersive display and remote interaction in automotive digital cloud showrooms. Existing technologies typically use high-precision 3D vehicle assets as the core, combined with a rendering pipeline that ensures consistency in materials, lighting, and physics, to reconstruct the vehicle's exterior and cabin details in virtual space. Through network distribution and multi-terminal adaptation, the consistency of the presentation effect is guaranteed. Users can freely browse, switch perspectives, access hotspot explanations, and interact with components in the showroom. Combined with voice narration, text and image information, and guided tour scripts, a controllable display rhythm can be achieved. It also supports linkage with product information such as parameter configuration and version comparison. This type of solution can break through time and geographical limitations, improve the efficiency and immersion of vehicle dissemination, and facilitate the recording and analysis of interactive data, supporting content iteration, personalized recommendations, and multi-scenario reuse. It has significant application value in digital marketing and online exhibitions.

[0003] Conventional solutions often lack explicit constraints on pre-action dependencies and component state semantics in complex interactions, making it difficult to reproduce the interaction sequence stably across different playbacks and terminals. At the same time, interaction scenarios are often loaded as a whole or at a fixed granularity, making it difficult to automatically trim them to the minimum necessary scope for the task, which can easily lead to uncertainty in loading overhead and user experience response. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a virtual reality-based interactive experience method for automotive digital cloud showrooms to solve the problems of unstable experience reproduction caused by the difficulty in explicitly solidifying the dependence of interactive actions and the semantics of component states, as well as the uncertainty of loading and response caused by the difficulty in tailoring interactive scenes to the minimum necessary range according to tasks.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] This invention provides a virtual reality-based interactive experience method for automotive digital cloud showrooms, which includes collecting 3D data of vehicle models, material textures, and component assembly relationships and unifying them into a vehicle coordinate system, constructing an interactive digital vehicle body, fixing motion description records of movable parts and interactive semantic annotations, and generating an interactive dependency sequence list.

[0008] The user experience intent is obtained based on the interaction dependency sequence list and converted into a task list. A set of necessary elements is generated for each task. The minimum necessary interaction field is selected from the set of necessary elements to generate the interaction script.

[0009] The minimum necessary interaction field is loaded into the virtual reality according to the interaction script. Actions are executed in the order of the task list to form candidate vehicle states. The consistency deviation is calculated and the candidate vehicle state with the smaller consistency deviation is selected. When the script determines that the task is completed, evidence bound to the task type is output.

[0010] Reproducible experience records are solidified through evidence, interaction scripts, actions, and vehicle digital entities, and the task sequence is re-scripted based on these reproducible experience records.

[0011] As a preferred embodiment of the interactive experience method for automotive digital cloud showroom based on virtual reality described in this invention, the specific steps of collecting vehicle model 3D data, material textures, and component assembly relationships and unifying them into the vehicle coordinate system to construct an interactive vehicle digital body are as follows: reading the assembly model from the vehicle model design data source, extracting 3D geometry, material textures, material parameters, and assembly relationships, solidifying component identifiers and component assembly hierarchical paths to form an input set, unifying the input set into the vehicle coordinate system, and performing rigid body transformation to obtain the overall vehicle geometric datum.

[0012] An interactive digital vehicle body is constructed on the vehicle's geometric datum. A mapping table of component identifiers, assembly level paths, geometric references, and material references is written with components as the smallest objects. The operable components are selected to form a list of movable components, and the spatial range description of the operable positions is solidified.

[0013] As a preferred embodiment of the virtual reality-based automotive digital cloud showroom interactive experience method of the present invention, the steps of solidifying the motion description records and interactive semantic annotations of movable parts and generating an interactive dependency sequence list are as follows: solidifying the motion description records of movable parts one by one and writing them into the interactive vehicle digital body; solidifying the interactive semantic annotations for each movable part and forming an interactive semantic annotation set; extracting actions from the interactive semantic annotation set to form an action list and solidifying an action identifier for each action.

[0014] The list of component states that have been changed for the action and the list of component states required for execution are fixed, and the dependency relationship records are matched and summarized into the relationship part of the interaction dependency sequence table. Starting from no preceding action, the execution order is iteratively written according to the action identifier string order to obtain the order part of the interaction dependency sequence table.

[0015] An interactive dependency sequence list is generated by combining the relational and sequential parts of the interactive dependency sequence list.

