An interactive interface generation method, system and vehicle

Abstract

The present application relates to the technical field of human-computer interaction, and discloses an interactive interface generation method and system and a vehicle, wherein the present application acquires semantic analysis on multi-modal data, determines user intention, acquires historical interface generation data related to the user intention, and then generates an interactive interface display plan based on the historical interface generation data, the user intention and vehicle context information, calls vehicle-mounted services for execution to receive service data packets returned by each vehicle-mounted service, and finally renders and generates an interactive interface based on the interactive interface display plan and the service data packets, so that the user intention is combined with the historical interface generation data to continuously learn user usage habits, an interface plan and service combination that meet the personalized needs of the user can be quickly generated, and multiple vehicle-mounted service interfaces can be dynamically called based on task planning capability to form an interface calling combination that adapts to current needs.

Description

Technical Field

[0001] This invention relates to the field of human-computer interaction technology, specifically to an interactive interface generation method, system, and vehicle. Background Technology

[0002] Most current in-vehicle interfaces use fixed layouts and function mappings. For example, they typically use fixed-level menus to rigidly map functions to interface elements. When users need to perform tasks such as navigation or media playback, they have to manually complete the operation layer by layer in multi-level menus, resulting in a lengthy interaction path and low interaction efficiency, which affects operational efficiency and safety in the driving environment. Summary of the Invention

[0003] This invention provides an interactive interface generation method, system, and vehicle to solve the problem that using a fixed-level menu and interface element mapping results in a lengthy user interaction path, affecting operational efficiency and security.

[0004] In a first aspect, the present invention provides an interactive interface generation method, the method comprising: acquiring multimodal data input from a user and performing semantic analysis on the multimodal data to determine the user intent; acquiring historical interface generation data related to the user intent; generating an interactive interface display plan based on the user intent, the historical interface generation data, and the current vehicle context information, the interactive interface display plan including at least an interface display layout scheme and at least one in-vehicle service to be invoked; invoking the in-vehicle service for execution and receiving service data packets returned by each in-vehicle service; and rendering and generating an interactive interface based on the interactive interface display plan and the service data packets.

[0005] The interactive interface generation method provided by this invention acquires multimodal user data input, performs semantic analysis on the multimodal data to determine user intent, obtains historical interface generation data related to user intent, and then generates an interactive interface display plan including an interface display layout scheme and in-vehicle services to be invoked based on the historical interface generation data, user intent, and vehicle context information. The method then invokes in-vehicle services to receive service data packets returned by each service. Finally, based on the interactive interface display plan and service data packets, the interactive interface is rendered. This method combines user intent with historical interface generation data, continuously learns user habits, and can quickly generate interface plans and service combinations that meet personalized user needs. It also dynamically invokes multiple in-vehicle service interfaces based on task planning capabilities, forming an interface call combination adapted to current needs. Finally, based on the interface plan and the service data returned by the in-vehicle services, the interactive interface is rendered, ensuring a high degree of matching between interface content and user needs, and improving the smoothness and intelligence level of in-vehicle interaction.

[0006] In one optional implementation, obtaining historical interface generation data related to the user intent includes: encoding the user intent and the current vehicle context information to obtain a query vector; retrieving a set of candidate interface generation experiences that are semantically related to the query vector from a preset memory database; and selecting available interface generation experiences with memory weights that meet the requirements from the set of candidate interface generation experiences based on the memory weights corresponding to each candidate interface generation experience, as historical interface generation data related to the user intent, wherein the memory weights are dynamically adjusted based on the creation time and access frequency.

[0007] This invention is based on experience in generating candidate interfaces through semantic association retrieval. It supports the parsing and matching of complex multi-dimensional requirements, adapts to complex contexts and composite tasks, and utilizes a dynamic memory weighting filtering mechanism to endow the interactive interface generation system with memory and self-optimization capabilities, ensuring that the filtered experience is in line with the current functional iteration of the in-vehicle system and changes in user interaction habits.

[0008] In one optional implementation, generating an interactive interface display plan based on the user intent, the historical interface generation data, and the vehicle's current context information includes: planning a task sequence based on the user intent, the vehicle's current context information, and the historical interface generation data, wherein the task sequence includes at least one in-vehicle service to be invoked, the execution order of each in-vehicle service, the execution agent assigned to each in-vehicle service, and the display priority and layout scheme of the corresponding interface output of each in-vehicle service; and determining security constraints for interface interaction based on the vehicle's current context information, wherein the security constraints are used to prohibit or enforce specific interface generation or interaction behaviors.

[0009] This invention combines user intent, historical experience, and vehicle context information to achieve intelligent decision-making in core task decomposition, execution sequence planning, and agent allocation. It can efficiently handle complex tasks, improve the flexibility of task planning, and dynamically generate safety constraint rules based on vehicle context information, taking into account both interactive experience and driving safety.

[0010] In an optional implementation, the method further includes: performing intent matching and security verification on the generated interactive interface display plan; if the verification fails, generating reflection information containing error analysis and correction strategies; and re-executing the step of planning the task sequence based on user intent, vehicle current context information, and the historical interface generation data based on the reflection information.

[0011] This invention, through a comprehensive verification, reflection mechanism, and replanning strategy, can not only obtain an interactive interface that best matches the user's intentions, but also enable the system to improve its strategies based on historical experience, thereby increasing problem-solving efficiency.

[0012] In one optional implementation, the step of rendering and generating the interactive interface based on the interactive interface display plan and the service data package includes: generating a structured interface description specification based on the interactive interface display plan and the service data package, wherein the interface description specification includes at least element tags, data binding relationships, and layout display constraints for the required interface elements; matching corresponding component templates from an atomic component library using a component selector based on the interface description specification, and binding the corresponding data in the service data package to the component templates to obtain an interface component set; performing layout calculations on the interface component set using a layout optimizer based on the interface description specification and the current vehicle context information to generate a layout scheme; and rendering and generating the interactive interface based on the layout scheme.

