Artificial intelligence-based interface interaction preloading method and system

Abstract

This invention relates to the field of artificial intelligence technology, and discloses an AI-based interface interaction preloading method and system. The method includes: collecting operation commands executed by the user on the maintenance and diagnosis interface to generate a user operation path sequence; performing path matching analysis between the user operation path sequence and the diagnostic process knowledge graph to obtain a candidate path set; selecting the downstream node with the highest path weight value in the candidate path set as the target node, and performing scenario-based preloading on the target interface of the target node to obtain a preloaded scenario package; monitoring whether an interface switching command to switch to the target interface is issued, and when the command is detected, calling the preloaded interface instance to the maintenance and diagnosis interface, and simultaneously activating the operation context instance; and interactively binding the data items of the auxiliary dataset with the interactive elements on the target interface to construct the interactive target interface. This invention can improve the responsiveness and operational consistency of interface switching in the maintenance and diagnosis system.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method and system for preloading user interface interactions based on artificial intelligence. Background Technology

[0002] In existing technologies, interface preloading methods in computer-aided equipment repair and diagnostic systems typically rely on static rules or user historical click frequencies for resource preloading. The system preloads all next-level interfaces pointed to by hyperlinks in the current interface, or preloads several frequently accessed interfaces based on global statistics. These methods lack dynamic awareness of the user's current operation path and a deep understanding of the internal logic of the diagnostic process, making it difficult to accurately predict the user's next operational intent in complex, non-linear repair processes, resulting in a low preloading hit rate. Low-hit-rate preloading not only fails to effectively shorten user waiting time but also causes ineffective use of system memory and processor resources due to loading large amounts of irrelevant interface data. On resource-constrained repair terminal equipment, it may even cause system response delays, affecting the normal progress of repair and diagnostic work.

[0003] In existing technologies, when a user switches to the prediction interface, the system still needs to temporarily query and load auxiliary data such as fault code tables, historical maintenance records, and technical announcements required by the interface from the maintenance data warehouse. The association between the data and interface controls can only be gradually established after the interface is displayed. This method makes it difficult to synchronize the visual presentation of the interface with the readiness of the background data. As a result, although the user sees the target interface, they cannot immediately perform operations such as querying or inputting data, and still need to wait for subsequent data loading and environment initialization, resulting in limited improvement in interaction smoothness. Therefore, there is an urgent need to develop an interface interaction preloading method that can deeply understand the diagnostic process based on the user's operation path and achieve collaborative pre-construction of interface resources and the operating environment. This would solve the technical problems of low preloading hit rate, poor resource utilization, and the need to wait for data preparation after the user reaches the target interface, thereby improving the continuity and response efficiency of human-computer interaction in the maintenance diagnostic system. Summary of the Invention

[0004] This invention provides an AI-based interface interaction preloading method and system to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides an AI-based interface interaction preloading method, comprising:

[0006] Collect the operation commands executed by the user on the maintenance and diagnostic interface, and generate the user operation path sequence;

[0007] The user operation path sequence is matched with the pre-stored diagnostic process knowledge graph to obtain a candidate path set.

[0008] The downstream node with the highest path weight value in the candidate path set is taken as the target node, and the target interface of the target node is preloaded in a scene-based manner to obtain the preloaded scene package of the target interface. The preloaded scene package includes a preloaded interface instance and an operation context instance.

[0009] Monitor whether the user issues an interface switching command to switch to the target interface. When the interface switching command is detected, call the preloaded interface instance to the maintenance and diagnostic interface, and simultaneously activate the operation context instance.

[0010] The data items of the auxiliary dataset in the target interface are interactively bound to the interactive elements on the target interface to construct an interactive target interface.

[0011] In a preferred embodiment, the step of collecting the operation commands executed by the user on the maintenance diagnostic interface and generating a user operation path sequence includes:

[0012] Listen for interactive events on the maintenance diagnostic interface, capture the user's triggering operations on the maintenance diagnostic interface, and generate an original operation record. The original operation record includes the control identifier of the triggered control and the trigger timestamp.

[0013] The original operation records are sorted according to the order of the trigger timestamps to obtain an operation record list;

[0014] Extract the interface node identifier corresponding to the control identifier from the operation record list;

[0015] The extracted interface node identifiers are concatenated according to the order of the trigger timestamps to generate a user operation path sequence.

[0016] In a preferred embodiment, the step of performing path matching analysis between the user operation path sequence and a pre-stored diagnostic process knowledge graph to obtain a candidate path set includes:

[0017] Using the user operation path sequence as query input, a path matching retrieval is performed in the pre-stored diagnostic process knowledge graph to locate the path segments in the diagnostic process knowledge graph that have the same node connection relationship as the user operation path sequence, and the path segments are used as the matching path segment set.

[0018] The pre-selected path segment whose path endpoint is completely consistent with the last node identifier of the user operation path sequence is taken as the target path segment.

[0019] Identify the complete diagnostic process tree in which the target path segment is located within the diagnostic process knowledge graph, and determine the current branch position of the target path segment within the complete diagnostic process tree;

[0020] Based on the current branch position, the downstream branch path is extracted from the complete diagnostic process tree;

[0021] The path weight value is calculated based on the path feature matching degree between the downstream branch path and the user's historical operation trend in the user operation path sequence.

[0022] Based on the path weight values, the user operation path sequence is matched with the pre-stored diagnostic process knowledge graph to obtain a set of candidate paths.

[0023] In a preferred embodiment, the path weight value includes:

[0024] Extract the node attributes of the last consecutive nodes in the user operation path sequence to generate a historical transfer feature vector. The historical transfer feature vector includes the operation type, operation object category, and operation result status between the last consecutive nodes.

[0025] Extract the node attributes of the starting node in the downstream branch path to generate candidate transition feature vectors;

[0026] Calculate the feature similarity between the historical transfer feature vector and the candidate transfer feature vector, and use the feature similarity as the basis for calculating the path weight value in the downstream branch path.

[0027] In a preferred embodiment, the formula for calculating the path weight value is:

[0028]

[0029] in, This refers to the path weight value of the downstream branch path. The first feature influence parameter is preset. The second feature influence parameter, The third characteristic influence parameter, This refers to the number of nodes in the downstream branch path that match historical trends. The total number of nodes contained in the downstream branch path. The current dwell time of the user on the endpoint node. The historical average dwell time of the user on the endpoint node. This represents the number of intersections between the set of path node attributes and the set of historical node attributes. The total number of elements in the node attribute set of the downstream branch path.

[0030] In a preferred embodiment, the downstream node with the highest path weight value in the candidate path set is selected as the target node, and the target interface of the target node is preloaded in a scenario-based manner to obtain a preloaded scenario package for the target interface. The preloaded scenario package includes a preloaded interface instance and an operation context instance, including:

[0031] Based on the node identifier of the target node, retrieve and call the interface layout file, interface style file and interface script file of the target interface from the interface resource database to generate the interface rendering file set of the target interface.

[0032] Load the interface layout file, the interface style file, and the interface script file from the interface rendering file set into the display buffer;

[0033] The rendering engine in the display buffer is used to parse and preprocess the set of interface rendering files to generate a preloaded interface instance.

[0034] Based on the node attribute information of the target node, diagnostic auxiliary data is extracted from the maintenance data warehouse of the target node. The diagnostic auxiliary data includes the fault code mapping table of the target interface, historical maintenance record summary, technical bulletin document, and spare parts inventory status information.