[0016] As a preferred embodiment of the interactive experience method for automotive digital cloud showrooms based on virtual reality described in this invention, the steps of obtaining user experience intent based on the interaction dependency sequence list and converting it into a task list, and generating a set of necessary elements for each task, are as follows: collecting user experience intent; guiding the user to locate the action identifier in the action list and determine the target action of the user experience intent according to the list of movable parts; tracing the preceding actions based on the interaction dependency sequence list and sorting them to generate an executable action sequence; solidifying each target action into a task item; and merging and deduplicating to obtain the task list order.

[0017] For each task item in the task list, read the action identifier in the sequence of executable actions for each task item, locate the component identifier based on the action identifier, extract the necessary information of the component from the interactive vehicle digital body, and form the necessary element set of the task item.

[0018] As a preferred embodiment of the interactive experience method for automotive digital cloud showroom based on virtual reality described in this invention, the step of selecting the minimum necessary interaction field from the set of necessary elements and generating an interaction script specifically involves: combining the set of necessary elements of all task items into a basic candidate interaction field; and generating multiple sets of candidate interaction fields by backtracking the parent node for each task item according to the assembly level path.

[0019] By determining whether the candidate interaction field covers all task items, the cumulative loading cost is calculated based on geometric references and material references. For each candidate interaction field, the proportion of the intersection between the necessary elements of the task item and the elements of the candidate interaction field is used as the cumulative coverage. The ratio of the cumulative loading cost to the cumulative coverage is used as the interaction field evaluation value. The candidate interaction field with the smallest interaction field evaluation value is selected as the minimum necessary interaction field.

[0020] Generate an interactive script in the order of the task list, consisting of a list of minimum necessary interactive field elements, a sequence of action prompts, and task completion criteria.

[0021] As a preferred embodiment of the interactive experience method for a car digital cloud showroom based on virtual reality described in this invention, the steps of loading the minimum necessary interactive field according to the interactive script to enter the virtual reality and executing actions in the order of the task list to form a candidate vehicle state are as follows: reading the minimum necessary interactive field element list in the interactive script, locating the geometric reference and material reference of each component in the element list one by one, loading the three-dimensional geometry, material texture and material parameters into a renderable appearance, placing the components according to the vehicle coordinate system and the overall vehicle geometric reference, binding the spatial range description of the operable position of the movable component and loading the motion description record and interactive semantic annotation of the movable component before entering the virtual reality.

[0022] The action prompt sequence is executed in the order of the task list. After the user triggers the operation within the spatial range of the operable location, the action identifier locates the component and updates its pose. The set of poses of each component in the vehicle coordinate system at the end of this action execution is recorded as the candidate vehicle state.

[0023] As a preferred embodiment of the virtual reality-based automotive digital cloud showroom interactive experience method of the present invention, the specific steps of calculating the consistency deviation and selecting the candidate vehicle state with the smaller consistency deviation are as follows: calculating the consistency deviation for multiple candidate vehicle states of the same action, and obtaining the consistency deviation by combining the interference penetration depth and the overtravel amount and normalizing the integral.

[0024] The candidate vehicle state with the smaller consistency deviation is selected as the retained vehicle state.

[0025] As a preferred embodiment of the interactive experience method for automotive digital cloud showroom based on virtual reality described in this invention, the step of outputting evidence bound to the task type when the script determines that the task is completed is as follows: after the action prompt sequence of the task item is executed, the observable result definition is checked according to the completion criterion. If it is true, the task is determined to be completed and evidence is output according to the result category. If it is not true, the last action prompt of the untrue observable result definition in the task item is returned and the best one is selected again.

[0026] Once the task list is complete, the evidence output by each task item will be summarized in the order of the task list and a correspondence will be established with the task items and action prompt sequences in the interactive script.

[0027] As a preferred embodiment of the interactive experience method for a virtual reality-based automotive digital cloud showroom described in this invention, the step of solidifying a reproducible experience record through evidence, interaction scripts, actions, and vehicle digital bodies includes: writing the basic part of the reproducible experience record in the order of the task list of the interaction script, and then writing the process part of the reproducible experience record in the order of the action prompt sequence of the task items.

[0028] As a preferred embodiment of the interactive experience method for automotive digital cloud showroom based on virtual reality described in this invention, the step of re-arranging the task order according to the reproducible experience record involves extracting the completion marker, the number of times the action is repeatedly triggered, and the task duration from the reproducible experience record, and determining the order of the task items according to the priority of incomplete tasks, the priority of more times the action is repeatedly triggered, and the priority of shorter task duration, thereby obtaining the target arrangement of the task items.

[0029] The task list order is generated by using the interaction dependency order list as a constraint. The action identifier set corresponding to the task list order is maintained. All task entries whose action prerequisites are included in the action identifier set are filtered in a loop and written according to the target. If there are no task entries to write, the remaining task entries are written in the order of the interaction dependency order list. The task list order in the interaction script is replaced with the task list order of the rearranged task list, and the rearranged interaction script is output.

