Tax self-service terminal interaction control system based on distributed architecture

Abstract

The application relates to the technical field of tax service, in particular to a tax self-service terminal interaction control system based on a distributed architecture, which comprises an interaction module, a feedback prompt module and the like, wherein the interaction module is used for providing a display user operation interface supporting transaction operation of a user, collecting interaction data and constructing the interaction data into an interaction data set; the feedback prompt module is used for judging whether operation is successful based on the interaction data, showing operation results to the user, reminding next operation when the operation is successful, and prompting failure and providing a solution when the operation fails; according to the operation results of the user, the application timely shows whether the operation is successful to the user, and guides the user to carry out subsequent steps through voice prompt; when the operation fails, the application details prompts the user about error causes and solutions through voice and visual modes, helps the user to understand and avoid repeated errors, and further reduces the operation psychological burden of the user.

Description

Technical Field

[0001] This invention relates to the field of tax service technology, and more specifically to an interactive control system for tax self-service terminals based on a distributed architecture. Background Technology

[0002] In recent years, tax self-service terminals have transformed from a supplementary channel into a primary means for many taxpayers, especially younger generations, to file and pay tax returns. The widespread adoption of self-service has significantly reduced the pressure on tax service halls and improved overall tax processing efficiency. However, this trend has revealed new challenges in service coverage: elderly users with lower digital literacy often find operating self-service terminals challenging. Current self-service terminal systems often prioritize efficiency in their interaction design, neglecting the specific needs of this user group, leading to frequent obstacles during operation, especially in sensitive areas involving financial transactions.

[0003] The core aspects of tax transactions not only require a high level of financial security from the system, but also place extremely high demands on the smoothness of the user experience and emotional reassurance. For users without a technical background, the interface of self-service terminals can appear complex, and the process logic unclear. If a failure occurs in a crucial payment step, users often feel helpless due to a lack of coping knowledge. Existing systems commonly suffer from delayed feedback mechanisms and vague information prompts, making it difficult for users to understand the reasons for failure and obtain effective operational guidance, thus triggering significant anxiety and insecurity. This negative experience not only directly reduces the success rate of individual transactions but also fundamentally erodes users' trust in self-service, hindering the inclusiveness of the service.

[0004] Therefore, a tax self-service terminal interactive control system based on a distributed architecture is proposed to solve the problems mentioned above. Summary of the Invention

[0005] Technical problems to be solved In view of the above-mentioned shortcomings of the existing technology, the present invention provides a tax self-service terminal interactive control system based on a distributed architecture, which can effectively solve the problems of poor interaction design in the existing tax self-service terminal interactive control system, which leads to poor user experience and easy errors for the elderly, and the failure to provide timely and clear feedback when transactions fail, causing users to feel anxious and uneasy.

[0006] Technical solution To achieve the above objectives, the present invention provides the following technical solution: This invention provides a tax self-service terminal interactive control system based on a distributed architecture, including an interaction module for displaying a user interface to support user transactions, and for collecting interaction data and constructing an interaction dataset. The feedback module determines whether an operation is successful based on interaction data and displays the result to the user; it prompts the user for the next step when the operation is successful, and displays a failure message and provides a solution when the operation fails. The fault handling module determines the error type of the operation failure based on the interactive dataset when the operation fails, including hardware-level errors, interface-level errors and distributed-level errors, and calls up the corresponding solutions. When an operation fails due to a distributed-level error, the system processing module triggers and implements a load scheduling strategy.

[0007] Furthermore, the method for constructing the interactive dataset is as follows: A two-dimensional coordinate system is established on the user interface. The boundary points of the interface frame of the operation area in the operation interface are mapped to the two-dimensional coordinate system to form the interface element frame of the operation area. Each interface element frame is assigned a unique interface element ID and stored in the interface element mapping table. When an interaction occurs, the interaction module captures the touch point coordinates (x, y) and timestamp t, as well as the pressure d and duration of the touch. The data is then merged into a single interaction data set; all interaction data are then merged in timestamp order to construct an interaction dataset.

[0008] Furthermore, when the touch point for capturing the interactive operation is a touch area, the coordinate range is determined based on the vertex coordinates of the touch area. The mathematical expression for determining the coordinate range is: In the formula, R represents the coordinate range of the touch area. Let be the coordinates of the i-th vertex in the coordinate range, and n be the number of vertices in the coordinate range; Let be the slope of the line segment from vertex i to vertex i+1. ; Let be the intercept of the line segment from the i-th vertex to the (i+1)-th vertex. .