[0035] The diagnostic auxiliary data is loaded into the operation context storage area of ​​the target node to generate an operation context instance;

[0036] Establish a reference relationship between the data items in the operation context instance and the corresponding interactive elements in the preloaded interface instance, and encapsulate the preloaded interface instance and the operation context instance with the established interactive element reference relationship to obtain the preloaded scene package of the target interface.

[0037] In a preferred embodiment, establishing the reference relationship between the data items in the operation context instance and the corresponding interactive elements in the preloaded interface instance includes:

[0038] Parse the interface structure tree of the preloaded interface instance and identify the data binding interaction elements in the interface structure tree;

[0039] Obtain the data item identifier of the diagnostic auxiliary data, and match the data binding interaction element with the data item identifier according to the binding rules in the preloaded interface instance to construct a mapping relationship table between the data binding interaction element and the corresponding data item;

[0040] The mapping table is stored in the operation context instance. When the preloaded interface instance is activated and displayed, the data binding interactive element obtains the data item content from the operation context instance according to the mapping table.

[0041] In a preferred embodiment, the step of monitoring whether the user issues an interface switching command to the target interface, and when the interface switching command is detected, calling the preloaded interface instance to the maintenance diagnostic interface, and simultaneously activating the operation context instance, includes:

[0042] When the instruction listening hook of the maintenance diagnostic interface detects that the navigation control in the target interface is triggered, a switching trigger signal is generated, and the switching trigger signal contains the interface identifier of the target interface.

[0043] The switching trigger signal is associated with the preloaded interface instance for verification to confirm that the interface identifier of the preloaded interface instance is consistent with the interface identifier in the switching trigger signal.

[0044] After verification, an interface switching instruction is generated and sent to the interface management service of the target interface to trigger the display call of the preloaded interface instance.

[0045] In a preferred embodiment, the step of interactively binding data items in the auxiliary dataset of the target interface with interactive elements on the target interface to construct an interactive target interface includes:

[0046] When the preloaded interface instance is invoked to the maintenance and diagnostic interface for display, the operation context instance is activated;

[0047] Based on the mapping table of the operation context instance, the correspondence between data binding interaction elements and corresponding data item identifiers in the mapping table is parsed.

[0048] Based on the correspondence, the data items in the auxiliary dataset are interactively bound to the interactive elements on the target interface to construct the interactive target interface.

[0049] To address the above problems, the present invention also provides an AI-based interface interaction preloading system, the system comprising:

[0050] The operation path acquisition module is used to collect the operation commands executed by the user on the maintenance and diagnostic interface and generate the user operation path sequence;

[0051] The path matching analysis module is used to perform path matching analysis between the user operation path sequence and the pre-stored diagnostic process knowledge graph to obtain a set of candidate paths.

[0052] The scene preloading module is used to take the downstream node with the highest path weight value in the candidate path set as the target node, and perform scene-based preloading on the target interface of the target node to obtain a preloaded scene package of the target interface. The preloaded scene package includes a preloaded interface instance and an operation context instance.

[0053] The instruction monitoring module is used to monitor whether the user issues an interface switching instruction to switch to the target interface. When the interface switching instruction is detected, the preloaded interface instance is called to the maintenance and diagnostic interface, and the operation context instance is activated at the same time.

[0054] The interaction construction module is used to interactively bind data items in the auxiliary dataset of the target interface with interactive elements on the target interface in order to construct an interactive target interface.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] 1. This invention achieves accurate prediction of user operation intentions by constructing a diagnostic process knowledge graph and introducing a dynamic weight calculation mechanism based on path feature matching. By performing structured matching analysis between the user's operation path sequence and the pre-stored diagnostic process knowledge graph, the precise location coordinates of the user in the current diagnostic process tree are identified. Path weight values ​​are calculated based on the feature similarity between historical transition feature vectors and candidate transition feature vectors, thereby dynamically selecting the downstream node that best matches the user's operation trend as the pre-loading target. This prediction strategy based on path logic and knowledge graph effectively avoids blind loading of irrelevant interfaces, significantly improves the pre-loading hit rate, and achieves efficient utilization of system memory and processor resources on resource-constrained maintenance terminal equipment.

[0057] 2. This invention achieves collaborative pre-construction of interface visual resources and the background operating environment by constructing a pre-loaded scene package containing pre-loaded interface instances and operation context instances. During the pre-loading stage, not only are interface layout files, style files, and script files called from the interface resource database to generate pre-loaded interface instances, but diagnostic auxiliary data such as fault code mapping tables, historical maintenance records, technical bulletin documents, and spare parts inventory status are also extracted from the maintenance data warehouse based on the node attribute information of the target nodes to generate operation context instances. Furthermore, reference relationships between data items and interface interaction elements are pre-established. When the user switches to the target interface, the system can directly call the pre-loaded interface instance to the display buffer, simultaneously activating the operation context instance and completing the interaction binding. This achieves a synchronized effect of instant interface presentation and immediate data availability, completely eliminating the waiting time after the user reaches the target interface and greatly improving the continuity of human-computer interaction and operational response efficiency during maintenance diagnosis. Attached Figure Description

[0058] Figure 1 A flowchart illustrating an AI-based interface interaction preloading method according to an embodiment of the present invention;

[0059] Figure 2 A functional block diagram of an AI-based interface interaction preloading system provided in an embodiment of the present invention;

[0060] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0061] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0062] This application provides an AI-based interface interaction preloading method. The execution subject of the AI-based interface interaction preloading method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the AI-based interface interaction preloading method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0063] Reference Figure 1 The diagram shown is a flowchart illustrating an AI-based interface interaction preloading method according to an embodiment of the present invention. In this embodiment, the AI-based interface interaction preloading method includes:

[0064] Collect the operation commands executed by the user on the maintenance and diagnostic interface, and generate the user operation path sequence;

[0065] In this embodiment of the invention, the step of collecting the operation instructions executed by the user on the maintenance and diagnostic interface and generating a user operation path sequence includes:

[0066] Listen for interactive events on the maintenance diagnostic interface, capture the user's triggering operations on the maintenance diagnostic interface, and generate an original operation record. The original operation record includes the control identifier of the triggered control and the trigger timestamp.

[0067] The original operation records are sorted according to the order of the trigger timestamps to obtain an operation record list;

[0068] Extract the interface node identifier corresponding to the control identifier from the operation record list;

[0069] The extracted interface node identifiers are concatenated according to the order of the trigger timestamps to generate a user operation path sequence.

[0070] Specifically, when maintenance personnel operate on the maintenance diagnostic interface, the interactive event listening service started in the system background monitors all user interactions occurring on the interface in real time. This monitoring is implemented through global hooks registered at the operating system level or event listeners mounted in the front-end framework. These listeners capture user actions triggered by various controls on the interface, including but not limited to mouse clicks, keyboard input, and touchscreen gestures. Each time a trigger is captured, the system immediately generates an original operation record, which is stored in the form of structured data. The control identifier is a unique identifier assigned by the system to each interactive control on the interface, such as a button ID or a menu item path code; the trigger timestamp is the system time recording the moment the operation occurred, accurate to the millisecond level, used to determine the order of operations later.

[0071] Specifically, because users may rapidly and continuously trigger multiple controls during operation, the generated original operation records may not be strictly arranged in chronological order when stored. Therefore, the system calls a sorting algorithm, using the trigger timestamp in each original operation record as the sorting key, to sort all original operation records in ascending order. After sorting, a list of operation records arranged in chronological order is obtained, where each record corresponds to a user operation, and the index order of the list accurately reflects the order in which the user operations occurred. This sorting action ensures that the subsequently generated path sequence can accurately reconstruct the user's operation trajectory during the maintenance and diagnosis process.