[0030] The beneficial effects of this invention are as follows: by constructing an interactive vehicle digital body from the vehicle design data source and solidifying the interactive semantic annotation and interactive dependency sequence table, the traceable association between actions and component states is realized, which is used to generate task lists and task item completion criteria, so that the interaction sequence can be reproduced and the results can be verified. By constructing candidate interaction fields and selecting the minimum necessary interaction fields to generate interaction scripts, the virtual reality can be entered according to the necessary elements of the task, which reduces the fluctuation of the loading range, achieves stable experience response and facilitates script re-arrangement. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a flowchart of a virtual reality-based interactive experience method for automotive digital cloud showrooms.

[0033] Figure 2 This is a heatmap comparing playback success rates based on reproducible experience records.

[0034] Figure 3 A flowchart for generating an interaction dependency sequence list.

[0035] Figure 4 A flowchart for calculating consistency deviation and selecting candidate vehicle states. Detailed Implementation

[0036] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0037] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0038] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0039] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a method for interactive experience in a virtual reality-based automotive digital cloud showroom, comprising the following steps:

[0040] S1. Collect vehicle model 3D data, material textures, and component assembly relationships and unify them into the vehicle coordinate system to construct an interactive vehicle digital body, solidify the motion description records of movable parts and interactive semantic annotations, and generate an interaction dependency sequence list.

[0041] The assembly model of the vehicle is read from the vehicle design data source, and 3D geometric data including the exterior, interior, glass, tires, lights, seats, door hood, trunk lid, steering wheel and center console components are exported. The material textures and material parameters corresponding to the components are read to give each component a renderable appearance. The assembly relationship data is read, including the parent-child assembly hierarchy and the reference relationship between the assembly datum plane and the assembly constraint. The component identifier and the component assembly hierarchy path are written into the list to obtain the input set used for unifying coordinates.

[0042] Vehicle design data source refers to the original engineering data carrier generated and saved during the vehicle design phase.

[0043] The component assembly hierarchy path refers to the hierarchy positioning string in the vehicle assembly model, which describes how any component is located from the root node of the whole vehicle along the parent and child assembly hierarchy levels to the component level.

[0044] All component geometry within the input set is unified to the same vehicle coordinate system. The vehicle coordinate system adopts the semantic definition of a traveling vehicle body, with the forward direction as the longitudinal direction, the left direction as the lateral direction, and the upward direction as the vertical direction. The origin is taken as the geometric center point of the vehicle's outer envelope projected onto the ground. The length unit of all geometry is unified to meters, and rigid body transformation is performed on each component to ensure that all components at the assembly level maintain their relative positions in the same datum, thus obtaining the vehicle's geometric datum in the vehicle coordinate system.

[0045] An interactive digital vehicle body is constructed on the vehicle's geometric datum. A mapping table of component identifiers, assembly level paths, geometric references, and material references is written with components as the smallest objects. Components are interactively filtered in the vehicle coordinate system, and all components that can be operated by the user, such as opening, sliding, rotating, pressing, and flicking, are included in the list of movable components. The list of movable components includes doors, trunk lids, seat rails and backrests, steering wheel adjustment mechanisms, window lifts, sun visors, and central control operable components. A unique component identifier, the assembly level path to which it belongs, and a spatial description of the operable position are assigned to each movable component, and the list of movable components is output.

[0046] The description of the spatial range of the operable position refers to the detailed definition of the spatial area, reachable direction range, and interaction restrictions that allow users to operate the movable parts during virtual interaction. The spatial area is represented by a closed space volume in the vehicle coordinate system and is bound to the part identifier. The reachable direction range is represented by a set of several allowed approach directions. The interaction restrictions include the allowed hand contact range and the description of restrictions that prohibit penetration.

[0047] For each movable part in the list of movable parts, a motion description record is fixed and written into the interactive vehicle digital body. The motion description record includes the motion type (rotation or sliding), motion axis or sliding direction, axis passing through points, starting and ending poses, limiting method and linkage relationship with the movable part. The linkage relationship is fixed by the way that the change of the pose of the front part causes the change of the effective space or obstruction boundary of the rear part (for example, there is an executable constraint association between the seat moving forward, the backrest folding down and the change of the trunk loading space). After the fixed is completed, a set of movable part motion description records is obtained.