[0009] Furthermore, the method for determining whether the operation is successful is as follows: To determine if an operation is valid, the conditions for a valid operation are: the pressure applied is greater than the touch threshold and the duration of the pressure is within a reasonable time threshold range; if the coordinates of the touch point of the valid operation are within the interface element box of the corresponding operation area, then the operation is successful.

[0010] Furthermore, the method for determining the error type is as follows: Extract the touch point coordinates of the failed operation and compare them with the corresponding interface element box of the operation area; When the distance between the touch point coordinates of the failed operation and the boundary of the interface element box is less than the preset judgment threshold, it is judged as a hardware-level error and offset detection is triggered. When the touch point coordinates are within a coordinate range and overlap with multiple interface element boxes, it is judged as an interface-level error and the valid area detection is triggered. When a user successfully completes an operation but fails after the operation request is sent, it is judged as a distributed-level error, triggering a load balancing strategy.

[0011] Furthermore, the offset detection method is as follows: Calculate the geometric distance between the touch point coordinates of the failed operation and the center point coordinates of the interface element box, and generate an offset vector; read the first Q interaction data of the failed operation in the interaction dataset, and generate an offset vector; If the same offset angle exists in Q consecutive interactive operations and the offset amount is greater than the preset offset threshold, it is determined to be a systematic offset. If Q consecutive interactive operations have the same offset angle and the offset changes over time, it is determined to be a gradual offset. If there are different offset angles and different offset amounts in Q consecutive interactive operations, but the difference distribution conforms to a normal distribution, then it is determined to be a random offset.

[0012] Furthermore, the solution to the aforementioned hardware-level error is: Based on the determined offset type, set the processing priority; systematic offsets have a higher priority than gradual offsets, which in turn have a higher priority than random offsets. For systemic offsets, perform global touch automatic calibration, based on voice prompts, and display a calibration progress bar. Dynamic compensation is performed for gradual offset. The calibration area is displayed on the user interface, and the user is prompted by voice to place their finger on the calibration area and keep it still. A calibration progress bar is also displayed until the calibration is completed. For random offsets, the user is prompted to try the operation again via voice.

[0013] Furthermore, the effective region detection method is as follows: Calculate the coordinate range of the touch point and the area of ​​the overlapping region of each interface element box, and select the overlapping region with the largest area as the determined interface element box.

[0014] Furthermore, the solution to the interface-level error is as follows: The system provides voice prompts to users to reselect the operation area and highlights the corresponding interface element box on the user interface.

[0015] Furthermore, the load scheduling strategy is as follows: Record the type of task being processed by the current node and determine the task level based on the preset priority sorting; select the node with the highest comprehensive score among the neighboring nodes as the target node, package the detailed information of the task being processed by the current node into a request, send it to the selected target node, wait for the target node's response and processing, and during the target node's processing, report the execution status to the source node and return the real-time updated data to the source node. The method for selecting the node with the highest overall score is as follows: Calculate the current load score, historical load performance score, future load prediction score, interaction impact score, real-time response capability score, and redundancy score of all neighboring nodes in the network topology of the source node, and use a weighted average method to calculate the comprehensive score of neighboring nodes.

[0016] Beneficial effects The technical solution provided by this invention has the following advantages compared with the prior art: This invention uses a feedback module to promptly display whether the user's operation was successful based on the result of the operation, and guides the user to the next step through voice prompts; when the operation fails, it provides the user with detailed prompts on the cause of the error and solutions through voice and visual means, helping the user understand and avoid repeating the same mistake, thereby reducing the user's psychological burden of operation.

[0017] This solution also analyzes interactive data in real time through a fault handling module to identify the causes of operation failures, including hardware-level errors, interface-level errors, and distributed-level errors, and provides corresponding solutions to ensure that the system can respond quickly to user needs. The feedback mechanism reduces user confusion and anxiety, allowing users to quickly understand the cause of errors, improve their confidence in operation, and enhance service continuity. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0019] Figure 1 This is a schematic diagram of the system operation process in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0021] The present invention will be further described below with reference to embodiments.

[0022] Example: See appendix Figure 1 This case proposes a tax self-service terminal interactive control system based on a distributed architecture. Addressing the difficulties elderly users face when operating the self-service terminal, it provides intuitive operation guidance and feedback based on voice and visual feedback, helping users understand and correct errors, improving the usability of the system, and making elderly users feel more relaxed and confident when performing tax operations.