[0072] Specifically, each record in the operation log list contains only the control identifier of the triggered control. This control identifier is an interface-level identity code and cannot be directly used to represent a step node in the diagnostic process. Therefore, the system needs to perform a mapping conversion: a control-node mapping table is pre-maintained in the system, recording the correspondence between each control identifier and its corresponding interface node identifier in the diagnostic process knowledge graph. The system iterates through the operation log list, and for each record in the list, queries the mapping table based on its control identifier to extract the corresponding interface node identifier. The interface node identifier is a predefined node number in the diagnostic process knowledge graph that represents an independent diagnostic step.

[0073] Specifically, after converting all control identifiers to interface node identifiers, the system obtains a list of interface node identifiers that perfectly corresponds to the order of the original operation record list. Subsequently, the system performs a concatenation operation: according to the order of the trigger timestamps, these interface node identifiers are sequentially connected to form an ordered sequence. This sequence can be a string connected by a specific delimiter or an array structure. The final generated user operation path sequence completely records all interface nodes traversed by the user from the moment they enter the current diagnostic task and their access order.

[0074] In summary, the above steps achieve refined capture and structured transformation of user actions on the maintenance diagnostic interface. Specifically, by listening to interaction events and generating raw operation records containing control identifiers and trigger timestamps, it ensures that every user operation is recorded completely and accurately, without losing any details that might affect intent judgment. By sorting the raw operation records based on trigger timestamps, it eliminates the problem of record order disorder that may occur in multi-threaded or asynchronous environments, ensuring the authenticity and timing accuracy of the operation path. By mapping control identifiers to interface node identifiers with process semantics, it completes semantic alignment from the interface layer to the knowledge graph layer, enabling the generated user operation path sequence to be directly understood and processed by the diagnostic process knowledge graph. By concatenating the interface node identifiers in chronological order, a structured and computable operation path sequence is finally generated. This series of actions lays a high-quality data foundation for subsequent path matching analysis based on the knowledge graph, enabling the system to accurately understand the user's precise position in the current diagnostic process, thereby significantly improving the accuracy and reliability of downstream node prediction. Meanwhile, this data collection process does not rely on any preset rules or statistical models, and can adapt to various complex and nonlinear maintenance operation scenarios, exhibiting wide applicability and robustness.

[0075] The user operation path sequence is matched with the pre-stored diagnostic process knowledge graph to obtain a candidate path set.

[0076] In this embodiment of the invention, the step of performing path matching analysis between the user operation path sequence and a pre-stored diagnostic process knowledge graph to obtain a candidate path set includes:

[0077] Using the user operation path sequence as query input, a path matching retrieval is performed in the pre-stored diagnostic process knowledge graph to locate the path segments in the diagnostic process knowledge graph that have the same node connection relationship as the user operation path sequence, and the path segments are used as the matching path segment set.

[0078] The pre-selected path segment whose path endpoint is completely consistent with the last node identifier of the user operation path sequence is taken as the target path segment.

[0079] Identify the complete diagnostic process tree in which the target path segment is located within the diagnostic process knowledge graph, and determine the current branch position of the target path segment within the complete diagnostic process tree;

[0080] Based on the current branch position, the downstream branch path is extracted from the complete diagnostic process tree;

[0081] The path weight value is calculated based on the path feature matching degree between the downstream branch path and the user's historical operation trend in the user operation path sequence.

[0082] Based on the path weight values, the user operation path sequence is matched with the pre-stored diagnostic process knowledge graph to obtain a set of candidate paths.

[0083] The path weight values ​​include:

[0084] Extract the node attributes of the last consecutive nodes in the user operation path sequence to generate a historical transfer feature vector. The historical transfer feature vector includes the operation type, operation object category, and operation result status between the last consecutive nodes.

[0085] Extract the node attributes of the starting node in the downstream branch path to generate candidate transition feature vectors;

[0086] Calculate the feature similarity between the historical transfer feature vector and the candidate transfer feature vector, and use the feature similarity as the basis for calculating the path weight value in the downstream branch path.

[0087] The formula for calculating the path weight value is:

[0088]

[0089] in, This refers to the path weight value of the downstream branch path. The first feature influence parameter is preset. The second feature influence parameter, The third characteristic influence parameter, This refers to the number of nodes in the downstream branch path that match historical trends. The total number of nodes contained in the downstream branch path. The current dwell time of the user on the endpoint node. The historical average dwell time of the user on the endpoint node. This represents the number of intersections between the set of path node attributes and the set of historical node attributes. The total number of elements in the node attribute set of the downstream branch path.

[0090] The system launches a graph database query engine, using the interface node identifiers and their connection order contained in the user's operation path sequence as search criteria. It then traverses all stored paths in the diagnostic process knowledge graph. The diagnostic process knowledge graph is a knowledge base organized in a graph structure, where nodes represent diagnostic steps and edges represent transition relationships between steps. The path matching retrieval process searches the graph database for continuous path segments with identical node connection relationships to the input sequence; that is, the node sequence and edge directions are completely consistent. After the retrieval is complete, the system collects all the path segments that meet the criteria, forming a matching path segment set. Each path segment in this set is a sub-path in the knowledge graph that perfectly matches the user's operation history.

[0091] The system examines each path segment in the matching path segment set one by one, extracting the identifier of the last node in each path segment. Simultaneously, the system obtains the identifier of the last interface node in the user's operation path sequence. Only when the endpoint node identifier of a path segment is exactly the same as the endpoint node identifier of the user's operation path sequence is the path segment selected. Through this filtering process, the system obtains one or more target path segments that not only perfectly match the user's historical operation paths but also whose endpoints are precisely the node the user is currently at, ensuring the accuracy of the starting point for subsequent analysis.

[0092] The diagnostic process knowledge graph consists of multiple complete diagnostic process trees, each representing a standard diagnostic process for a type of fault. The system locates the complete diagnostic process tree to which a target path segment belongs based on the attributes of its starting node or key nodes within the path. Then, within the complete diagnostic process tree, the system identifies the subtree region occupied by the target path segment and determines the specific coordinates of the segment's ending node within the entire tree, including the node's process stage, branch level, and depth within that branch. The current branch position is a structured description that precisely pinpoints the user's current location within the standard diagnostic process.

[0093] The system starts from the current branch position and traverses downstream along the diagnostic process tree. Downstream branch paths refer to all possible paths extending outwards from the current node along the tree structure. These paths may be single linear paths or split into multiple parallel paths at branch nodes. The system extracts all possible downstream branch paths; each downstream branch path is a complete pathway composed of a series of continuous nodes and edges, representing various diagnostic approaches the user might take next.

[0094] The system analyzes the user's historical operation records prior to the current diagnostic task, extracting the user's operational preferences in different diagnostic scenarios to form historical operation trend characteristics. For each downstream branch path, the system analyzes the characteristics of the starting segment, including the path's direction, node attribute types, and transition conditions. The system compares these characteristics with the user's historical operation trend characteristics to determine the degree of consistency between the downstream branch path and the user's consistent operating habits. The higher the consistency of path characteristics, the more the downstream branch path aligns with the user's personalized operational preferences. Based on this consistency, the system assigns a path weight value to each downstream branch path, which quantifies the likelihood of the user selecting that path.

[0095] The system appends the calculated path weights to each downstream branch path, forming a list of candidate paths with weight labels. This list is the candidate path set, and each path in the set contains complete node sequence information and its corresponding path weight value. The candidate path set is arranged in descending order of weight value; the path with the highest weight value represents the downstream direction that best matches the user's historical operation trend, while the path with the lower weight value represents the less likely direction.