[0048] Based on the list of movable parts, an interactive semantic annotation is fixed for each movable part. The semantic annotation consists of a set of executable actions of the part, display information entries and observable results. After the semantic annotation is completed, an interactive semantic annotation set is obtained.

[0049] The set of executable actions for a component is used to limit the types of operations that a user can perform on a movable component in virtual reality.

[0050] Display information entries to present functional and parameter descriptions related to the component when an action is triggered.

[0051] The definition of observable results is used to specify the verifiable results that should be produced after an action is completed (e.g., whether it is in place, whether it can be closed, whether it is reached, and whether the occlusion has changed).

[0052] All actions are extracted from the interaction semantic annotation set to form an action list. An action identifier is assigned to each action in the action list. The action identifier is obtained by concatenating the component identifier corresponding to the action with the operation type. A list of changed component states and a list of component states required for execution are assigned to each action in the action list. The list of changed component states of the preceding candidate action is compared with the list of component states required for execution of the subsequent candidate action. The number of identical component state entries in the list of changed component states and the list of component states required for execution is taken as the number of matching entries. The ratio of the number of matching entries to the total number of component state entries required for execution of the subsequent candidate action is taken as the action dependency of the preceding candidate action on the subsequent candidate action. If the list of component state required for execution of the subsequent candidate action is empty, the action dependency is recorded as zero. When the action dependency is non-zero, a dependency relationship record is generated. All dependency relationship records are summarized in field order to obtain the relationship part of the interaction dependency sequence table.

[0053] A component status entry is a comparable record consisting of a component identifier, a status name, and a status value description. The status name indicates the status type, and the status value description indicates the range or position of the status value in the execution decision.

[0054] Actions without any preceding actions are filtered to form a set of currently executable actions. Initially, all actions in the set have no preceding action dependencies and can be executed first. When there are multiple actions in the set, their order is determined by the string order of their action identifiers and written into the execution order. Actions already written into the execution order are removed from the set. The process continues until all actions are executed according to their dependencies, resulting in the order part of the interaction dependency order table. When multiple actions enter the set in the same round, their order is still determined by the string order of their action identifiers. The interaction dependency order table consists of a relational part and an order part.

[0055] S2. Obtain user experience intent based on the interaction dependency sequence list and convert it into a task list. Generate a set of necessary elements for each task, select the minimum necessary interaction field from the set of necessary elements, and generate an interaction script.

[0056] The user experience intent is collected sequentially. First, the component identifiers and component assembly hierarchy paths in the list of movable components are presented. After the user selects any component, only the set of executable actions corresponding to the component in the interaction semantic annotation set is presented. The user then selects an operation type, and the selection is written into the record of the component identifier and operation type. The corresponding action identifier is immediately located in the action list as the target action corresponding to this experience intent. This process is repeated until the collection ends, resulting in a set of experience intents consisting of multiple target actions.

[0057] For each target action in the experience intent set, the preceding actions of the target action are traced step by step based on the relational part of the interaction dependency sequence table. Starting from the target action, all dependency records pointing to the target action are searched, and the preceding actions in the dependency records are added to the unprocessed set. The tracing is repeated for each preceding action in the unprocessed set until no more preceding actions are added, resulting in the complete action set corresponding to the target action. The complete action set is sorted by the order part of the interaction dependency sequence table to obtain an executable action sequence consistent with the dependency relationship. One executable action sequence corresponding to a target action is solidified into a task item. Multiple task items are then merged according to the order part of the interaction dependency sequence table. If multiple task items contain the same action, only the position of the action's first appearance in the overall sequence is retained and subsequent duplicates are deleted, so that all actions appear only once in the task list, resulting in a definite task list and task order.

[0058] For each task item in the task list, the action identifier in the executable action sequence of the task item is read one by one, and the corresponding component identifier is located according to the action identifier. The necessary information of the component is extracted from the mapping table of interactive vehicle digital bodies, the motion description record of movable parts, the interactive semantic annotation set, the movable part list, the action list and the interaction dependency sequence table to form the necessary element set of the task item. The necessary element set includes the component identifier, component assembly level path, geometric reference, material reference, spatial range description of operable position, motion description record of movable parts, display information items in interactive semantic annotation and observable result definition.

