An intelligent interaction method, system and device for a vehicle display terminal

Abstract

The application discloses an intelligent interaction method, system and device for a vehicle display terminal, belongs to the technical field of vehicle intelligent interaction management, and comprises the following steps: monitoring abnormal conditions of vehicle state sensing and information priority dynamic conflict conditions, and determining whether to monitor, optimize and process and select a notification mode for abnormal conditions of a vehicle-mounted intelligent terminal pushing process, so that the accuracy of dynamic adaptation of a pushing strategy to a driving state and information demand is improved, the timeliness of key dispatching information transmission and the execution accuracy are further improved, and the problem of low interaction management efficiency of drivers and vehicles in the prior art is solved.

Description

Technical Field

[0001] This invention relates to the field of vehicle intelligent interactive management technology, and in particular to an intelligent interactive method, system and device for vehicle display terminals. Background Technology

[0002] Vehicle display terminals are a core pillar for the safe, efficient, and user-friendly operation of modern automobiles, especially commercial and public transportation vehicles. By deeply integrating hardware, software, and scenarios, they solve the core contradictions in information processing within the driving environment.

[0003] As the direct interface for drivers, the driver display terminal not only presents information and responds to operations, but also serves as an intelligent hub connecting drivers, vehicles, and the operations management center. This hub function relies heavily on efficient collaboration with the in-vehicle intelligent terminal—the vehicle's communication center. The in-vehicle intelligent terminal contains shared memory or a high-speed message bus (such as Automotive Ethernet). It is responsible for real-time collection of vehicle data (such as speed and direction), processing satellite and network signals, parsing dispatch center instructions, and then accurately pushing information (such as station arrival notifications and traffic congestion notifications) to the driver display terminal, or converting it into clear voice announcements. Therefore, the driver display terminal is the "intelligent manifestation" of this collaborative system. Through seamless integration with the underlying in-vehicle intelligent terminal, it jointly ensures driving safety, optimizes operational processes, and meets the traceability requirements for data recording and supervision.

[0004] To achieve efficient interactive management of drivers and vehicles, the existing method is as follows: After the vehicle display terminal is bound to the driver's ID, driven by multimodal perception and scenario-based intelligent kernel, the main interface or dedicated window dynamically visualizes key statuses according to the driving scenario, such as bus stops, Beidou signals, network connection, camera status, and the working status of the vehicle intelligent terminal (including but not limited to vehicle positioning mode, recording status, and vehicle speed). At the same time, the screen brightness can be automatically adjusted according to changes in ambient light to ensure comfortable display all day long. In terms of communication and interaction management, the in-vehicle intelligent terminal can deeply integrate the cloud ecosystem and dispatch instructions, and send the integrated and dispatched content to the vehicle display terminal to achieve intelligent message delivery: non-emergency notifications (such as ordinary notification instructions) are displayed as icons in the status bar, while emergency instructions (such as highway warnings, congestion instructions, etc.) are triggered by priority voice broadcast through multimodal fusion to maximize driver focus. The vehicle display terminal can feed back its interaction status (such as message read, voice playback completed) to the in-vehicle intelligent terminal, which then uploads it to the cloud; when the vehicle is put into P gear or turned off, the intelligent system intelligently determines that the scenario is converted to a pending state, allowing the driver to centrally view and process all the central messages during the trip.

[0005] For example, Chinese invention patent CN112613629B discloses a vehicle interaction method, system, and vehicle with external interaction function, including: obtaining the location information of the reserved vehicle; sending the location information to a server so that the server can determine whether the location information meets preset conditions; and displaying a first interaction identifier on the user terminal when the preset conditions are met, wherein the first interaction identifier corresponds to a second interaction identifier displayed on the display device of the reserved vehicle, and the second interaction identifier is used to identify the reserved vehicle.

[0006] The existing technology focuses on location verification and identifier matching mechanisms, but it cannot meet the dynamic interaction needs of "driver status as the center" during driving. Especially in complex scenarios such as multiple information superposition and high-load driving, it is difficult to avoid problems such as information push mismatch with driving status and delay in key command transmission, thus exposing the adaptation gap of the existing technology in the field of human-machine collaborative interaction during driving.