[0096] In summary, the above steps achieve a deep understanding and high-precision prediction of user operation intentions. Using the user's operation path sequence as query input for retrieval within the diagnostic process knowledge graph accurately locates the user's current position within the diagnostic process, avoiding the limitations of traditional methods that rely solely on isolated nodes. By filtering for target path segments that perfectly match the last node identifier of the user's operation path sequence, the starting point of the analysis is perfectly aligned with the user's current state, eliminating prediction errors caused by positional offsets. Identifying the current branch position of the target path segment within the complete diagnostic process tree allows the system to understand the user's current process stage from a global perspective, providing a structured context for subsequent path prediction.

[0097] Specifically, after the aforementioned steps generate a set of candidate paths, this step will further refine the method for determining the path weight values. By comparing and analyzing the attribute characteristics of the last consecutive nodes in the user operation path sequence with the starting nodes in the downstream branch paths, each downstream branch path will be assigned a weight that reflects its degree of conformity with the user's operating habits.

[0098] Further, the node attributes of the last consecutive nodes in the user operation path sequence are extracted to generate a historical transition feature vector. The system first identifies the consecutive nodes at the end of the user operation path sequence. These consecutive nodes represent the diagnostic steps continuously executed by the user in a recent period. The last consecutive nodes can be the last two or the last three nodes of the sequence, the specific number determined by the system's preset sliding window length. For each last consecutive node, the system reads the attribute information stored in that node from the diagnostic process knowledge graph. The node attributes include the operation type, operation object category, and operation result status corresponding to the node. The operation type refers to the nature of the operation performed by the user in this diagnostic step, such as a detection operation, parameter setting operation, or result confirmation operation; the operation object category refers to the type of equipment component or fault phenomenon targeted by this step, such as a power module, communication interface, or sensor unit; the operation result status refers to the diagnostic result or system feedback status obtained after executing this step, such as detection passed, detection abnormal, or parameter out of range. The system arranges and combines these attribute information according to the order of the nodes in the sequence to form a multi-dimensional historical transfer feature vector. This vector is stored in memory in the form of structured data and is used to characterize the user's recent operation transfer pattern in the current diagnostic task.

[0099] Furthermore, the system extracts the node attributes of the starting nodes in the downstream branch paths to generate candidate transition feature vectors. For each downstream branch path originating from the current branch position, the system identifies the first node of that path as the starting node. The starting node is the first diagnostic step the user will enter if they select this downstream branch path. The system reads the node attributes of this starting node from the diagnostic process knowledge graph. These node attributes also include the operation type, operation object category, and operation result status corresponding to the node. Corresponding to the extraction method of historical transition feature vectors, the system organizes these attribute information into candidate transition feature vectors with the same structure and dimensions as the historical transition feature vectors. This candidate transition feature vector reflects the characteristic pattern of the first step the user will face if they continue the diagnosis along this downstream branch path.

[0100] Furthermore, the feature similarity between the historical transfer feature vector and the candidate transfer feature vector is calculated, and this feature similarity is used as the basis for calculating the path weight value in the downstream branch path. The system compares the historical transfer feature vector and the candidate transfer feature vector item by item. The comparison method is to check whether the attribute values ​​of corresponding dimensions in the two vectors are consistent or similar.

[0101] For example, if the operation type in the historical transfer feature vector is a detection operation, and the operation type in the candidate transfer feature vector is also a detection operation, then it is considered a match for that dimension; if the operation objects are of the same category, it is also considered a match; if the operation result states are related, it is also counted as a match. The system counts the number of successful matches in all dimensions and divides the number of successful matches by the total number of dimensions to obtain a ratio between 0 and 1, which is the feature similarity. The higher the feature similarity, the more closely the starting node features of the downstream branch path match the user's recent operation transfer features, and the greater the likelihood that the user will choose that path. The system records this feature similarity as a key calculation basis for subsequently assigning path weight values ​​to downstream branch paths. This completes the basic data preparation for path weight values.

[0102] The path weight value of the downstream branch path is a dimensionless real number used to characterize the comprehensive matching degree between the downstream branch path and the user's operating habits and current operating status.

[0103] The first feature influence parameter is a preset constant coefficient used to adjust the contribution ratio of the first item in the path weight value calculation. The value of the first feature influence parameter is preset by the system according to the characteristics of the diagnostic scenario. For example, in scenarios that emphasize the importance of path structure matching, a larger value can be set, so that the first item has a greater impact on the final weight value.

[0104] This is the second feature influence parameter, used to adjust the contribution ratio of the second item in the path weight value calculation. The setting of the second feature influence parameter reflects the importance of the time factor in path selection prediction and can be personalized according to the operation rhythm characteristics of different user groups.

[0105] This is the third feature influence parameter, used to adjust the contribution ratio of the third term in the path weight value calculation. The design of the third feature influence parameter reflects the importance of node attribute feature overlap in path prediction.

[0106] To determine the number of nodes in the downstream branch path that match historical trends, the system analyzes the user's historical operation records prior to the current diagnostic task, extracting node types or node sequence patterns frequently accessed by the user in similar diagnostic scenarios, forming a historical operation trend feature library. Then, for each downstream branch path, the system checks each node in the path one by one, determining whether the node's attributes match the node patterns in the historical operation trend feature library. The matching criteria include whether the operation type corresponding to the node belongs to the user's preferred operation category, whether the operation object corresponding to the node is consistent with the device component frequently handled by the user, and whether the node's operation result status matches common statuses in the user's historical diagnoses. The number of nodes in the entire downstream branch path that meet the matching conditions is then counted.

[0107] This refers to the total number of nodes contained in the downstream branch path. This parameter is obtained directly from the downstream branch path itself by counting the nodes in the path. The total number of nodes reflects the length of the downstream branch path, i.e., how many diagnostic steps the user would need to perform if they continued the diagnosis along this path.

[0108] The current dwell time of the user at the final node refers to the last node in the user's operation path sequence, i.e., the diagnostic step the user is currently in. The current dwell time is measured as the time interval from when the user enters the diagnostic interface corresponding to that node to the current moment. The system records the timestamp of each user entering each interface and calculates the difference between the timestamp and the current time to obtain the dwell time at the current node. The dwell time reflects the user's thinking time, operational complexity, or degree of difficulty encountered in the current diagnostic step.

[0109] The historical average dwell time of the user on the endpoint node is a baseline value derived from statistical analysis of the user's historical operation data. The system collects all dwell time records of the user on this type of node in multiple past diagnostic tasks, calculates the arithmetic mean of these records, and obtains the historical average dwell time.

[0110] To determine the intersection of the path node attribute set and the historical node attribute set, the system first establishes a node attribute set for all nodes in the downstream branch path. This set contains various attributes for each node in the path, such as operation type, operation object category, and operation result status. Simultaneously, the system establishes a historical node attribute set for all nodes involved in the user's historical operations. This set contains attribute information for all nodes accessed by the user in previous diagnostic tasks. Then, the system calculates the intersection of these two sets, identifying attribute items that appear in both the downstream branch path node attributes and the user's historical node attributes, and counts the number of these shared attribute items to obtain the preloaded scene package. .

[0111] The total number of node attribute set elements in the downstream branch path is obtained by deduplicating the attributes of all nodes in the downstream branch path, that is, calculating the total number of different attribute items involved in the path.

[0112] The downstream node with the highest path weight value in the candidate path set is taken as the target node, and the target interface of the target node is preloaded in a scene-based manner to obtain the preloaded scene package of the target interface. The preloaded scene package includes a preloaded interface instance and an operation context instance.