[0013] The component selector of this invention can accurately match the required components based on the interface description specification and quickly combine them through data binding to adapt to the new interface requirements corresponding to personalized and complex tasks. Furthermore, through the layout optimizer, which combines the interface description specification with the current context information of the vehicle, the layout scheme is dynamically calculated to ensure that it fits the real-time in-vehicle scenario and enhances the safety of interaction.

[0014] In one optional implementation, before matching the corresponding component template from the atomic component library, the method further includes: retrieving historical interface templates from the interface database based on the interface description specification; if a matching historical interface template is found from the interface database, rendering and generating an interactive interface based on the historical interface template; if no matching historical interface template is found from the interface database, performing the step of matching the corresponding component template from the atomic component library using a component selector.

[0015] This invention introduces a historical interface template retrieval mechanism, which prioritizes the reuse of templates matched in the interface database. For frequently occurring interaction needs in vehicle scenarios, mature templates can be directly called. Only the binding data needs to be updated according to the current service data package, which significantly shortens the rendering time and avoids affecting the operation of other core vehicle functions due to excessive rendering resources.

[0016] In an optional implementation, the method further includes: collecting interaction logs between the user and the interactive interface; calculating a reward signal based on the interaction logs; and updating the policy network parameters of the layout optimizer and the component preference parameters of the component selector based on the reward signal.

[0017] This invention collects real-time interaction logs between users and the interactive interface, calculates reward signals, and updates the layout optimizer and component selector using the reward signals to achieve adaptive evolution of the rendering strategy and continuously adapt to user habits.

[0018] In an optional implementation, the method further includes: storing the interface generation data corresponding to the interactive interface as interface generation experience in a memory database, and updating the memory weights corresponding to each interface generation experience in the memory database.

[0019] This invention uses the full amount of interface generation data of the current generation of interactive interface as interface generation experience, stores it in the memory database, and updates the memory weights corresponding to each interface generation experience, thereby maintaining the lightweight and efficient nature of the memory database.

[0020] In an optional implementation, the method further includes: uploading the updated policy network parameters of the layout optimizer and the component preference parameters of the component selector to a federated aggregator for aggregation to obtain a global model update, so as to distribute the updated global model to each vehicle locally; and receiving and applying the global model update to the local model.

[0021] In one optional implementation, the step of performing semantic analysis on the multimodal data to determine user intent includes: performing feature fusion processing on the acquired multimodal data to obtain a multimodal feature vector; acquiring historical user intent-related data, wherein the user intent-related data includes at least historical dialogues and user preferences; concatenating the multimodal feature vector with the user intent-related data to obtain feature enhancement information; and performing semantic analysis on the feature enhancement information to determine user intent.

[0022] This invention integrates feature information from different modalities to generate multimodal feature vectors, enabling accurate parsing of ambiguous commands. It also incorporates historical intent-related data, allowing for the understanding of deeper needs based on users' past interaction habits and improving the accuracy of intent recognition.

[0023] Secondly, the present invention provides an interactive interface generation system, the system comprising an intent determination module, an intelligent agent module, and an interface generation module, wherein: the intent determination module is used to acquire multimodal data input from a user and perform semantic analysis on the multimodal data to determine the user's intent; the intelligent agent module is used to acquire historical interface generation data related to the user's intent; generate an interactive interface display plan based on the user's intent, the historical interface generation data, and the vehicle's current context information, the interactive interface display plan including at least an interface display layout scheme and at least one in-vehicle service to be invoked; invoke the in-vehicle service to execute, and receive service data packets returned by each in-vehicle service; the interface generation module is used to render and generate an interactive interface based on the interactive interface display plan and the service data packets.

[0024] Thirdly, the present invention provides a vehicle, the vehicle including an interactive interface generation system, the system including a controller, the controller including a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the interactive interface generation method of the first aspect or any corresponding embodiment described above.

[0025] The present invention has the following technical effects:

[0026] The interactive interface generation method provided by this invention acquires multimodal user data input, performs semantic analysis on the multimodal data to determine user intent, obtains historical interface generation data related to user intent, and then generates an interactive interface display plan including an interface display layout scheme and in-vehicle services to be invoked based on the historical interface generation data, user intent, and vehicle context information. The method then invokes in-vehicle services to receive service data packets returned by each service. Finally, based on the interactive interface display plan and service data packets, the interactive interface is rendered. This method combines user intent with historical interface generation data, continuously learns user habits, and can quickly generate interface plans and service combinations that meet personalized user needs. It also dynamically invokes multiple in-vehicle service interfaces based on task planning capabilities, forming an interface call combination adapted to current needs. Finally, based on the interface plan and the service data returned by the in-vehicle services, the interactive interface is rendered, ensuring a high degree of matching between interface content and user needs, and improving the smoothness and intelligence level of in-vehicle interaction. Attached Figure Description

[0027] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0028] Figure 1 This is a schematic diagram of the first type of interactive interface generation method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a second process for generating an interactive interface according to an embodiment of the present invention; Figure 3 This is a logical architecture example diagram of the process of generating multimodal feature vectors according to an embodiment of the present invention; Figure 4 This is an example diagram of the process logic architecture for determining user intent according to an embodiment of the present invention; Figure 5This is an example diagram of the process architecture for generating an interactive interface to display planning and safety constraints according to an embodiment of the present invention. Figure 6 This is a structural block diagram of the interface generation module according to an embodiment of the present invention; Figure 7 This is an example diagram of a feedback learning logic framework according to an embodiment of the present invention; Figure 8 This is a structural block diagram of an interactive interface generation system according to an embodiment of the present invention; Figure 9 This is a logical architecture example diagram of an interactive interface generation method according to an embodiment of the present invention; Figure 10 This is a structural block diagram of a vehicle according to an embodiment of the present invention; Figure 11 This is a schematic diagram of the hardware structure of the controller according to an embodiment of the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0030] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0031] Current in-vehicle interfaces mostly adopt fixed layouts and function mappings, requiring drivers to navigate through layers of menus, resulting in low interaction efficiency. Related solutions propose using multimodal feature input to a large language model to identify user intent and generate interface descriptions. While this reduces manual search operations, interface generation relies on a static Domain-Specific Language (DSL), making it difficult to handle complex tasks and contexts. Furthermore, it cannot dynamically combine multiple interfaces and lacks memory and learning capabilities. To address these issues, this invention proposes an interactive interface generation system with planning, memory, self-reflection, and federated learning capabilities, achieving multimodal semantic understanding, task planning, dynamic interface calls, and adaptive interface generation.