[0007] The above-mentioned technology has at least the following technical problems:

[0008] In the process of driver-vehicle interaction management, there may be problems such as incomplete vehicle-to-vehicle state perception and coordination (e.g., failure to capture driver's hand movements in real time under high operational load scenarios, while the hands are the direct actuators for driver control and operation interaction, and their frequency, force, and pattern are one of the most direct indicators of the driver's real-time operational load) and dynamic conflicts in information priorities (e.g., regular station notifications and sudden road condition warnings are triggered at the same time). These issues cause the response of the in-vehicle intelligent terminal to be inconsistent with the driver's real-time operational load. Existing technologies usually adopt single-state judgment (e.g., judging the driving scenario based solely on vehicle speed) and static message hierarchical logic, which cannot fully adapt to the dynamically changing scenario requirements during driving (e.g., temporary stops, allocation differences under sudden road conditions) and the driver's real-time operational state. There is still a shortcoming that the presentation of interactive information does not match the driving state. This results in the failure to push critical dispatch information to the driver in high operational load scenarios (e.g., highway overtaking, congested following), which may further cause safety risks of delayed reception of emergency dispatch instructions, ultimately leading to low efficiency in driver-vehicle interaction management. Summary of the Invention

[0009] To address the technical problem of low efficiency in driver and vehicle interaction management in existing technologies, embodiments of the present invention provide an intelligent interaction method, system, and device for a vehicle display terminal. The technical solution is as follows:

[0010] On the one hand, an intelligent interaction method for vehicle display terminals is provided. This method includes: assessing the real-time operational load based on abnormal situations perceived by the vehicle status and dynamic conflicts in information priorities; determining whether to monitor abnormal situations during the push process of the in-vehicle intelligent terminal based on the assessment results; if monitoring is performed, dynamically selecting a method to adapt the urgency of the push notification according to the monitoring results after the abnormal situation monitoring ends; if no monitoring is performed, taking measures to reduce the push latency of the push information in the in-vehicle intelligent terminal and the in-vehicle display terminal.

[0011] On the other hand, an intelligent interaction system for a vehicle display terminal is provided. This system applies an intelligent interaction method for a vehicle display terminal, including: a load anomaly monitoring module for assessing the real-time operational load based on abnormal situations perceived by the vehicle status and dynamic conflicts in information priorities; a push anomaly assessment module for deciding whether to monitor abnormal situations in the push process of the in-vehicle intelligent terminal based on the assessment results; a notification method adaptation module for dynamically selecting a method to adapt to the urgency of push notifications based on the monitoring results after monitoring is completed if monitoring is performed; and a delay processing module for reducing the push delay of push information in the in-vehicle intelligent terminal and the in-vehicle display terminal if monitoring is not performed.

[0012] On the other hand, a vehicle display terminal device is provided, which applies an intelligent interaction method for a vehicle display terminal, including: a display screen, input buttons, a communication interface, and a processor; the display screen is used to display the working status of the vehicle-mounted intelligent terminal, view dispatch center messages, and view local recordings and videos; the input buttons are used to provide physical interaction between the driver and the display screen; the communication interface is used to access a local storage device and read local data; the processor is used to coordinate the logical operations of the intelligent interaction system and manage the data transmission and reception and protocol parsing of the communication interface.

[0013] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0014] 1. Based on abnormal situations and dynamic conflicts in information priorities arising from vehicle status perception, an assessment reflecting real-time operational load is conducted. This helps to accurately map driving scenarios to operational loads, providing data support for optimizing information push strategies. The assessment results determine whether to monitor abnormal situations during the push process of the in-vehicle intelligent terminal. If monitoring is conducted, after the monitoring period, the appropriate method for adjusting the urgency of push notifications is dynamically selected based on the monitoring results. This helps to achieve efficient adaptation of information transmission under different load scenarios, thereby avoiding unnecessary information from interfering with driver focus and ensuring the real-time delivery of core dispatch instructions. If no monitoring is conducted, abnormal push notifications help to dynamically correct and optimize the push mechanism, thereby ensuring the driver receives complete critical information and reducing the risk of dispatch deviations due to missing information.

[0015] 2. By conducting state-aware anomaly assessment and information priority dynamic conflict anomaly assessment, the upper limit number of anomalies and the abnormal task interaction value are obtained respectively. When the upper limit number of anomalies does not exceed the specified number of qualified state types, the corresponding driving scenario risk level is further quantified. Compared to existing technologies that rely solely on a single parameter (such as vehicle speed) to determine scenarios and lack risk level quantification standards, thus failing to accurately match driving safety requirements, this approach helps achieve multi-dimensional quantification and classification of driving scenario risks, thereby improving the scenario adaptability and safety redundancy of dispatch instruction pushes. Furthermore, when the abnormal task interaction value does not exceed the specified abnormal task interaction value, compared to existing technologies that use static message hierarchical logic and have low information-to-driving-requirement matching, dynamic matching of priority interaction information and instructions is performed. This helps achieve precise alignment between priority information and the driver's real-time operating status, thereby improving the response efficiency and execution accuracy of information interaction.