[0113] In summary, by extracting the node attributes of the last consecutive nodes in the user's operation path sequence, the system generates a historical transfer feature vector that reflects the user's recent operation transfer patterns. By extracting the node attributes of the starting nodes in downstream branch paths, the system generates candidate transfer feature vectors representing the entry features of each downstream path. By calculating the feature similarity between the historical transfer feature vector and the candidate transfer feature vector, the system obtains a numerical basis for quantifying the degree of conformity between downstream branch paths and user operation habits. This similarity will be directly used to determine the path weight value.

[0114] In this embodiment of the invention, the downstream node with the highest path weight value in the candidate path set is taken as the target node, and the target interface of the target node is preloaded in a scene-based manner to obtain a preloaded scene package of the target interface. The preloaded scene package includes a preloaded interface instance and an operation context instance, including:

[0115] Based on the node identifier of the target node, retrieve and call the interface layout file, interface style file and interface script file of the target interface from the interface resource database to generate the interface rendering file set of the target interface.

[0116] Load the interface layout file, the interface style file, and the interface script file from the interface rendering file set into the display buffer;

[0117] The rendering engine in the display buffer is used to parse and preprocess the set of interface rendering files to generate a preloaded interface instance.

[0118] Based on the node attribute information of the target node, diagnostic auxiliary data is extracted from the maintenance data warehouse of the target node. The diagnostic auxiliary data includes the fault code mapping table of the target interface, historical maintenance record summary, technical bulletin document, and spare parts inventory status information.

[0119] The diagnostic auxiliary data is loaded into the operation context storage area of ​​the target node to generate an operation context instance;

[0120] Establish a reference relationship between the data items in the operation context instance and the corresponding interactive elements in the preloaded interface instance, and encapsulate the preloaded interface instance and the operation context instance with the established interactive element reference relationship to obtain the preloaded scene package of the target interface.

[0121] The step of establishing the reference relationship between the data items in the operation context instance and the corresponding interactive elements in the preloaded interface instance includes:

[0122] Parse the interface structure tree of the preloaded interface instance and identify the data binding interaction elements in the interface structure tree;

[0123] Obtain the data item identifier of the diagnostic auxiliary data, and match the data binding interaction element with the data item identifier according to the binding rules in the preloaded interface instance to construct a mapping relationship table between the data binding interaction element and the corresponding data item;

[0124] The mapping table is stored in the operation context instance. When the preloaded interface instance is activated and displayed, the data binding interactive element obtains the data item content from the operation context instance according to the mapping table.

[0125] Based on the node identifier of the target node, the system retrieves and calls the interface layout file, interface style file, and interface script file of the target interface from the interface resource database to generate the interface rendering file set of the target interface. The system obtains the node identifier of the target node stored in the diagnostic process knowledge graph, and this node identifier corresponds one-to-one with the interface index in the interface resource database.

[0126] The interface resource database is a dedicated storage system for all diagnostic interface-related files. The interface layout file defines the position, size, and arrangement of each control in the interface and uses a markup language to describe the structural skeleton of the interface.

[0127] The interface style file defines the visual presentation attributes of the controls, including appearance features such as color, font, border, and background; the interface script file contains the behavioral logic of interactive controls in the interface, such as the function triggered when a button is clicked, the data loading method of a drop-down menu, etc.

[0128] The system performs a precise search in the interface resource database based on the node identifier, finds the three files that uniquely correspond to the node, reads these files from the disk storage medium, and combines them to form an interface rendering file set. This file set completely contains all the source materials required to build the target interface.

[0129] The display buffer is a dedicated storage area in system memory used to temporarily store UI resources that will be processed by the rendering engine. The system writes the three files from the UI rendering file set into the display buffer as binary data streams and allocates contiguous storage space for these files within the buffer.

[0130] During loading, the system simultaneously records the starting address and data length of each file in the buffer, enabling the rendering engine to accurately locate and access these resources. The purpose of the display buffer is to isolate disk I / O operations from rendering processing, allowing the rendering engine to read interface files at memory access speeds and avoiding the impact of disk read latency on preloading efficiency.

[0131] The rendering engine is a core processing component in the system, specifically responsible for converting interface description files into displayable memory objects.

[0132] The rendering engine first reads the UI layout file from the display buffer, parses its markup structure, and creates a corresponding control object tree in memory based on the layout description. Each control object contains attributes such as its type identifier, position coordinates, and size. Next, the rendering engine reads the UI style file, applies style rules to the control object tree, and sets visual attributes such as color and font for each control object. Finally, the rendering engine loads the UI script file, associating the event handling functions defined in the script with the control objects, but does not execute these functions at this stage. After parsing and preprocessing, a complete and structured control object tree is formed in memory; this object tree is the preloaded UI instance. This instance has completed all layout calculations and style applications and possesses all the conditions for display; the only difference is that it has not yet been submitted to the screen for display.

[0133] Furthermore, based on the node attribute information of the target node, diagnostic auxiliary data is extracted from the maintenance data warehouse of the target node. The diagnostic auxiliary data includes the fault code mapping table of the target interface, historical maintenance record summary, technical bulletin document, and spare parts inventory status information.

[0134] The system then accesses the node attribute information of the target node stored in the diagnostic process knowledge graph. These attributes contain data indexes associated with the node, pointing to specific datasets in the maintenance data warehouse. The maintenance data warehouse is a database system that centrally stores all maintenance-related data.

[0135] The system performs retrieval operations based on the data index, extracting a fault code mapping table from the data warehouse. This table records the correspondence between fault symptoms and fault codes, along with a detailed explanation and handling suggestions for each fault code. It also extracts historical maintenance record summaries, which contain key information recorded during past maintenance of the same or similar equipment, including fault symptoms, diagnostic processes, replaced parts, and maintenance results. Furthermore, it extracts technical bulletin documents, which are technical notices issued by equipment manufacturers, containing solutions to known problems, software update instructions, and operational precautions. Finally, it extracts spare parts inventory status information, reflecting the currently available spare parts types, quantities, and storage locations. This diagnostic auxiliary data is read from the data warehouse in a structured format and prepared for loading into the operational context.

[0136] The operation context storage area is an independent data space in memory that the system allocates specifically for each preloaded interface, used to store all dynamic data that the interface needs to access during its operation.

[0137] The system organizes the extracted fault code mapping table, historical maintenance record summaries, technical bulletin documents, and spare parts inventory status information according to a predefined data structure and writes them into the operation context storage area. During the writing process, the system creates an index for each data item to facilitate quick location and access later. After the writing is complete, a complete data set is formed in the operation context storage area; this set is the operation context instance.

[0138] The system traverses the control object tree within the preloaded interface instance to identify all interactive elements requiring data support. These elements include dropdown selection boxes, data tables, text input boxes, and information prompts. For each data-required interactive element, the system parses its bound data request identifier, which indicates the type of data the element needs to access. Simultaneously, the system traverses the data items in the operation context instance, obtaining the index identifier and data structure of each data item. Based on the matching relationship between the data request identifier and the data item index identifier, the system finds the corresponding data item for each interactive element and establishes a reference relationship between them. This reference relationship is established by adding a data pointer field to the interactive element's object structure, which points to the memory address of the corresponding data item in the operation context instance. After establishing the reference relationship, the interactive element knows where to retrieve data when it is activated.

[0139] The system associates and packages the preloaded UI instance and the operation context instance into a unified logical unit. This encapsulation can be achieved by adding a context pointer field to the preloaded UI instance, pointing to the operation context instance; or by creating a new encapsulation object that holds references to both the UI instance and the context instance.