[0032] According to an embodiment of the present invention, an embodiment of an interactive interface generation method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0033] This embodiment provides an interactive interface generation method, which can be used in an interactive interface generation system. The interactive interface generation system includes at least an intent determination module, an agent module, an interface generation module, and a feedback learning module. Figure 1 This is a flowchart of an interactive interface generation method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Obtain the user's multimodal data input, and perform semantic analysis on the multimodal data to determine the user's intent.

[0034] This invention does not limit the type of multimodal data, which may include voice, gaze, and gesture recognition data. User voice, gaze, and gesture data can be collected through microphone arrays, cameras, and gesture sensors. Then, voice analysis technology (combined with algorithms such as Natural Language Processing (NLP) and computer vision) can be used to analyze the core demands of the multimodal data and clarify the user's intent. The user intent can be composed of a domain, intent, and slot. For example, if a passenger says "order a latte to be delivered to my workplace," the multimodal data can be analyzed to identify the ordering intent and extract the slot: {product = latte, delivery location = workplace}.

[0035] Step S102: Obtain historical interface generation data related to user intent.

[0036] The interactive interface generation system or cloud-based system of this invention has a memory database that stores historical trajectory data of generated interactive interfaces (i.e., interface generation data), including but not limited to interface layout, service call data, user interaction preferences, etc. This is only an example and is used to provide reference data for interface generation in the future.

[0037] In this embodiment of the invention, after determining the user's intent, the intelligent agent module can retrieve historical interface data related to the user's intent from the memory database. For example, based on the determined intent {product = latte coffee, delivery location = workplace}, the module can retrieve trajectory data of historical interactive interfaces generated based on the query features such as "latte coffee" and "workplace" from the memory database, discover the user's preference for target brand coffee, and obtain the specific address data of the workplace. This is just an example.

[0038] Step S103: Generate an interactive interface display plan based on user intent, historical interface generation data, and the vehicle's current context information.

[0039] The interactive interface display plan includes at least an interface display layout scheme and at least one in-vehicle service to be invoked.

[0040] The intelligent agent module of this invention can integrate user intent, historical interface generation data, and current vehicle context information (including but not limited to vehicle speed, gear position, screen configuration, driving load, road conditions, and weather information) through an algorithm model. It comprehensively considers the optimal presentation form of the interface and the required services to generate an interactive interface display plan. The interactive interface display plan includes an interface display layout scheme and at least one in-vehicle service to be invoked. The interface display layout scheme is used to constrain "what to display, what to display first, where to display, and what interactions are allowed", including but not limited to the display priority and display layout of different in-vehicle services. The in-vehicle service to be invoked can be determined based on user intent and historical interface generation data. For example, if it is determined that the user prefers a target brand of coffee, it can be determined to invoke a food delivery service. If the user wants to navigate to their workplace, it can be determined to invoke a navigation service. This is just an example.

[0041] Step S104: Invoke the vehicle service to execute and receive the service data packets returned by each vehicle service.

[0042] In this embodiment of the invention, the in-vehicle service is a sub-agent agent. Based on the invocation command, it can trigger the execution of corresponding tasks and generate service data packets. These service data packets are organized in the form of key-value pairs, lists, tables, or time series, and carry timestamps, data source identifiers, validity periods, and necessary confidence fields. Examples include: navigation services returning a route list and key turning points; weather services returning hourly / daily forecasts; and food delivery services returning a list of stores, a list of products, and prices. Specifically, if the agent module determines to invoke the food delivery service, it can query the food delivery service interface in the interface description directory, dynamically assemble the invocation command, and send it outwards. The food delivery service interface sends a call command and then obtains a service data packet from the food delivery service. Based on the call command containing "target brand latte" and "unit address," the food delivery service obtains the coffee shops that can deliver to the target brand, as well as the list and price of latte products at those coffee shops. It then integrates the coffee shops, the latte product list, and the prices into a service data packet and feeds the generated service data packet back to the interface generation module. This is just one example; it demonstrates how to introduce multiple sub-agent agents (navigation, entertainment, vehicle control, etc.) within the core agent, improving the processing capability for complex tasks by scheduling tasks and sharing feedback.

[0043] Step S105: Based on the interactive interface display planning and service data package, render and generate the interactive interface.

[0044] The interface generation module of this invention can fill the content of the service data package into the corresponding interface module according to the interface layout scheme in the interactive interface display plan, and complete the visualization drawing of the interface elements through the rendering engine. Finally, the drawn interactive interface is displayed on terminals such as the vehicle central control screen and the instrument panel. In a specific example, the interface generation module can display a list of stores and a latte coffee product card in the interactive interface based on the service data package fed back by the food delivery service. After the user selects the product and confirms the order, the payment interface can be called to complete the order. This is just an example.

[0045] The interactive interface generation method provided in this embodiment acquires multimodal user input data, performs semantic analysis on the multimodal data to determine user intent, and obtains historical interface generation data related to user intent. Then, based on the historical interface generation data, user intent, and vehicle context information, it generates an interactive interface display plan that includes an interface display layout scheme and in-vehicle services to be invoked. It calls in-vehicle services to execute and receives service data packets returned by each in-vehicle service. Finally, based on the interactive interface display plan and service data packets, it renders and generates the interactive interface. This method combines user intent with historical interface generation data, continuously learns user habits, and can quickly generate interface plans and service combinations that meet personalized user needs. It also dynamically calls multiple in-vehicle service interfaces based on task planning capabilities to form an interface call combination adapted to the current needs. Finally, based on the interface plan and the service data returned by the in-vehicle services, it renders and generates the interactive interface, ensuring that the interface content is highly matched with user needs and improving the smoothness and intelligence level of in-vehicle interaction.