[0023] The interactive control system includes an interaction module, a feedback and prompting module, a fault handling module, a safety monitoring module, and a data management module.

[0024] The interaction module displays the user interface on the self-service terminal and supports user voice and voice input, ensuring easy browsing and transaction operations. It also records user interaction data during transactions and constructs it into an interaction dataset.

[0025] The feedback module displays the result of each interaction with the user, indicating whether the operation was successful or failed. Upon successful operation, it prompts the user with voice instructions to proceed to the next step. Upon failed operation, it alerts the user with both voice and a pop-up message to indicate the error.

[0026] The fault handling module, when an operation fails, traverses the interaction data in the interaction dataset, analyzes and determines the error type, and retrieves the corresponding solution based on the error type. This solution is displayed on the user interface via the feedback module, and simultaneously, the interaction module plays voice prompts explaining the reason for the failed operation and the solution, allowing the user to correct the operation according to the solution. By analyzing and providing feedback on user failures through the fault handling module, the system's usability and user experience are improved. Clearly indicating the specific steps leading to failure, this targeted analysis helps users understand their mistakes and avoid repeated operations. After identifying the cause of failure, the module can quickly retrieve the corresponding solution and display it directly on the user interface via the feedback module, providing users with immediate solutions when encountering problems, reducing user confusion and anxiety. The voice prompts explaining the reason for failure and the solution, played through the interaction module, make information delivery more vivid. Through the combination of voice and visual prompts, users can more clearly understand the problem and its solution, ensuring the accuracy of information delivery and reducing user frustration during operation, especially when encountering complex financial transactions.

[0027] When a user transaction fails due to a distributed error, the system processing module triggers and implements a load scheduling strategy to effectively reduce the load pressure on individual terminal nodes, reduce system failures or response delays caused by insufficient resources, thereby improving the overall stability and reliability of the system and ensuring the continuity and smoothness of user operations.

[0028] Specifically, the interactive dataset is constructed as follows: A two-dimensional coordinate system is established on the user interface of the self-service terminal, with the lower left corner of the interface as the origin, the X-axis extending from left to right, and the Y-axis extending from bottom to top. In each interface, the boundary points of the interface frame of the operating area are mapped onto the two-dimensional coordinate system to form the interface element frame of the operating area; each interface element frame is assigned a unique interface element ID; a preset interface element mapping table is used in the system to record the IDs of all interactive interface elements and their coordinate ranges in different operating interfaces.

[0029] When a user interacts with the screen by touching it, the interaction module captures the touch point coordinates (x, y) and timestamp t, as well as the pressure d and duration of the touch. These data are then merged into a single interaction dataset. During a user's continuous operations, all interaction data are merged in timestamp order to form an interaction dataset. The interaction dataset uses a spatiotemporal matrix format, such as: In the formula, Represents the timestamp of the k-th interaction operation. Touch point coordinates Pressure and press duration In this context, a user's touch on the screen may be an area, such as a pat of the entire palm, in which case the coordinate point is a coordinate range. In the formula, R represents the coordinate range of the touch area. Represents the coordinates of any point within the coordinate range; Let be the coordinates of the i-th vertex in the coordinate range, and n be the number of vertices in the coordinate range; Let be the slope of the line segment from vertex i to vertex i+1. ; Let be the intercept of the line segment from the i-th vertex to the (i+1)-th vertex. .

[0030] Each time a user performs an interaction, the system first determines whether the action is valid. The criteria for this determination are: whether the pressure applied exceeds the touch threshold, and whether the duration of the pressure is within a reasonable time threshold. Only when both criteria are met is the action recorded as valid. By judging the impact of pressure and duration on the interaction, accidental touches, such as clothing touching the screen or water dripping onto it, can be effectively filtered out.

[0031] When a user interaction is deemed valid, the system determines success based on whether the touch point coordinates fall within the corresponding interface element frame. If the touch point coordinates are within the interface element frame, the interaction is considered successful, and the feedback module plays a voice message to confirm success and prompts the user to proceed to the next step. If the interaction fails, the system notifies the user of the error via both voice and a pop-up message, specifying the error type. The method for analyzing error types is as follows: Define error types, including: hardware-level errors, which are systematic offsets that occur when the user performs multiple consecutive operations, such as all touch points being offset by a fixed distance in a certain direction; or single operation offsets exceeding a threshold and the offset direction conforms to common offset directions in historical calibration data.