[0140] After encapsulation, a bidirectional access relationship is established between the preloaded UI instance and the operation context instance. The UI instance can access data in the context instance at any time through the reference relationship, and the context instance can also notify the UI instance of data updates through backreferences. This encapsulated logical unit is the preloaded scene package, which is stored in the preload area of ​​the system memory, waiting to be activated and used. As a whole, the preloaded scene package ensures that when the user switches to the target interface, the UI visual resources and the background data environment are ready simultaneously, achieving a seamless operation experience.

[0141] In summary, based on the node identifier of the target node, the system retrieves and calls the interface layout file, interface style file, and interface script file from the interface resource database, generating a set of interface rendering files for the target interface. The interface layout file, interface style file, and interface script file from the rendering file set are loaded into the display buffer, preparing the data for subsequent rendering processing. Based on the rendering engine in the display buffer, the system parses and preprocesses the interface rendering file set, generating a preloaded interface instance, which is a memory-based interface object with completed layout calculations and style applications. Based on the node attribute information of the target node, diagnostic auxiliary data is extracted from the maintenance data warehouse, obtaining the fault code mapping table, historical maintenance record summaries, technical bulletin documents, and spare parts inventory status information required for the target interface to run. The preloaded interface instance and operation context instance, which have established interactive element reference relationships, are encapsulated to obtain a preloaded scene package for the target interface. This scene package is a complete, instantly activated operation unit containing all the resources required for interface display and data access.

[0142] The system accesses the tree structure of control objects stored in memory for the preloaded UI instance. This tree structure, starting from the root control, nests layer by layer through parent-child relationships, fully describing the hierarchical composition of the UI. The system employs a depth-first traversal algorithm, visiting each control object in the tree layer by layer, starting from the root control.

[0143] For each accessed control object, the system checks whether its property set contains data binding tags.

[0144] The system accesses the diagnostic auxiliary data stored in the operation context instance, iterates through each data item, and reads the data item identifier corresponding to each data item. The data item identifier is a unique name assigned by the system to each type of diagnostic auxiliary data.

[0145] Simultaneously, the system reads the predefined binding rules from the preloaded interface instance. These rules are stored in the interface instance's metadata as a configuration table, specifying which data item identifier each data-bound interactive element should be associated with. The binding rules can take the form of a correspondence between control types and data item types, such as specifying that all dropdown selection boxes are bound to a fault code mapping table by default; or they can be an explicit mapping between control identifiers and data item identifiers.

[0146] According to the binding rules, the system matches each control in the list of data-bound interactive elements with the data item identifier in the operation context instance, finding the corresponding data item for each control. After matching, the system creates a new mapping table, which is stored in key-value pairs. The key of each row is a unique identifier for the data-bound interactive element, which can be a path or control ID in the interface structure tree. The value of each row is the memory address or data item identifier of the corresponding data item. The mapping table completely records where each control that needs data should obtain the data from.

[0147] The system writes the completed mapping table as a data field into the storage area of ​​the operation context instance, making it a component of the operation context instance. The storage location of the mapping table within the operation context instance is fixed for quick access later. When the user subsequently issues a screen switching command and the preloaded screen instance is invoked to the display screen, the system simultaneously activates the operation context instance.

[0148] During activation, each data-bound interactive element in the preloaded interface instance performs an initialization operation, which includes looking up the corresponding data item identifier in the mapping table based on its own unique identifier, and then locating the specific data content stored in the operation context instance through the data item identifier.

[0149] Once the location is successfully established, the interactive element loads the data content into its own display buffer. For example, a dropdown selector loads its option list, a data table loads its row records, and a text input box loads its default text. After loading, the interactive element refreshes its display state, making the data content visible on the interface. At this point, a dynamic, real-time accessible connection is established between the data-bound interactive element and the data items in the operation context instance, ensuring that users can immediately obtain the required data when interacting with these controls, without needing to query the database again or wait for data to load.

[0150] In summary, by parsing the interface structure tree of the preloaded interface instance, the system identified all interactive elements requiring data support, forming a list of data-bound interactive elements. By obtaining the data item identifiers of diagnostic auxiliary data and matching them according to binding rules, the system found the corresponding data source for each data-bound interactive element and constructed a mapping table. By storing the mapping table in the operation context instance, the system solidified the association between data-bound interactive elements and data items. When the preloaded interface instance is activated and displayed, the interactive element can instantly retrieve the data item content from the operation context instance according to the mapping table, realizing on-demand access and instant presentation of data.

[0151] Monitor whether the user issues an interface switching command to switch to the target interface. When the interface switching command is detected, call the preloaded interface instance to the maintenance and diagnostic interface, and simultaneously activate the operation context instance.

[0152] In this embodiment of the invention, the step of monitoring whether the user issues an interface switching command to switch to the target interface, and when the interface switching command is detected, calling the pre-loaded interface instance to the maintenance and diagnostic interface, and simultaneously activating the operation context instance, includes:

[0153] When the instruction listening hook of the maintenance diagnostic interface detects that the navigation control in the target interface is triggered, a switching trigger signal is generated, and the switching trigger signal contains the interface identifier of the target interface.

[0154] The switching trigger signal is associated with the preloaded interface instance for verification to confirm that the interface identifier of the preloaded interface instance is consistent with the interface identifier in the switching trigger signal.

[0155] After verification, an interface switching instruction is generated and sent to the interface management service of the target interface to trigger the display call of the preloaded interface instance.

[0156] Pre-set command listener hooks on the maintenance diagnostic interface. A command listener hook is a small program fragment residing in the interface event handling layer, attached to all navigation controls that might trigger interface transitions. Navigation controls include menu items, buttons, links, and any interface element capable of causing interface navigation. The function of the command listener hook is to capture events as soon as these controls are triggered and to prevent the default event handling process, allowing the system to intervene and execute the activation operation of the pre-loaded scene package.

[0157] The triggering action can be any interaction with the navigation control, such as a mouse click, keyboard Enter key press, or touchscreen tap. After capturing the triggering event, the instruction listener hook reads the target interface information associated with the triggered navigation control. This information is usually stored in the control's properties, such as recording the target interface's identifier as a data property.

[0158] The instruction listening hook generates a switching trigger signal based on the information it reads. This signal is a structured event object containing a key field: the target interface's identifier. The interface identifier is a unique code assigned to each diagnostic interface by the system, such as the filename of the interface resource or a primary key value in a database. After generation, the switching trigger signal is sent to the system's event processing center for further verification and processing.

[0159] The system accesses the preloaded interface instance stored in the preloaded scene package, reads the instance's metadata, and extracts the corresponding interface identifier. Simultaneously, it extracts the target interface's identifier from the switching trigger signal. The system compares these two identifiers character by character to check for complete consistency. During the comparison, the system considers details such as capitalization and namespaces to ensure there is no ambiguity. If the comparison shows that the two identifiers are identical, it means that the interface instance in the preloaded scene package is indeed the target interface the user wants to switch to, and the verification passes.

[0160] The interface switching command is a higher-level event object that contains complete information about the target interface and a memory address reference of the preloaded scene package. The interface switching command is sent to the system's interface management service, the core component responsible for managing the entire lifecycle of interface display, including interface creation, display, hiding, and destruction. Upon receiving the interface switching command, the interface management service locates the position of the preloaded interface instance in the display buffer based on the memory address reference of the preloaded scene package carried in the command.

[0161] The preloaded interface instance is submitted from the display buffer to the screen frame buffer, triggering the graphics processing unit to perform the final rendering output of the interface, so that the interface content is displayed on the maintenance and diagnostic interface. Simultaneously with the display call, the interface management service performs the activation operation of the operation context instance.