[0046] This embodiment provides an interactive interface generation method, which can be used in an interactive interface generation system. The interactive interface generation system includes at least an intent determination module, an agent module, an interface generation module, and a feedback learning module. Figure 2 This is a flowchart of an interactive interface generation method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S201: Obtain the user's multimodal data input, perform semantic analysis on the multimodal data, and determine the user's intent.

[0047] Specifically, user intent can be determined through the following steps: perform feature fusion processing on the acquired multimodal data to obtain multimodal feature vectors; acquire user intent-related data from historical records, which includes at least historical dialogues and user preferences; concatenate the multimodal feature vectors with the user intent-related data to obtain feature enhancement information; perform semantic analysis on the feature enhancement information to determine user intent.

[0048] The intent determination module designed in this embodiment of the invention may include a multimodal input module and a semantic understanding module. The multimodal input module can capture speech through an in-vehicle microphone array and obtain clear speech through noise suppression and sound source localization. Then, it can obtain speech text through Automatic Speech Recognition (ASR) technology. A camera captures the driver's eye movement trajectory and captures the gaze point, and a gesture sensor recognizes gesture data. Then, the multimodal data is time-synchronized and fused to obtain a multimodal feature vector. The process of generating the multimodal feature vector can be referred to as follows: Figure 3 As shown.

[0049] In the data fusion process of this invention, multimodal pronoun disambiguation technology (GazePointAR) can be used to dissolve pronouns in speech using gaze and gesture data features. To alleviate contextual constraints, the semantic understanding module can introduce a vector database as external memory to obtain historical user intent-related data from the vector database. This intent-related data includes at least historical dialogues and user preferences. Specifically, the historical user intent-related data can adopt a hierarchical and controllable memory mechanism, including at least historical dialogues (covering dialogues and states within the current ignition cycle or trip, used for short-term contextual connections), user preferences (retaining information related to long-term user preferences across trips, such as frequently used destinations, preferred content, and interface density), where user preferences can be weighted using a forgetting curve to avoid interference from outdated preferences, and driving scenarios (information strongly associated with specific scenarios, such as route preferences to the company being linked to weekday morning rush hour), which are prioritized for retrieval in similar scenarios. As an example, during retrieval, user identification, time window, similarity threshold, and memory weight can be used for joint filtering. Cross-trip memory retrieval can be disabled during user switching or privacy mode.

[0050] The semantic understanding module of this invention can encode multimodal feature vectors, concatenate the encoded multimodal feature vectors with user intent-related data to obtain feature enhancement information, and then input the feature enhancement information into a preset Large Language Model (LLM) for semantic understanding, outputting the domain, intent, and slot representing the user intent. The process logic architecture for determining the user intent can be found in [reference needed]. Figure 4 As shown in the specific embodiment, the driver issues the instruction "navigate to the company and play today's weather forecast". The speech is captured by the microphone array and generated into text. At the same time, the camera confirms the driver's gaze direction to help eliminate the destination reference "company". Then, the user intent-related data is concatenated with the multimodal feature vector, and then semantic understanding is performed to generate two intents: {navigation, destination = company} and {play weather, type = today's weather}.

[0051] This invention integrates feature information from different modalities to generate multimodal feature vectors, enabling accurate parsing of ambiguous commands. It also incorporates historical intent-related data, allowing for the understanding of deeper needs based on users' past interaction habits and improving the accuracy of intent recognition.

[0052] Step S202: Obtain historical interface generation data related to user intent.

[0053] Specifically, step S202 above includes: Step S2021: Encode the user intent and the current vehicle context information to obtain the query vector.

[0054] Step S2022: Retrieve a set of candidate interfaces that are semantically related to the query vector from the preset memory database.

[0055] Step S2023: Based on the memory weights corresponding to each candidate interface generation experience, select usable interface generation experiences that meet the memory weight requirements from the candidate interface generation experience set, and use them as historical interface generation data related to user intent. The memory weights are dynamically adjusted based on the creation time and access frequency.

[0056] In the memory database designed in this invention, each interface generation experience has a corresponding memory weight. The memory weight can be dynamically adjusted using the Ebbinghaus forgetting curve. The memory weight is related to the creation time and access frequency. The amount of memory retained decays exponentially over time. For example, the time interval between the current time and the experience creation time is calculated, and the memory weight is decayed based on the product of the decay coefficient and the time interval. If the interface generation experience is accessed later, the most recent access time can replace the previous creation time, and the time interval is recalculated to achieve the effect of refreshing the memory through access. The access frequency reflects the practical value of the interface generation experience. The more times it is accessed and the more recent the experience, the higher the reinforcement factor. Then, the time decay factor and the access frequency factor are integrated to obtain the final memory weight of each interface generation experience. This is just an example to achieve the effect of retaining frequently used information and preventing information overload.

[0057] The intelligent agent module in this embodiment of the invention can encode user intent and vehicle current context information to obtain a query vector. The specific encoding method is not limited. For example, a semantic encoding model can be used to encode the user intent into text to generate a semantic vector, and the vehicle context information can be numerically encoded to generate a context feature vector. Then, the semantic vector and the context feature vector are weighted and fused to output the query vector. This is just an example. Then, a vector similarity calculation algorithm can be used to compare the semantic correlation between the query vector and all experience vectors in the memory database one by one, and screen out the interface generation experience with similarity higher than a preset threshold to form a candidate interface generation experience set.