[0032] A UI-level error occurs when the coordinates of a user's touch point fall within the frames of multiple UI elements simultaneously; in this case, it is considered an area overlap error in the UI.

[0033] Distributed-level error, meaning the current terminal node cannot respond to user interaction in a timely manner.

[0034] The error type is determined by extracting the touch point coordinates of the failed operation and comparing them with the UI element box of the operation area corresponding to that transaction. When the distance between the touch point coordinates of a failed operation and the boundary of the UI element box is less than a preset threshold, the offset detection procedure is triggered. Specifically: Calculate the geometric distance between the touch point coordinates of the failed operation and the center point coordinates of the interface element box, and generate an offset vector. The calculation formula is as follows: In the formula, This represents the offset between the touch point coordinates and the center point coordinates. The coordinates of the touch point where the operation failed. These are the coordinates of the center point of the UI element's frame. The offset vector includes: X-axis offset. Y-axis offset Offset angle .

[0035] Then, read the Q previous interaction data of the failed operation from the interaction dataset, calculate the geometric distance between the corresponding touch point coordinates and the center point coordinates of the interface element box, and generate the offset vector.

[0036] If the same offset angle exists in Q consecutive interactive operations, and the offset amount is greater than the preset offset threshold, it is determined to be a systematic offset.

[0037] If Q consecutive interactive operations exhibit the same offset angle, and the offset changes over time (e.g., increases linearly), it is determined that the offset is a gradual shift caused by temperature changes. Temperature changes cause the physical materials of the touchscreen to expand or contract, resulting in a slight change in the position of the touch sensor.

[0038] If Q consecutive interactive operations have different offset angles and offset amounts, but the difference distribution conforms to a normal distribution, it is determined to be a random offset. During device operation, random noise caused by electromagnetic interference, aging of electronic components, or temperature changes may lead to touch point reading errors.

[0039] Based on the identified offset type, processing priorities are set; systematic offsets have higher priority than gradual offsets, which in turn have higher priority than random offsets. In offset type identification and matching, matching is performed according to priority order; once a match is successful, matching stops, and the offset type is determined. Different processing strategies are adopted based on the offset type, including: for systematic offsets, global touch calibration is performed, and detailed logs are recorded for subsequent hardware maintenance; for gradual offsets, dynamic compensation is performed (e.g., using linear extrapolation to predict the current offset), and an early warning is issued; for random offsets, the offset is recorded, and when the frequency exceeds a set threshold, a full calibration is triggered.

[0040] Simultaneously, when the offset-type processing strategy is triggered, the feedback module will issue processing prompts to the user based on the interaction module. For systematic offsets, a voice announcement will be made: "Due to a system anomaly, your touch operation has been affected. Automatic calibration is now in progress. Please try again later," and a calibration progress bar will be displayed on the user interface. For gradual offsets, a calibration area will be displayed on the user interface, and a voice announcement will be made: "Your operation has been affected by environmental factors. Please try to keep your finger on the calibration area. The system is correcting the touch sensitivity." After the user's finger is placed on the calibration area, a calibration progress bar will be displayed. For random offsets, a voice announcement will be made: "Your touch has a detection error. The system is working to improve recognition accuracy. Please try again."

[0041] When the touch point coordinates are within a range, check if the touch point's coordinate range overlaps with multiple interface element boxes. If the touch point's coordinate range overlaps with multiple interface element boxes, calculate the area of ​​each overlapping region using the following formula: In the formula, Let m be the area of ​​the overlapping region between the j-th interface element box and the coordinate range of the touch point; m is the number of vertices in the overlapping region. Let be the coordinates of the j-th vertex in the overlapping region. Select the overlapping region with the largest area from all overlapping regions as the defined interface element box. If the selected interface element box is not the corresponding interface element box in the transaction process, a voice prompt will announce "Please reselect the correct operation area," and the corresponding interface element box will be highlighted on the user interface.

[0042] Once the user successfully completes the operation, the system resource status of each request is monitored when the user's operation request is issued. The system determines the current load of the terminal node and checks if the load exceeds the set resource threshold. If so, the system enters the resource shortage handling process, i.e., load scheduling is implemented. A voice announcement is made stating, "The system is currently under high load; please wait and refresh the page later," while the system processing progress and refresh area are displayed on the user interface.