[0059] Candidate interaction fields are constructed, and the minimum necessary interaction field is selected. The basic candidate interaction field is constructed as the union of the sets of necessary elements of all task items. For each task item, according to the component assembly hierarchy path, multiple sets of candidate interaction fields are generated by progressively backtracking from containing all parent node components to containing only the nearest parent node component. For each candidate interaction field, it is determined whether it covers all task items by the correspondence between the task item and the set of necessary elements of the task item. The interaction field evaluation value is calculated, and the minimum interaction field evaluation value is selected as the minimum necessary interaction field. The expression is:

[0060] ;

[0061] ;

[0062] in, Indicates the evaluation score of the interaction field. This indicates the cumulative loading cost. Represents the dummy variable for integration. Indicates the number of task entries. Indicates the first The set of required elements for each task entry The set of required elements representing the candidate interaction field. Indicates the task entry index. This represents the total number of triangular faces corresponding to geometric references in the candidate interaction field. This represents the total number of triangular faces corresponding to the overall vehicle geometric datum. This indicates the total number of bytes occupied by material maps and material parameters in the candidate interaction field. This indicates the total number of bytes occupied by the vehicle's texture maps and material parameters.

[0063] After the minimum necessary interaction field is determined, the interaction script is generated according to the order of the tasks in the task list. The interaction script consists of a list of minimum necessary interaction field elements, a sequence of action prompts corresponding to the task items, and the task item completion criteria.

[0064] The minimum necessary interaction field element list refers to the list of elements in the minimum necessary interaction field that cover all task items.

[0065] The action prompt sequence corresponding to a task entry refers to the prompt order generated according to the executable action sequence of the task entry determined by the interaction dependency order table, and the spatial range description of binding action identifiers to each action and pointing to the operable position of the corresponding component.

[0066] The task completion criterion refers to the definition of observable results corresponding to the actions related to the task item, taken from the interaction semantic annotation set.

[0067] S3. Load the minimum necessary interaction field into virtual reality according to the interaction script, execute actions in the order of the task list to form candidate vehicle states, calculate the consistency deviation and select the candidate vehicle state with the smaller consistency deviation, and output evidence bound to the task type when the script determines that the task is completed.

[0068] The system reads the minimum necessary interactive field element list from the interaction script, locates the geometric and material references of each component in the element list, loads the corresponding 3D geometric data, material maps, and material parameters into a renderable appearance, and places the components according to the vehicle coordinate system and the overall vehicle geometric reference to keep the relative positions of each component unchanged at the assembly level. For each movable component in the element list, the system loads the spatial range description of the movable component's operable position, binds the spatial range description of the operable position to the component identifier of the movable component, and then loads the corresponding movable component motion description record and interaction semantic annotation. After loading is completed, the system enters the virtual reality environment and uses the first task entry in the interaction script as the current task entry.

[0069] The task list in the interaction script is read sequentially, and the current task item is executed one by one according to the action prompt sequence. Each action prompt is bound to an action identifier and points to the spatial range description of the operable position of the corresponding component. After the user triggers the operation within the spatial range description of the operable position, the corresponding component identifier is located according to the action identifier. The motion description record of the movable component is read, and the user operation is converted into the rotation or sliding of the component according to the motion type in the motion description record. The component pose is constrained by the limit method so that it does not exceed the range of the start pose and the end pose. The set of poses of each component in the vehicle coordinate system at the end of this action is recorded as a candidate vehicle state. When the user triggers the same action repeatedly, the mapping and constraint process is repeated to form multiple candidate vehicle states.

[0070] The pose set refers to the pose set of the movable parts corresponding to the current action and the pose set of the subsequent parts that have a linkage relationship described in the motion description record of the movable parts.

[0071] For multiple candidate vehicle states generated by the same action, the consistency deviation is calculated for each state, and the best one is retained. For each candidate vehicle state, during the process from triggering to ending the action, the penetration depth or unreasonable contact depth between components is calculated based on geometric references. Furthermore, the out-of-bounds travel amount of the movable component's pose relative to the initial and final poses is calculated based on the movable component's motion description record. The expression is as follows:

[0072] ;

[0073] in, Indicates the amount of consistency deviation. This indicates the duration of the action from its triggering to its termination. Represents a time variable. This indicates the number of component pairs involved in the interference check. Indicates the index of the component pair. Indicates time No. The penetration depth or unreasonable contact depth of individual components. Indicates the first Each movable component corresponds to the upper bound of the reference depth used for normalization. Indicates the number of movable parts involved in the travel inspection. Indicates the index of the movable part. Indicates time No. The overtravel range of a movable part Indicates the first One movable part is used for the normalized reference travel upper limit.

[0074] For each component pair involved in the interference check, it is recommended to determine the upper limit of the reference depth according to priority. If the vehicle design data source provides the mating distance, stop distance, assembly gap, or equivalent geometric constraint distance of the component pair in the assembly constraints, then the distance is converted to the length unit of the vehicle coordinate system and used as the upper limit of the reference depth. If the assembly constraints do not provide an available distance, calculate the local thickness of the two components along the normal direction through the solid near the contact point and contact normal obtained by the interference detection (the distance between the entry point and the exit point can be obtained by ray intersection), and take the smaller value of the local thickness of the two components as the upper limit of the reference depth of the component pair.