[0058] The embodiments of the present invention can filter out usable interface generation experiences that meet the requirements of memory weight from the candidate interface generation experience set based on the memory weight corresponding to each candidate interface generation experience. For example, a memory weight threshold can be set to filter out usable interface generation experiences with memory weight greater than the memory weight threshold, which can be used as historical interface generation data related to user intent.

[0059] This invention is based on experience in generating candidate interfaces through semantic association retrieval. It supports the parsing and matching of complex multi-dimensional requirements, adapts to complex contexts and composite tasks, and utilizes a dynamic memory weighting filtering mechanism to endow the interactive interface generation system with memory and self-optimization capabilities, ensuring that the filtered experience is in line with the current functional iteration of the in-vehicle system and changes in user interaction habits.

[0060] Step S203: Based on user intent, historical interface generation data, and the vehicle's current context information, generate an interactive interface display plan. This plan includes at least an interface layout scheme and at least one in-vehicle service to be invoked.

[0061] Specifically, the interactive interface display plan is generated through the following steps: Based on user intent, current vehicle context information, and historical interface generation data, a task sequence is planned. The task sequence includes at least one in-vehicle service to be invoked, the execution order of each in-vehicle service, the execution agent assigned to each in-vehicle service, and the display priority and layout scheme of the corresponding interface output of each in-vehicle service; Based on the current vehicle context information, safety constraints for interface interaction are determined. These safety constraints are used to prohibit or force specific interface generation or interaction behaviors.

[0062] The intelligent agent module of this invention obtains the current context information of the vehicle, including but not limited to information such as vehicle speed, gear position, screen configuration, driving load, road conditions and weather. It is used to determine the planning of display priority and interaction method. For example, during vehicle driving, the navigation service has a higher priority than the entertainment service, and only voice interaction is allowed. This is just an example.

[0063] The intelligent agent module designed in this embodiment of the invention can perform hierarchical reinforcement learning to plan task sequences based on user intent, historical interface generated data, and the current vehicle context information. This can include, but is not limited to, various in-vehicle services, execution agents assigned to these services, tool / interface calls at each step, expected output data structures, and the display priority and layout scheme of the corresponding interfaces for each in-vehicle service. Specifically, based on user intent and historical interface generated data, core tasks can be determined. For example, if parsing the user intent determines that the core tasks are navigation and weather playback, then the in-vehicle services to be called include navigation and weather services. Execution agents are assigned to each in-vehicle service. The navigation agent is called to obtain the route and start voice navigation, while the weather agent obtains weather data from the weather service interface. This enables dynamic discovery and invocation of in-vehicle service interfaces, overcoming the limitations of related solutions that can only call... By addressing the rigidity of pre-registered plugins, and integrating sensor data such as vehicle speed, location, and weather, a hierarchical reinforcement learning strategy is employed between high-level planning and low-level execution to achieve personalized decision-making based on vehicle information. For instance, if the vehicle is currently in motion, a self-reflection mechanism can determine that navigation is the primary task and a higher priority action. The central planner in the intelligent agent module can use hierarchical reinforcement learning to determine the execution order of various in-vehicle services. For example, it can first call the navigation agent to obtain the route and start navigation, then activate the weather agent to obtain weather data from the weather service interface. In subsequent displays, the navigation interface can be prioritized, and navigation cards and key turn prompts can be set to be displayed on the main screen, while weather summary cards can be displayed on the secondary screen or HUD. This is just one example to ensure that important driving information is obtained in a timely manner and to achieve personalized decision-making based on vehicle information.

[0064] The intelligent agent module in this embodiment of the invention can also determine safety constraints for interface interaction based on the current context information of the vehicle. These constraints are used to prohibit or enforce specific interface generation or interaction behaviors, such as prohibiting complex input while driving, allowing interaction only on the secondary screen, or allowing operation only by voice. This is just one example; the process architecture for generating the interactive interface display plan and safety constraints can be found in [reference needed]. Figure 5 As shown.

[0065] This invention utilizes a hierarchical reinforcement learning strategy, combining user intent, historical experience, and vehicle context information, to achieve intelligent decision-making in core task decomposition, execution sequence planning, and agent allocation. It can efficiently handle complex tasks, improve the flexibility of task planning, and dynamically generate safety constraint rules based on vehicle context information, taking into account both interactive experience and driving safety.

[0066] Step S204: Invoke the vehicle services and receive the service data packets returned by each vehicle service. For details, please refer to [link to relevant documentation]. Figure 1 Step S104 of the illustrated embodiment will not be described again here.

[0067] In one alternative implementation, the generated interactive interface display plan is subjected to intent matching and security verification; if the verification fails, reflection information containing error analysis and correction strategies is generated; based on the reflection information, the steps of planning the task sequence based on user intent, vehicle current context information and historical interface generation data are re-executed.

[0068] After generating the interactive interface display plan, the intelligent agent module of this embodiment can perform a full-dimensional verification and evaluation of the generated plan and the output of each step, including intent matching and security verification. Specifically, it can detect whether the slot information of the semantic understanding output is complete to avoid service abnormalities caused by the omission of key information. It can also detect whether the interface display plan is consistent with the user's intent. For example, if the user's intent is high-speed navigation, it can check whether the high-speed priority route service is called. This is just an example. It can also detect whether driving safety constraints are triggered, such as whether a complex touch interface is planned during driving or whether secondary confirmation is required (such as changing the navigation destination, adjusting the volume, etc.). This is just an example. If the verification is successful, the successful experience can be recorded and the generated interface display plan can be stored in the memory database for subsequent planning.

[0069] If the verification fails, a self-reflection loop can be triggered to generate reflection information according to the specifications. For example, reflection text and structured reflection tags containing the failure type, possible causes, and correction strategies can be generated to facilitate quick system parsing. The failure experience can also be written back to the memory database to strengthen the warning effect of failure cases. Subsequently, the structured reflection tags can be read to initiate targeted replanning or supplementary inquiries. For example, if the error is due to an incomplete slot, supplementary inquiries can be sent to the user. If the failure is due to a strategy adaptation issue, the interactive interface display plan can be regenerated according to the correction strategy. This is just an example.