[0043] The load scheduling strategy is implemented as follows: The system records the type of task being processed by the current node and determines the task level based on a preset priority ranking to ensure that important tasks are prioritized when load balancing strategies are implemented across multiple terminal nodes. It evaluates the status of neighboring nodes, selecting the node with the highest overall score as the target node. The system packages the detailed information of the task being processed by the current node into a request and sends it to the selected target node, awaiting its response and processing. During processing, it periodically reports the execution status back to the source node, returning real-time updated data so that the source node can display the system's processing progress on the user interface in real time.

[0044] The method for selecting the node with the highest overall score is as follows: Calculate the current load score, historical load performance score, future load prediction score, interaction impact score, real-time response capability score, and redundancy score of all neighboring nodes in the network topology of the source node, and use a weighted average method to calculate the comprehensive score of the neighboring nodes. Wherein: Current load score The value used to represent the CPU and memory usage status of a node reflects its idle state. The calculation formula is as follows: In the formula, CPU represents CPU utilization and MEM represents memory utilization.

[0045] Historical load performance score This is used to measure a node's performance in processing requests over a period of time, including the average response time of historical requests and the percentage of requests successfully processed within a certain period. The calculation formula is: In the formula, ART represents the average response time, MRT represents the maximum response time, and SR represents the task success rate.

[0046] The future load prediction score represents the predicted load change trend of neighboring nodes over a future period of time. It is obtained by collecting historical data and using time series models for prediction.

[0047] Interaction Impact Rating This method is used to analyze the connectivity and mutual influence of neighboring nodes in the network topology to assess their load capacity. It involves counting the number of directly connected neighboring nodes and the number of requests and interactions between other neighboring nodes and this neighboring node over a period of time. The calculation formula is as follows: In the formula, C is the number of nodes connected to a neighboring node in the network topology; MC represents the maximum number of connections in the network topology; IF represents the interaction frequency of neighboring nodes; and MIF represents the maximum interaction frequency.

[0048] Real-time response capability score This is used to measure the rapid response capability of neighboring nodes under sudden events; it is calculated by recording the average fault recovery time of neighboring nodes when they recover to normal status after a failure or high load, and measuring the average response time of neighboring nodes during a load surge. The formula is as follows: In the formula, FRT represents the fault recovery time, and MRT represents the maximum fault recovery time.

[0049] Redundancy score This is used to evaluate the ability of neighboring nodes to employ redundancy strategies in the network topology to ensure continuous service capability; it assesses the ratio of available redundant resources (such as backup servers, standby connections) to total resources for neighboring nodes, and records the frequency with which the node implements redundancy strategies. The calculation formula is as follows: In the formula, R represents redundant resources, TR represents total resources, RI represents the number of times the redundancy strategy is implemented, and MRI represents the maximum number of implementations.

[0050] By comprehensively considering multiple scoring dimensions, a more comprehensive and intelligent decision-making mechanism is introduced in the selection process of neighboring nodes. This mechanism can more accurately identify nodes with lighter loads, thereby optimizing task distribution, reducing resource waste, and improving the load balancing efficiency of the entire distributed architecture system. Furthermore, through social influence scoring and redundancy scoring, it helps select nodes that are more closely connected in the network topology and where redundancy strategies are implemented. This effectively reduces the risk of single points of failure and enhances the overall stability of the system.

[0051] 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. A tax self-service terminal interactive control system based on a distributed architecture, characterized in that, It includes an interaction module, which displays a user interface to support user transaction operations, and collects interaction data and builds it into an interaction dataset; The feedback and prompt module determines whether the operation was successful based on the interaction data and displays the operation result to the user; When an operation is successful, prompt the user for the next step; when an operation fails, display a failure message and provide a solution. The fault handling module determines the error type of the operation failure based on the interactive dataset when the operation fails, including hardware-level errors, interface-level errors and distributed-level errors, and calls up the corresponding solutions. When an operation fails due to a distributed-level error, the system processing module triggers and implements a load scheduling strategy.

2. The tax self-service terminal interactive control system based on a distributed architecture as described in claim 1, characterized in that, The method for constructing the interactive dataset is as follows: A two-dimensional coordinate system is established on the user interface. The boundary points of the interface frame of the operation area in the operation interface are mapped to the two-dimensional coordinate system to form the interface element frame of the operation area. Each interface element frame is assigned a unique interface element ID and stored in the interface element mapping table. When an interaction occurs, the interaction module captures the touch point coordinates (x, y) and timestamp t, as well as the pressure d and duration of the touch. The data is then merged into a single interaction data set; all interaction data are then merged in timestamp order to construct an interaction dataset.