[0075] The upper limit of the reference stroke is taken from the initial and final poses, motion type, motion axis, or sliding direction in the motion description record of the movable part. If the motion type is sliding, the upper limit of the reference stroke is the displacement of the final pose relative to the initial pose along the sliding direction. If the motion type is rotation, the upper limit of the reference stroke is the angular amplitude of the final pose relative to the initial pose about the motion axis.

[0076] After the action prompt sequence of the current task item is completed, the completion criteria of the current task item in the interaction script is read, and the observable result definitions corresponding to the completion criteria are checked one by one in the vehicle retention state. When the component state item required by the observable result definition is met in the vehicle retention state, the current task item is determined to be completed and evidence is output. The result category pointed to by the observable result definition corresponding to the completion criteria of the current task item is used as the task type, and evidence including task item, action identifier sequence, component identifier, component state item and state value description is output. The result category is taken from whether it is in place, whether it can be closed, whether it is reached, and whether the occlusion has changed, as pointed to by the observable result definition. The evidence can directly correspond to the observable result definitions that have been published in the interaction semantic annotation set. If the completion criteria are not met, the last action prompt corresponding to the unmet observable result definition in the current task item is returned, execution continues and candidate vehicle states are re-formed, consistency deviation is calculated and selected, until the current task item is completed and the next task item is entered.

[0077] Once the task list is complete, the evidence output by each task item is summarized in the order of the task list and a correspondence is established with the task items and action prompt sequences in the interactive script.

[0078] S4. Solidify reproducible experience records through evidence, interaction scripts, actions, and vehicle digital entities, and re-arrange the task sequence according to the reproducible experience records.

[0079] Write the basic part of the reproducible experience record in the order of the task list in the interaction script, and write the mapping table, movable part list, movable part motion description record, interaction semantic annotation set, action list and interaction dependency order table that have been fixed in the vehicle digital body. Also write the minimum necessary interaction field element list, task items, action prompt sequence corresponding to task items and task item completion criteria in the interaction script.

[0080] For each task entry, the process of reproducible experience record is written in the order of the task entry action prompt sequence, the action identifier sequence formed by the actual trigger of the task entry is written, for each trigger in the action identifier sequence, the pose set of candidate vehicle states formed by each trigger and the consistency deviation amount are written, and the pose set of candidate vehicle states that are retained after selection is written. When the task entry completion criterion is met, the evidence output by the task entry is written.

[0081] Figure 2 To compare the playback success rate based on reproducible experience records, the left subgraph A (closed) represents the playback result when the interaction semantic annotation set, interaction dependency order list, task completion criteria, evidence, and reproducible experience records are not formed into a complete closed loop (i.e., the traceability and comparison state with the incomplete reproducibility mechanism). The right subgraph A (open) represents the playback and verification result based on reproducible experience records after the closed loop is fully enabled according to this invention. The horizontal axis represents the terminal level (low, medium, high), and the vertical axis represents the network level (good network, normal network, poor network). The value of each heatmap cell represents the completion of task list playback under the corresponding network and terminal combination conditions. When A is closed, a closed-loop solidification of the interaction semantic annotation set, interaction dependency sequence list, and task item completion criteria is not formed. The replay success rate fluctuates significantly with network jitter and terminal computing power decline. When A is open, the action prompt sequence corresponding to the task item is solidified according to the interaction dependency sequence list, and the task item completion criteria are defined and generated based on the observable results in the interaction semantic annotation set. During the replay process, the action identifier sequence is reproduced through reproducible experience records, and the completion criteria are checked one by one while retaining the vehicle status and corresponding evidence is output. This improves the overall replay success rate under various network and terminal combinations and reduces fluctuations, verifying that the present invention achieves reproducible interaction sequence and verifiable results.

[0082] From the reproducible experience records, for each task item, the number of times the same action identifier appears repeatedly in the action identifier sequence of the task item is extracted as the action repetition trigger count. The duration from the first action trigger to the completion criterion being met is the task duration. Whether the task item's completion criterion is met is the completion mark (completed if the completion criterion is met, incomplete if not). Without changing the task item action prompt sequence and task item completion criterion, the task item sorting priority is as follows: incomplete task items are placed before completed task items. Among task items with the same completion mark, those with more action repetition trigger counts are placed first. Among task items with the same action repetition trigger counts, those with shorter task durations are placed first. If they are still the same, the order is determined by the string order of the first action identifier in the executable action sequence of the task item. All task items are compared and sorted sequentially according to the sorting priority to obtain the target arrangement of task items.