[0070] This invention, through a comprehensive verification, reflection mechanism, and replanning strategy, can not only obtain an interactive interface that best matches the user's intentions, but also enable the system to improve its strategies based on historical experience, thereby increasing problem-solving efficiency.

[0071] Step S205: Based on the interactive interface display planning and service data package, render and generate the interactive interface.

[0072] Specifically, the interactive interface is rendered and generated through the following steps: Based on the interactive interface display plan and service data package, a structured interface description specification is generated. The interface description specification includes at least the element tags, data binding relationships, and layout display constraints of the required interface elements; based on the interface description specification, historical interface templates are retrieved from the interface database; if a matching historical interface template is found in the interface database, the interactive interface is rendered and generated based on the historical interface template.

[0073] The service data packets obtained in this embodiment of the invention determine the content filling, and the interactive interface display plan determines the information organization method and interaction process, thereby making the interface specifications executable and secure.

[0074] like Figure 6 As shown, the interface generation module designed in this embodiment of the invention includes a specification generator, a first component selector, a second component selector, an interface database, an atomic component library, a layout optimizer (RL-based), a context and layout constraint module, and a rendering engine. The User Interface (UI) specification generator receives the interactive interface display plan and service data package output by the intelligent agent module, and can then call a Large Language Model (LLM) to convert it into a structured interface specification, such as a UI description in JSON format. The interface description specification includes at least element tags (i.e., the types of required components), data binding relationships, layout display constraints, and priorities of the required interface elements, achieving a structured translation of rendering requirements. This avoids rendering deviations caused by ambiguous requirement transmission and ensures a high degree of consistency between the final interface and the planning scheme. Specifically, element tags are structured identifier fields in the interface specification used to describe the semantics and constraints of each interface element. They drive the atomic component library retrieval and component matching to avoid uncertain matching caused by relying on natural language descriptions. Element tags include at least: semantic role tags (e.g., navigation main task card, weather summary card, optional list, confirmation button) and data type tags (e.g., route node sequence, time series, store list, SKU). The list, interaction type tags (click / swipe / voice confirmation / secondary confirmation), and safety constraint tags (prohibit complex input while driving, only secondary screen can be interacted with, only voice can be operated, etc.) are just examples.

[0075] The interface database designed in this embodiment of the invention is stored locally and used to save historical interface templates and for retrieval. The interface templates include, but are not limited to, historically generated interface specifications and their layout results (cached and reused), reusable interface pattern fragments, such as the "list-details-confirm" process template, user preference-related interface parameters, such as font size, card density, summary level, and historical feedback summaries of interface generation and layout optimization (used for sorting and selection), which are only examples.

[0076] The first component selector designed in this embodiment of the invention is a cache / mode selector. It can first search the interface database for a historical interface template that matches the interface description specification. If a matching historical interface template is found in the interface database, it can be directly rendered based on the historical interface template to generate an interactive interface. In an optional implementation, the component selector can directly search the interface database for a historical interface description specification that matches the current interactive interface display plan and service data package. If it exists, the historical interface description specification is directly obtained as the current interface description specification, and then the interface database is searched based on the current interface description specification to reduce resource consumption. This is just an example.

[0077] This invention introduces a historical interface template retrieval mechanism, which prioritizes the reuse of templates matched in the interface database. For frequently occurring interaction needs in vehicle scenarios, mature templates can be directly called. Only the binding data needs to be updated according to the current service data package, which significantly shortens the rendering time and avoids affecting the operation of other core vehicle functions due to excessive rendering resources.

[0078] In one optional implementation, if no matching historical interface template is found in the interface database, based on the interface description specification, the corresponding component template is matched from the atomic component library through the component selector, and the corresponding data in the service data packet is bound to the component template to obtain an interface component set; based on the interface description specification and the current vehicle context information, the layout optimizer performs layout calculations on the interface component set to generate a layout scheme; based on the layout scheme, the interactive interface is rendered and generated.

[0079] like Figure 6 As shown, if the first component selector fails to find a matching historical interface template in the interface database, the second component selector can match component instances from the atomic component library based on the interface description specification and complete data binding, outputting a set of interface components for the layout optimizer to perform layout retrieval. Specifically, the atomic component library stores visualization templates, size ranges, screen adaptation rules, and matching metadata with data types for each type of atomic component such as cards, lists, buttons, and charts. The second component selector can search for suitable component templates in the atomic component library based on the element tags in the interface description specification. The second component selector selects the corresponding component template based on the element tags and data constraints in the interface description specification, and can bind the corresponding data in the service data package to the component template to obtain a set of interface components. If a certain type of component is missing, a design template or generative model can be called to dynamically create a new component style, which is only an example.

[0080] In this embodiment of the invention, the second component selector can output the selected set of interface components to the layout optimizer. The layout optimizer performs layout calculations on the set of interface components according to the interface description specifications (such as component priority and size) and the current vehicle context information (such as vehicle speed, driving status, and screen size), determines the position and size of each component, and obtains a layout scheme. The layout optimizer can use reinforcement learning algorithms to iteratively generate multiple layout schemes in a simulated training or evaluation environment, and then reward users based on feedback indicators such as user dwell time and click frequency, so that the interface can dynamically adapt to individual preferences and reduce driving interference, thereby improving the layout accuracy of the layout optimizer. To ensure real-time performance and safety, small-scale exploration and updates are only performed when safety constraints are met during the online phase, and the layout strategy can be evaluated offline in a simulated environment or traffic split experiment.

[0081] In the layout scheme generation process of this invention, the context and layout constraint module can provide the layout optimizer with information such as real-time driving status, number and size of screens, and user preferences, and set hard and soft constraints for the layout optimizer. For example, when driving at high speed, complex operation interfaces are prohibited from being displayed on the main driving screen, and only non-critical information is presented on the secondary screen to improve driving safety and interactive experience. As an example only, the rendering engine can render components onto appropriate screens according to the layout scheme output by the layout optimizer and process visual effects such as animation and theme colors. The rendered components are updated in real time through data binding to ensure information synchronization.