[0016] 3. In high-load scenarios (such as congested or accelerating road sections), where inefficiency in vehicle dispatching command interaction leads to service and dispatching deviations, anomaly assessment of vehicle status is performed to obtain abnormal vehicle status perception values. If the abnormal vehicle status perception value exceeds a predetermined driving anomaly perception value, a safety information pending interaction mode is entered. After the safety information pending interaction mode ends, priority interaction information and dynamic matching of commands are initiated. Compared to the shortcomings of existing technologies, such as disordered information push and delayed command reception in high-load scenarios, this technology helps to achieve a dynamic balance between safety and dispatching requirements in high-load scenarios, thereby improving the interaction efficiency of vehicle dispatching commands and the accuracy of service execution.

[0017] 4. By dynamically matching priority interaction information and instructions to obtain interaction matching values, and optimizing the interaction information response when the interaction matching value falls within the unqualified range, this method addresses the shortcomings of existing technologies, such as the lack of a closed-loop optimization mechanism for information response and the absence of corrective measures after unqualified matching, which leads to persistently low interaction efficiency. This method helps to achieve dynamic optimization and continuous iteration of the entire interaction information response process, thereby realizing closed-loop improvement of driver and vehicle interaction management and enhancing the robustness and adaptability of the overall interaction.

[0018] 5. When multiple information is pushed simultaneously, it is necessary to pay attention to the common needs of interaction efficiency, voice broadcasting, and screen display. By obtaining the interaction efficiency evaluation value and matching the notification method based on the interaction efficiency evaluation value, compared with the shortcomings of existing technologies that have the disadvantages of interaction channel conflict when multiple information is superimposed, which can easily lead to information redundancy or perception overload, it helps to achieve optimal allocation and resource scheduling of interaction channels in multi-information scenarios, thereby achieving a dual improvement in interaction efficiency and driving comfort, and ensuring decision-making efficiency. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating an intelligent interaction method for a vehicle display terminal provided in an embodiment of this application;

[0021] Figure 2 A schematic diagram illustrating the dynamic conflict and anomaly assessment of information priority in an intelligent interaction method for a vehicle display terminal, provided in an embodiment of this application.

[0022] Figure 3 A schematic diagram illustrating the dynamic matching of interactive information and instructions in an intelligent interaction method for a vehicle display terminal provided in an embodiment of this application;

[0023] Figure 4 This is a priority classification diagram of push information provided in the embodiments of this application;

[0024] Figure 5 This is a schematic diagram of the structure of an intelligent interactive system for a vehicle display terminal provided in an embodiment of this application. Detailed Implementation

[0025] Embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of the present disclosure are shown in the drawings, it should be understood that embodiments of the present disclosure may be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure.

[0026] It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure. In the description of the embodiments of this disclosure, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "this embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects.

[0027] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0028] Example 1: This application provides an intelligent interaction method for a vehicle display terminal, which includes the following steps:

[0029] Load anomaly monitoring: Based on abnormal situations perceived by vehicle status and dynamic conflicts in information priorities, real-time operational load anomaly assessment is performed to reflect the real-time operational load. By monitoring load anomalies, it is possible to accurately quantify the driver's real-time operational load and abnormal states, providing data support for subsequent interaction strategy adjustments and avoiding load misjudgments caused by judging based on a single parameter.

[0030] Push Anomaly Assessment: Based on the assessment results, it is determined whether to monitor abnormal situations in the push process of in-vehicle intelligent terminals. By conducting push anomaly assessment, it is helpful to realize the on-demand allocation of push monitoring resources, avoid system computing power redundancy caused by indiscriminate monitoring, and improve the pertinence and response efficiency of interactive decisions.

[0031] Notification method adaptation: If monitoring is performed, after the abnormal situation monitoring ends, the method of pushing notifications based on the monitoring results will be dynamically selected to adapt to the urgency of the abnormal situation. By adapting the notification method, it is helpful to achieve dynamic matching between the information push method, the urgency of the driving scenario and the driver's real-time load status, thereby improving the accuracy of key information transmission while ensuring driving focus.