3. The tax self-service terminal interactive control system based on a distributed architecture as described in claim 2, characterized in that, When the touch point of the captured interactive operation is a touch area, the coordinate range is determined based on the vertex coordinates of the touch area. The mathematical expression for determining the coordinate range is: In the formula, R represents the coordinate range of the touch area. Represents the coordinates of any point within the coordinate range; Let be the coordinates of the i-th vertex in the coordinate range, and n be the number of vertices in the coordinate range; Let be the slope of the line segment from vertex i to vertex i+1. ; Let be the intercept of the line segment from the i-th vertex to the (i+1)-th vertex. .

4. The tax self-service terminal interactive control system based on a distributed architecture according to claim 3, characterized in that, The method for determining whether the operation was successful is as follows: To determine if an operation is valid, the conditions for a valid operation are: the pressure applied is greater than the touch threshold and the duration of the pressure is within a reasonable time threshold range; if the coordinates of the touch point of a valid operation are within the interface element box of the corresponding operation area, then the operation is successful.

5. The tax self-service terminal interactive control system based on a distributed architecture according to claim 4, characterized in that, The method for determining the error type is as follows: Extract the touch point coordinates of the failed operation and compare them with the corresponding interface element box of the operation area; When the distance between the touch point coordinates of the failed operation and the boundary of the interface element box is less than the preset judgment threshold, it is judged as a hardware-level error and offset detection is triggered. When the touch point coordinates are within a coordinate range and overlap with multiple interface element boxes, it is judged as an interface-level error and the valid area detection is triggered. When a user successfully completes an operation but fails after the operation request is sent, it is judged as a distributed-level error, triggering a load balancing strategy.

6. The tax self-service terminal interactive control system based on a distributed architecture according to claim 5, characterized in that, The offset detection method is as follows: Calculate the geometric distance between the touch point coordinates of the failed operation and the center point coordinates of the interface element box, and generate an offset vector; read the first Q interaction data of the failed operation in the interaction dataset, and generate an offset vector; If the same offset angle exists in Q consecutive interactive operations and the offset amount is greater than the preset offset threshold, it is determined to be a systematic offset. If Q consecutive interactive operations have the same offset angle and the offset changes over time, it is determined to be a gradual offset. If there are different offset angles and different offset amounts in Q consecutive interactive operations, but the difference distribution conforms to a normal distribution, then it is determined to be a random offset.

7. The tax self-service terminal interactive control system based on a distributed architecture according to claim 6, characterized in that, The solution to the hardware-level error is: Based on the determined offset type, set the processing priority; systematic offsets have a higher priority than gradual offsets, which in turn have a higher priority than random offsets. For systemic offsets, perform global touch automatic calibration, based on voice prompts, and display a calibration progress bar. Dynamic compensation is performed for gradual offset. The calibration area is displayed on the user interface, and the user is prompted by voice to place their finger on the calibration area and keep it still. A calibration progress bar is also displayed until the calibration is completed. For random offsets, the user is prompted to try the operation again via voice.

8. The tax self-service terminal interactive control system based on a distributed architecture according to claim 5, characterized in that, The method for detecting the effective region is as follows: Calculate the coordinate range of the touch point and the area of ​​the overlapping region of each interface element box, and select the overlapping region with the largest area as the determined interface element box.

9. The tax self-service terminal interactive control system based on a distributed architecture as described in claim 8, characterized in that, The solution to the interface-level error is as follows: The system provides voice prompts to users to reselect the operation area and highlights the corresponding interface element box on the user interface.

10. The tax self-service terminal interactive control system based on a distributed architecture according to claim 5, characterized in that, The load scheduling strategy is as follows: Record the type of task being processed by the current node and determine the task level based on the preset priority sorting; select the node with the highest comprehensive score among the neighboring nodes as the target node, package the detailed information of the task being processed by the current node into a request, send it to the selected target node, wait for the target node's response and processing, and during the target node's processing, report the execution status to the source node and return the real-time updated data to the source node. The method for selecting the node with the highest overall score is as follows: Calculate the current load score, historical load performance score, future load prediction score, interaction impact score, real-time response capability score, and redundancy score of all neighboring nodes in the network topology of the source node, and use a weighted average method to calculate the comprehensive score of neighboring nodes.

Images