[0082] The component selector of this invention can accurately match the required components based on the interface description specification and quickly combine them through data binding to adapt to the new interface requirements corresponding to personalized and complex tasks. Furthermore, through the layout optimizer, which combines the interface description specification with the current context information of the vehicle, the layout scheme is dynamically calculated to ensure that it fits the real-time in-vehicle scenario and enhances the safety of interaction.

[0083] In one alternative implementation, user interaction logs with the interactive interface can be collected; based on the interaction logs, a reward signal is calculated, and the policy network parameters of the layout optimizer and the component preference parameters of the component selector are updated based on the reward signal.

[0084] The feedback collector in this embodiment of the invention can collect user interaction logs with the interactive interface, and then feed them back to the reinforcement learning updater through a privacy protection mechanism. The reinforcement learning updater calculates reward signals based on the interaction logs. Specifically, the reward signals are automatically calculated from the interaction logs collected during operation, and include at least the key component reach rate / click rate (such as a turn prompt being confirmed, a product card being clicked, etc.), key task completion time (the shorter the time, the higher the reward), effective dwell time index of the main information area (which can be calculated based on the gaze point or dwell time normalized), negative feedback penalties such as accidental touch, undo, and repeated questioning, as well as driving safety penalties (such as complex interaction attempts on the driver's screen at high speed, and the time the gaze is away from the road exceeding a threshold, etc.), which are only examples.

[0085] The embodiments of the present invention can continuously update the layout strategy based on the reward signal to achieve continuous adaptation. The updater can also be combined with chain feedback training to integrate text feedback into the UI specification generator, thereby improving the accuracy of the large language model in interface design.

[0086] The feedback learning module in this embodiment of the invention can combine federated learning and Chain of Hindsight techniques to update the policy network parameters of the layout optimizer, making its output more user-preferred and safer within the same context. It can also update the component preference parameters of the component selector, such as the priority of interface pattern fragments, card density, and summarization weight, thereby influencing subsequent interface specification generation and component matching results. See the detailed feedback learning logic framework diagram below. Figure 7 As shown, by dynamically adjusting the interface layout and function order based on feedback rewards, a balance between personalization and security can be achieved.

[0087] This invention collects real-time interaction logs between users and the interactive interface, calculates reward signals, and updates the layout optimizer and component selector using the reward signals to achieve adaptive evolution of the rendering strategy and continuously adapt to user habits.

[0088] Furthermore, the updated strategy network parameters of the layout optimizer and the component preference parameters of the component selector are uploaded to the federated aggregator for aggregation to obtain a global model update, which is then distributed to each vehicle's local machine; the updated global model is received and applied to the local model.

[0089] This invention allows the updated policy network parameters of the layout optimizer and the component preference parameters of the component selector to be uploaded to a federated aggregator for aggregation. A weighted average is then calculated to obtain a global model update. By uploading lightweight model parameters to the federated aggregator, privacy and bandwidth control are better guaranteed compared to the original user data. The updated global model serves as the initialization prior and periodic alignment baseline for each vehicle model. After being distributed to each vehicle locally, it undergoes self-reflection and adjustment based on its own memory. Through multiple iterations, the semantic understanding and interface generation accuracy of vehicles in different cities are significantly improved, demonstrating the effectiveness of federated learning and chain feedback training. This achieves group experience transfer + local personalization, realizing a continuous learning closed loop with privacy protection.

[0090] In one optional implementation, the interface generation data corresponding to the interactive interface is stored in the memory database as interface generation experience, and the memory weights corresponding to each interface generation experience in the memory database are updated.

[0091] After the interactive interface is rendered, this invention can automatically extract all the interface generation data generated in this instance, integrate it into standardized interface generation experience, and store it synchronously in the memory database. Then, based on the Ebbinghaus forgetting curve, the memory weight of all interface generation experience in the memory database can be updated in real time to enhance the reuse value of high-quality experience.

[0092] This invention uses the full amount of interface generation data of the current generation of interactive interface as interface generation experience, stores it in the memory database, and updates the memory weights corresponding to each interface generation experience, thereby maintaining the lightweight and efficient nature of the memory database.

[0093] This embodiment also provides an interactive interface generation system, such as Figure 8 As shown, the interactive interface generation system includes at least an intent determination module 801, an intelligent agent module 802, and an interface generation module 803. The intent determination module 801 is used to acquire multimodal data input from the user and perform semantic analysis on the multimodal data to determine the user's intent. The intelligent agent module 802 is used to acquire historical interface generation data related to the user's intent; based on the user's intent, historical interface generation data, and the vehicle's current context information, it generates an interactive interface display plan, which includes at least an interface display layout scheme and at least one in-vehicle service to be invoked; it invokes the in-vehicle service for execution and receives service data packets returned by each in-vehicle service; the interface generation module 803 is used to render and generate the interactive interface based on the interactive interface display plan and service data packets. For detailed explanations, please refer to the above embodiments, which will not be repeated here. The overall logical architecture of the interactive interface generation method can be found in [reference needed]. Figure 9 The example shown is for illustrative purposes only.

[0094] This embodiment also provides a vehicle, such as Figure 10 As shown, the vehicle includes an interactive interface generation system 1001. The system includes a controller, which includes a memory and a processor. The memory and the processor are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to perform the interactive interface generation method described in the above embodiments.

[0095] Figure 11 This is a schematic diagram of the structure of a controller provided in an embodiment of the present invention.

[0096] The following is a detailed reference. Figure 11 The diagram illustrates a structural schematic suitable for implementing a controller in an embodiment of the present invention. The controller may include a processor (e.g., a central processing unit, graphics processing unit, etc.) 1101, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 1102 or a program loaded from memory 1108 into random access memory (RAM) 1103. The RAM 1103 also stores various programs and data required for controller operation. The processor 1101, ROM 1102, and RAM 1103 are interconnected via a bus 1104. An input / output (I / O) interface 1105 is also connected to the bus 1104.