[0162] Activating an operation context instance means changing the operation context instance from an inactive state to an accessible state. Specifically, this involves setting an activation flag so that data items in the operation context instance can be accessed by interactive elements in the preloaded interface instance.

[0163] Once activated, the data-bound interactive elements in the preloaded interface instance read the corresponding data items from the operation context instance based on the previously established mapping table and populate the display area of ​​the interactive elements. The entire display invocation and activation process is completed in a very short time, with almost no noticeable delay for the user.

[0164] Furthermore, after the preloaded interface instance is displayed and the operation context instance is activated, the system switches the current focus to the newly displayed interface and releases the display resources of the old interface. Simultaneously, the system maintains the activity of the command listening hook to prepare for capturing the next interface switching operation. Thus, the entire process from the user triggering the navigation controls to seeing the fully operable target interface is complete; all preloaded resources are accurately activated at crucial moments, achieving seamless interface switching.

[0165] In summary, the above steps achieve precise triggering and immediate activation of the preloaded scene package. The instruction listening hook allows the system to capture the event and generate a switching trigger signal the moment the user triggers the navigation controls, avoiding the need for polling or delayed detection to determine user intent, as required by traditional solutions, thus significantly improving response speed. The association verification operation ensures that the preloaded interface instance is completely consistent with the target interface the user actually wants to switch to, eliminating display errors that may result from mismatched identifiers and guaranteeing the accuracy and reliability of diagnostic operations.

[0166] The data items of the auxiliary dataset in the target interface are interactively bound to the interactive elements on the target interface to construct an interactive target interface.

[0167] The step of interactively binding data items in the auxiliary dataset of the target interface with interactive elements on the target interface to construct an interactive target interface includes:

[0168] When the preloaded interface instance is invoked to the maintenance and diagnostic interface for display, the operation context instance is activated;

[0169] Based on the mapping table of the operation context instance, the correspondence between data binding interaction elements and corresponding data item identifiers in the mapping table is parsed.

[0170] Based on the correspondence, the data items in the auxiliary dataset are interactively bound to the interactive elements on the target interface to construct the interactive target interface.

[0171] Activating an operation context instance means changing the operation context instance from a preloaded state to an accessible state. The specific process includes setting the access flag of the operation context instance to an active state, so that the data items in the instance can be accessed externally.

[0172] Simultaneously, the system registers the memory address of the operation context instance to the activity context list of the interface management service, so that interactive elements in the preloaded interface instance can locate the instance through the context manager. After activation, the fault code mapping table, historical maintenance record summary, technical bulletin documents, and spare parts inventory status information stored in the operation context instance all enter the ready state, waiting to be invoked by interactive elements.

[0173] The mapping table is a data structure stored in key-value pair format. Each record contains two main fields: the first field is a unique identifier for the data-bound interactive element, which can be the path of the interactive element in the interface structure tree or the control identifier of the interactive element; the second field is the corresponding data item identifier, which indicates the index name of the data item that the interactive element needs to access in the operation context instance. The system reads each record in the mapping table line by line, parsing out the data item that each data-bound interactive element should be associated with. During the parsing process, the system simultaneously verifies the validity of the data item identifier, confirming that the data item with the corresponding identifier does exist in the operation context instance, ensuring that subsequent binding operations can be executed successfully.

[0174] Interaction binding is a refined process executed on a per-element basis. For dropdown selection elements, the system finds the corresponding data item identifier based on the mapping table, reads the option list data corresponding to that data item identifier from the operation context instance, and then loads the option list data into the dropdown selection's data source, enabling the dropdown selection to expand and display these options when clicked. For table-type interactive elements, the system reads the record set pointed to by the corresponding data item identifier, fills each row of data in the record set into the row container of the data table, and maps the data to the corresponding column cells according to the table column configuration.

[0175] For text input fields, the system reads the corresponding default value or tooltip text and sets it as the initial display content of the input field. For message boxes, the system reads the corresponding tooltip content and stores it in the message buffer of the message box, which is then displayed immediately when the user hovers over or clicks the tooltip icon. After each interactive element completes data loading, the system triggers the element's UI refresh method to update its display state.

[0176] The integration process includes redrawing and checking the overall interface layout to ensure no layout errors occur after data loading; binding and checking the event handlers of interactive elements to ensure user actions correctly trigger predefined functions; and initializing the focus state so that users can begin operations immediately upon entering the interface. After integration, all interactive elements in the preloaded interface instance are dynamically linked to data items in the operation context instance. Users can instantly access the corresponding data content for any interactive element's operation without any waiting or background requests.

[0177] The target interactive interface is a visually complete and data-ready operable interface, featuring the following characteristics: every interactive element is correctly displayed according to a predefined layout and style; each interactive element requiring data support has already loaded specific data from the operation context instance; the event handlers for each interactive element are ready to respond to user clicks, inputs, selections, etc.; a dynamic reference relationship is maintained between the entire interface and the operation context instance, allowing interactive elements to detect and refresh their display when data in the operation context instance is updated via a mapping table. The target interactive interface is submitted to the final rendering queue of the display system and presented on the maintenance diagnostic interface for users to perform subsequent diagnostic operations.

[0178] In summary, when the preloaded interface instance is invoked for display on the maintenance and diagnostic interface, the system activates the operation context instance, making its data items accessible. Based on the mapping table of the operation context instance, the system parses the correspondence between each data-bound interactive element and its corresponding data item identifier, providing clear guidance for subsequent data loading. Based on the parsed correspondence, the system performs interactive binding operations on the data items in the auxiliary dataset and the interactive elements on the target interface one by one, loading the data into the interactive elements and refreshing the display, ultimately constructing a visually complete, data-ready, and immediately operable interactive target interface.

[0179] like Figure 2 The diagram shown is a functional block diagram of an AI-based interface interaction preloading system provided in an embodiment of the present invention.

[0180] The AI-based interface interaction preloading system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the AI-based interface interaction preloading system 100 may include an operation path acquisition module 101, a path matching and analysis module 102, a scene preloading module 103, an instruction monitoring module 104, and an interaction construction module 105. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0181] In this embodiment, the functions of each module / unit are as follows:

[0182] The operation path acquisition module 101 is used to acquire the operation commands executed by the user on the maintenance and diagnostic interface and generate the user operation path sequence.

[0183] The path matching analysis module 102 is used to perform path matching analysis between the user operation path sequence and the pre-stored diagnostic process knowledge graph to obtain a candidate path set.

[0184] The scene preloading module 103 is used to take the downstream node with the highest path weight value in the candidate path set as the target node, and perform scene-based preloading on the target interface of the target node to obtain a preloaded scene package of the target interface. The preloaded scene package includes a preloaded interface instance and an operation context instance.

[0185] The instruction monitoring module 104 is used to monitor whether the user issues an interface switching instruction to switch to the target interface. When the interface switching instruction is detected, the preloaded interface instance is called to the maintenance and diagnostic interface, and the operation context instance is activated at the same time.

[0186] The interaction construction module 105 is used to interactively bind data items in the auxiliary dataset of the target interface with interactive elements on the target interface to construct an interactive target interface.

[0187] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0188] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0189] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0190] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0191] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0192] Finally, 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.