[0083] The target arrangement is modified using an interaction dependency order list as a constraint. The set of action identifiers corresponding to the order of the newly written task list is maintained. From the task items in the task list that have not yet been written, all task items whose prerequisite actions are included in the action identifier set are selected as the current writable set. When the current writable set is empty, it is determined that the remaining task items cannot be satisfied by the target arrangement under the interaction dependency order list constraint. Therefore, writing all unwritten task items according to the target arrangement order is stopped, and instead, the remaining task items are written according to the order determined by the order part of the interaction dependency order list. When there are multiple task items in the current writable set, they are selected for writing according to the target arrangement order. If the target arrangement order is still tied, the order is determined by the string order of the first action identifier and written accordingly. After writing, the action identifier set is updated, and the selection and writing are repeated until all task items are written. After completion, the task list order in the interaction script is replaced with the rearranged task list order, keeping the minimum necessary interaction field element list, the action prompt sequence corresponding to the task item, and the task item completion criteria unchanged. The rearranged interaction script is then output.

[0084] In summary, this invention achieves traceable association between actions and component states by constructing an interactive digital vehicle body from vehicle design data sources and solidifying interactive semantic annotations and interaction dependency sequence lists. This is used to generate task lists and task completion criteria, ensuring that the interaction sequence is reproducible and the results are verifiable. By constructing candidate interaction fields and selecting the minimum necessary interaction field, an interaction script is generated, enabling entry into virtual reality based on the necessary elements of the task. This reduces loading range fluctuations, achieves stable experience response, and facilitates script re-arrangement.

[0085] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for interactive experience in a car digital cloud showroom based on virtual reality, characterized in that: include, Collect vehicle 3D data, material textures and component assembly relationships and unify them into the vehicle coordinate system, construct an interactive vehicle digital body, solidify the motion description records of movable parts and interactive semantic annotations, and generate an interaction dependency sequence list; The user experience intent is obtained based on the interaction dependency sequence list and converted into a task list. A set of necessary elements is generated for each task. The minimum necessary interaction field is selected from the set of necessary elements to generate the interaction script. The minimum necessary interaction field is loaded into the virtual reality according to the interaction script. The actions are executed in the order of the task list to form candidate vehicle states. The consistency deviation is calculated and the candidate vehicle state with the smaller consistency deviation is selected. When the script determines that the task is completed, the evidence bound to the task type is output. Reproducible experience records are solidified through evidence, interaction scripts, actions, and vehicle digital entities, and the task sequence is re-scripted based on the reproducible experience records. The steps for selecting the minimum necessary interaction field from the set of necessary elements and generating the interaction script are as follows: combine the set of necessary elements of all task items into the basic candidate interaction field, and backtrack the parent node for each task item according to the assembly hierarchy path to generate multiple sets of candidate interaction fields. By determining whether the candidate interaction field covers all task items, the cumulative loading cost is calculated based on geometric references and material references. For each candidate interaction field, the proportion of the intersection between the necessary elements of the task item and the elements of the candidate interaction field is used as the cumulative coverage. The ratio of the cumulative loading cost to the cumulative coverage is used as the interaction field evaluation value. The candidate interaction field with the smallest interaction field evaluation value is selected as the minimum necessary interaction field. Generate an interactive script in the order of the task list, consisting of a list of minimum necessary interactive field elements, a sequence of action prompts, and task completion criteria.

2. The interactive experience method for a car digital cloud showroom based on virtual reality as described in claim 1, characterized in that: The process involves collecting 3D vehicle data, material textures, and component assembly relationships, unifying them into a vehicle coordinate system, and constructing an interactive digital vehicle entity. The specific steps are as follows: The assembly model is read from the vehicle design data source, and 3D geometry, material maps, material parameters and assembly relationships are extracted. The component identifiers and component assembly hierarchical paths are solidified to form an input set. The input set is unified to the vehicle coordinate system and rigid body transformation is performed to obtain the vehicle geometric datum. An interactive digital vehicle body is constructed on the vehicle's geometric datum. A mapping table of component identifiers, assembly level paths, geometric references, and material references is written with components as the smallest objects. The operable components are selected to form a list of movable components, and the spatial range description of the operable positions is solidified.