[0097] Typically, the following devices can be connected to I / O interface 1105: input devices 1106 including, for example, a touchscreen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 1107 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; memory 1108 including, for example, magnetic tape, hard disk, etc.; and communication devices 1109. Communication device 1109 allows the controller to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 11 A controller with various devices is shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0098] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 1109, or installed from a memory 1108, or installed from a ROM 1102. When the computer program is executed by the processor 1101, it performs the functions defined in the interactive interface generation method of the embodiments of the present invention.

[0099] Figure 11 The controller shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0100] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the interactive interface generation method shown in the above embodiments is implemented.

[0101] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0102] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and all such modifications and variations fall within the defined scope.

Claims

1. A method for generating an interactive interface, characterized in that, The method includes: Acquire multimodal data input from the user, and perform semantic analysis on the multimodal data to determine the user's intent; Obtain historical interface generation data related to the user intent; Based on the user intent, the historical interface generation data, and the vehicle's current context information, an interactive interface display plan is generated. The interactive interface display plan includes at least an interface display layout scheme and at least one in-vehicle service to be invoked. The system invokes the in-vehicle services to execute and receives service data packets returned by each in-vehicle service. Based on the interactive interface display plan and the service data package, the interactive interface is rendered and generated.

2. The method according to claim 1, characterized in that, The acquisition of historical interface generation data related to the user intent includes: The user intent and the vehicle's current context information are encoded to obtain a query vector; Retrieve a set of candidate interfaces semantically related to the query vector from a preset memory database to generate an experience set; Based on the memory weights corresponding to the candidate interface generation experiences, usable interface generation experiences that meet the memory weight requirements are selected from the candidate interface generation experience set and used as historical interface generation data related to user intent. The memory weights are dynamically adjusted based on the creation time and access frequency.

3. The method according to claim 1, characterized in that, The step of generating an interactive interface display plan based on the user intent, the historical interface generation data, and the vehicle's current context information includes: Based on user intent, current vehicle context information, and historical interface generation data, a task sequence is planned. The task sequence includes at least one in-vehicle service to be invoked, the execution order of each in-vehicle service, the execution agent assigned to each in-vehicle service, and the display priority and layout scheme of the corresponding interface output of each in-vehicle service. Based on the current context information of the vehicle, safety constraints for interface interaction are determined, which are used to prohibit or enforce specific interface generation or interaction behaviors.

4. The method according to claim 3, characterized in that, The method further includes: Perform intent matching and security checks on the generated interactive interface display plan; If the verification fails, a reflection message containing error analysis and correction strategies will be generated. Based on the reflected information, the steps of planning the task sequence based on user intent, current vehicle context information, and historical interface generation data are re-executed.

5. The method according to claim 1, characterized in that, The step of rendering and generating the interactive interface based on the interactive interface display plan and the service data package includes: Based on the interactive interface display plan and service data package, a structured interface description specification is generated. The interface description specification includes at least the element labels, data binding relationships, and layout display constraints of the required interface elements. Based on the interface description specification, the corresponding component template is matched from the atomic component library through the component selector, and the corresponding data in the service data package is bound to the component template to obtain the interface component set; Based on the interface description specification and the current vehicle context information, the layout optimizer performs layout calculations on the set of interface components to generate a layout scheme. Based on the layout scheme, an interactive interface is rendered and generated.

6. The method according to claim 5, characterized in that, Before matching the corresponding component template from the atomic component library, the method further includes: Based on the interface description specifications, historical interface templates are retrieved from the interface database. If a matching historical interface template is retrieved from the interface database, an interactive interface is generated based on the historical interface template. If no matching historical interface template is found in the interface database, the step of matching the corresponding component template from the atomic component library using the component selector is executed.

7. The method according to claim 5, characterized in that, The method further includes: Collect user interaction logs with the interactive interface; Based on the interaction logs, a reward signal is calculated, and the policy network parameters of the layout optimizer and the component preference parameters of the component selector are updated based on the reward signal.

8. The method according to claim 2, characterized in that, The method further includes: The interface generation data corresponding to the interactive interface is stored in the memory database as interface generation experience, and the memory weights corresponding to each interface generation experience in the memory database are updated.

9. The method according to claim 7, characterized in that, The method further includes: The updated policy network parameters of the layout optimizer and the component preference parameters of the component selector are uploaded to the federated aggregator for aggregation to obtain a global model update, which is then distributed to each vehicle locally. Receive and apply the global model update to the local model.

10. The method according to claim 1, characterized in that, The step of performing semantic analysis on the multimodal data to determine user intent includes: The acquired multimodal data is subjected to feature fusion processing to obtain multimodal feature vectors; Obtain user intent-related data from historical records, including at least historical conversations and user preferences; The multimodal feature vectors are concatenated with user intent-related data to obtain feature enhancement information; Semantic analysis is performed on the enhanced feature information to determine the user's intent.

11. An interactive interface generation system, characterized in that, The system includes an intent determination module, an agent module, and an interface generation module, wherein: The intent determination module is used to acquire the user's multimodal data input and perform semantic analysis on the multimodal data to determine the user's intent; The intelligent agent module is used to acquire historical interface generation data related to the user intent; based on the user intent, the historical interface generation data and the current vehicle context information, generate an interactive interface display plan, the interactive interface display plan including at least an interface display layout scheme and at least one in-vehicle service to be invoked; invoke the in-vehicle service to execute, and receive service data packets returned by each in-vehicle service; The interface generation module is used to render and generate an interactive interface based on the interactive interface display plan and the service data package.

12. A vehicle, characterized in that, The vehicle includes an interactive interface generation system, the system including a controller, the controller including a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the interactive interface generation method of any one of claims 1 to 10.

Images