Claims

1. A method for preloading user interface interactions based on artificial intelligence, characterized in that, The method includes: Collect the operation commands executed by the user on the maintenance and diagnostic interface, and generate the user operation path sequence; The user operation path sequence is matched with the pre-stored diagnostic process knowledge graph to obtain a candidate path set. The downstream node with the highest path weight value in the candidate path set is taken as the target node, and the target interface of the target node is preloaded in a scene-based manner to obtain the preloaded scene package of the target interface. The preloaded scene package includes a preloaded interface instance and an operation context instance. Monitor whether the user issues an interface switching command to switch to the target interface. When the interface switching command is detected, call the preloaded interface instance to the maintenance and diagnostic interface, and simultaneously activate the operation context instance. The data items of the auxiliary dataset in the target interface are interactively bound to the interactive elements on the target interface to construct an interactive target interface.

2. The AI-based interface interaction preloading method as described in claim 1, characterized in that, The process of collecting user operation commands executed on the maintenance and diagnostic interface and generating a user operation path sequence includes: Listen for interactive events on the maintenance diagnostic interface, capture the user's triggering operations on the maintenance diagnostic interface, and generate an original operation record. The original operation record includes the control identifier of the triggered control and the trigger timestamp. The original operation records are sorted according to the order of the trigger timestamps to obtain an operation record list; Extract the interface node identifier corresponding to the control identifier from the operation record list; The extracted interface node identifiers are concatenated according to the order of the trigger timestamps to generate a user operation path sequence.

3. The AI-based interface interaction preloading method as described in claim 1, characterized in that, The step of performing path matching analysis between the user operation path sequence and a pre-stored diagnostic process knowledge graph to obtain a candidate path set includes: Using the user operation path sequence as query input, a path matching retrieval is performed in the pre-stored diagnostic process knowledge graph to locate the path segments in the diagnostic process knowledge graph that have the same node connection relationship as the user operation path sequence, and the path segments are used as the matching path segment set. The pre-selected path segment whose path endpoint is completely consistent with the last node identifier of the user operation path sequence is taken as the target path segment. Identify the complete diagnostic process tree in which the target path segment is located within the diagnostic process knowledge graph, and determine the current branch position of the target path segment within the complete diagnostic process tree; Based on the current branch position, the downstream branch path is extracted from the complete diagnostic process tree; The path weight value is calculated based on the path feature matching degree between the downstream branch path and the user's historical operation trend in the user operation path sequence. Based on the path weight values, the user operation path sequence is matched with the pre-stored diagnostic process knowledge graph to obtain a set of candidate paths.

4. The AI-based interface interaction preloading method as described in claim 3, characterized in that, The path weight values ​​include: Extract the node attributes of the last consecutive nodes in the user operation path sequence to generate a historical transfer feature vector. The historical transfer feature vector includes the operation type, operation object category, and operation result status between the last consecutive nodes. Extract the node attributes of the starting node in the downstream branch path to generate candidate transition feature vectors; Calculate the feature similarity between the historical transfer feature vector and the candidate transfer feature vector, and use the feature similarity as the basis for calculating the path weight value in the downstream branch path.

5. The AI-based interface interaction preloading method as described in claim 4, characterized in that, The formula for calculating the path weight value is: ； in, This refers to the path weight value of the downstream branch path. The first feature influence parameter is preset. The second feature influence parameter, The third characteristic influence parameter, This refers to the number of nodes in the downstream branch path that match historical trends. The total number of nodes contained in the downstream branch path. The current dwell time of the user on the endpoint node. The historical average dwell time of the user on the endpoint node. This represents the number of intersections between the set of path node attributes and the set of historical node attributes. The total number of elements in the node attribute set of the downstream branch path.

6. The AI-based interface interaction preloading method as described in claim 1, characterized in that, The downstream node with the highest path weight value in the candidate path set is selected as the target node, and the target interface of the target node is preloaded in a scenario-based manner to obtain a preloaded scenario package for the target interface. The preloaded scenario package includes a preloaded interface instance and an operation context instance, including: Based on the node identifier of the target node, retrieve and call the interface layout file, interface style file and interface script file of the target interface from the interface resource database to generate the interface rendering file set of the target interface. Load the interface layout file, the interface style file, and the interface script file from the interface rendering file set into the display buffer; The rendering engine in the display buffer is used to parse and preprocess the set of interface rendering files to generate a preloaded interface instance. Based on the node attribute information of the target node, diagnostic auxiliary data is extracted from the maintenance data warehouse of the target node. The diagnostic auxiliary data includes the fault code mapping table of the target interface, historical maintenance record summary, technical bulletin document, and spare parts inventory status information. The diagnostic auxiliary data is loaded into the operation context storage area of ​​the target node to generate an operation context instance; Establish a reference relationship between the data items in the operation context instance and the corresponding interactive elements in the preloaded interface instance, and encapsulate the preloaded interface instance and the operation context instance with the established interactive element reference relationship to obtain the preloaded scene package of the target interface.

7. The AI-based interface interaction preloading method as described in claim 6, characterized in that, The step of establishing the reference relationship between the data items in the operation context instance and the corresponding interactive elements in the preloaded interface instance includes: Parse the interface structure tree of the preloaded interface instance and identify the data binding interaction elements in the interface structure tree; Obtain the data item identifier of the diagnostic auxiliary data, and match the data binding interaction element with the data item identifier according to the binding rules in the preloaded interface instance to construct a mapping relationship table between the data binding interaction element and the corresponding data item; The mapping table is stored in the operation context instance. When the preloaded interface instance is activated and displayed, the data binding interactive element obtains the data item content from the operation context instance according to the mapping table.

8. The AI-based interface interaction preloading method as described in claim 1, characterized in that, The monitoring process involves checking whether the user issues a screen switching command to the target screen. When the screen switching command is detected, the pre-loaded screen instance is invoked to the maintenance and diagnostic screen, and the operation context instance is activated simultaneously, including: When the instruction listening hook of the maintenance diagnostic interface detects that the navigation control in the target interface is triggered, a switching trigger signal is generated, and the switching trigger signal contains the interface identifier of the target interface. The switching trigger signal is associated with the preloaded interface instance for verification to confirm that the interface identifier of the preloaded interface instance is consistent with the interface identifier in the switching trigger signal. After verification, an interface switching instruction is generated and sent to the interface management service of the target interface to trigger the display call of the preloaded interface instance.

9. The AI-based interface interaction preloading method as described in claim 1, characterized in that, The step of interactively binding data items in the auxiliary dataset of the target interface with interactive elements on the target interface to construct an interactive target interface includes: When the preloaded interface instance is invoked to the maintenance and diagnostic interface for display, the operation context instance is activated; Based on the mapping table of the operation context instance, the correspondence between data binding interaction elements and corresponding data item identifiers in the mapping table is parsed. Based on the correspondence, the data items in the auxiliary dataset are interactively bound to the interactive elements on the target interface to construct the interactive target interface.

10. An AI-based interface interaction preloading system, characterized in that, The system for implementing the AI-based interface interaction preloading method of claim 1 includes: The operation path acquisition module is used to collect the operation commands executed by the user on the maintenance and diagnostic interface and generate the user operation path sequence; The path matching analysis module is used to perform path matching analysis between the user operation path sequence and the pre-stored diagnostic process knowledge graph to obtain a set of candidate paths. The scene preloading module is used to take the downstream node with the highest path weight value in the candidate path set as the target node, and perform scene-based preloading on the target interface of the target node to obtain a preloaded scene package of the target interface. The preloaded scene package includes a preloaded interface instance and an operation context instance. The instruction monitoring module is used to monitor whether the user issues an interface switching instruction to switch to the target interface. When the interface switching instruction is detected, the preloaded interface instance is called to the maintenance and diagnostic interface, and the operation context instance is activated at the same time. The interaction construction module is used to interactively bind data items in the auxiliary dataset of the target interface with interactive elements on the target interface in order to construct an interactive target interface.

Images