3. The interactive experience method for a car digital cloud showroom based on virtual reality as described in claim 2, characterized in that: The fixed movable component motion description record and interaction semantic annotation are used to generate an interaction dependency sequence table. The specific steps are as follows: The motion description records of each movable part in the list of movable parts are solidified and written into the digital body of the interactive vehicle. Interactive semantic annotations are solidified for each movable part to form an interactive semantic annotation set. Actions are extracted from the interactive semantic annotation set to form an action list and an action identifier is solidified for each action. The list of component states that have been changed for the action and the list of component states required for execution are fixed, and the dependency relationship records are matched and summarized into the relationship part of the interaction dependency sequence table. Starting from no preceding action, the execution order is iteratively written according to the action identifier string order to obtain the order part of the interaction dependency sequence table. An interactive dependency sequence list is generated by combining the relational and sequential parts of the interactive dependency sequence list.

4. The interactive experience method for a car digital cloud showroom based on virtual reality as described in claim 3, characterized in that: The steps for obtaining user experience intent based on the interaction dependency sequence list and converting it into a task list, and generating a set of necessary elements for each task, are as follows: Collect user experience intent, guide users to locate action identifiers in the action list according to the list of movable parts and determine the target action of user experience intent, trace the preceding actions based on the interaction dependency sequence list and sort them to generate an executable action sequence, solidify each target action into a task item, merge and remove duplicates to obtain the task list order. For each task item in the task list, read the action identifier in the sequence of executable actions for each task item, locate the component identifier based on the action identifier, extract the necessary information of the component from the interactive vehicle digital body, and form the necessary element set of the task item.

5. The interactive experience method for a car digital cloud showroom based on virtual reality as described in claim 4, characterized in that: The steps for loading the minimum necessary interactive field into virtual reality according to the interactive script, and executing actions in the order of the task list to form candidate vehicle states are as follows: Read the minimum necessary interactive field element list in the interaction script, locate the geometric reference and material reference of each component in the element list one by one, load the 3D geometry, material map and material parameters into a renderable appearance, place the components according to the vehicle coordinate system and the vehicle geometric reference, bind the spatial range description of the operable position of the movable component, load the motion description record of the movable component and the interaction semantic annotation, and then enter the virtual reality. The action prompt sequence is executed in the order of the task list. After the user triggers the operation within the spatial range of the operable location, the action identifier locates the component and updates its pose. The set of poses of each component in the vehicle coordinate system at the end of this action execution is recorded as the candidate vehicle state.

6. The interactive experience method for a car digital cloud showroom based on virtual reality as described in claim 5, characterized in that: The specific steps for calculating the consistency deviation and selecting the candidate vehicle state with the smaller consistency deviation are as follows: For multiple candidate vehicle states with the same action, calculate the consistency deviation, combine the interference penetration depth and the overtravel amount, and normalize the integral to obtain the consistency deviation. The candidate vehicle state with the smaller consistency deviation is selected as the retained vehicle state.

7. The interactive experience method for a car digital cloud showroom based on virtual reality as described in claim 6, characterized in that: When the script determines that the task is complete, it outputs evidence linked to the task type. The specific steps are as follows: After the action prompt sequence of the task item is completed, the observable result definition is checked according to the completion criterion. If it is true, the task is determined to be completed and evidence is output according to the result category. If it is false, the last action prompt of the untrue observable result definition in the task item is returned and the best one is selected again. Once the task list is complete, the evidence output by each task item will be summarized in the order of the task list and a correspondence will be established with the task items and action prompt sequences in the interactive script.

8. The interactive experience method for a car digital cloud showroom based on virtual reality as described in claim 7, characterized in that: The process of solidifying reproducible experience records through evidence, interaction scripts, actions, and vehicle digital entities includes: writing the basic part of the reproducible experience record in the order of the task list of the interaction script, and then writing the process part of the reproducible experience record in the order of the action prompt sequence of the task items.

9. The interactive experience method for a car digital cloud showroom based on virtual reality as described in claim 8, characterized in that: The specific steps for re-scripting the task sequence based on reproducible experience records are as follows: The completion markers, number of repeated action triggers, and task durations of each task item are extracted from the reproducible experience records. The order of tasks is determined by prioritizing those that are not completed, those that have more repeated action triggers, and those that have shorter task durations, resulting in a target ranking of the task items. The task list order is generated by using the interaction dependency order list as a constraint. The action identifier set corresponding to the task list order is maintained. All task entries whose action prerequisites are included in the action identifier set are filtered in a loop and written according to the target. If there are no task entries to write, the remaining task entries are written in the order of the interaction dependency order list. The task list order in the interaction script is replaced with the task list order of the rearranged task list, and the rearranged interaction script is output.

Images