[0032] Delay processing: If no monitoring is performed, measures will be taken to reduce the push latency of push information in the vehicle intelligent terminal and vehicle display terminal in case of abnormal push information. By delaying the push, it is possible to quickly correct the information lag caused by abnormal push, reduce the interference of transmission delay on driving decisions, and ensure the timely delivery of core scheduling information.

[0033] like Figure 1 The figure shows a flowchart of an intelligent interaction method for a vehicle display terminal provided by an embodiment of the present invention. As shown in the figure: Based on the abnormal situation of vehicle status perception and the dynamic conflict of information priority, an assessment reflecting the real-time operation load is performed; based on the assessment result, it is determined whether to monitor the abnormal situation of the push process of the vehicle intelligent terminal (i.e., to perform dynamic matching of interactive information and instructions); if monitoring is performed, after the abnormal situation monitoring ends, a method to adapt the urgency of the push notification is dynamically selected according to the result of the abnormal situation monitoring; if no monitoring is performed, processing to reduce the push delay of push information in the vehicle intelligent terminal and the vehicle display terminal is adopted.

[0034] In the early stages of designing the intelligent interaction method for a vehicle display terminal provided in this application, a database system with standardized data storage and management functions has been constructed. In addition to realizing the orderly storage, efficient retrieval, and standardized management of various types of data during the interaction process, the system is also pre-configured with multiple types of setting parameters to support intelligent interaction decisions, including but not limited to the specified upper limit number of state types and the specified driving anomaly perception values. The various values ​​are directly set by technical personnel. These parameters are all directly set by technical personnel in combination with the actual driving scenario requirements, vehicle operation specifications, and safety standards, providing basic data support for the smooth operation of core logics such as anomaly assessment and information matching in the subsequent intelligent interaction process.

[0035] In this embodiment, the coordinated efforts of load anomaly monitoring, push anomaly assessment, notification method adaptation, and delayed processing help to achieve closed-loop management of the entire human-machine interaction process in driving scenarios. This helps to solve core pain points such as incomplete state perception and dynamic conflicts in information priorities, thereby helping to achieve adaptive optimization of in-vehicle intelligent interaction, improve the efficiency of driver-vehicle interaction management, reduce safety risks caused by improper information transmission, and ensure the safety of the driving process and the accuracy of operation scheduling.

[0036] Preferably, the assessment reflecting the real-time operational load includes state perception anomaly assessment and information priority dynamic conflict anomaly assessment. State perception anomaly assessment means acquiring data to evaluate abnormal situations in vehicle state perception coordination and deciding whether to process the corresponding interactive information to reduce the interaction delay caused by perception anomalies, such as directly writing push information (including but not limited to vehicle arrival notifications, road congestion notifications, road construction notifications, route adjustment notifications, etc.) into the shared memory inside the vehicle intelligent terminal, adjusting screen brightness, etc. Information priority dynamic conflict anomaly assessment means acquiring data to measure abnormal task interaction and deciding whether to process the corresponding interactive information to adapt to the interaction delay caused by abnormal push resource allocation, such as allocating corresponding buffers for interactive information, etc.

[0037] In this embodiment, data reflecting abnormal vehicle state perception and coordination is used as a decision-making condition for reducing interaction delays caused by perception anomalies. This helps to accurately locate interaction bottlenecks caused by incomplete perception and coordination, thereby triggering targeted optimization processing, avoiding system redundancy caused by blind processing, and ensuring real-time coordination between vehicle state and driver operation. This reduces the interaction delay of scheduling instructions caused by perception anomalies and improves the timeliness of information transmission in high-load scenarios. Data reflecting abnormal task interaction is used as a decision-making condition for processing interaction delays caused by abnormal push resource allocation. This helps to improve the dynamic adaptability of push resource allocation, reduce information congestion caused by resource competition, ensure the smooth flow of core task interaction channels, and optimize resource utilization efficiency in different scenarios. By simultaneously evaluating state perception anomalies and dynamic conflict anomalies in information priority, it helps to achieve full-dimensional coverage and synchronous control of the two types of core anomalies during the interaction process, avoiding anomaly omissions caused by single evaluation, providing comprehensive and accurate decision-making basis for subsequent processing, and helping to systematically solve the problems of interaction delay and adaptation imbalance.

[0038] Preferably, the specific process of state perception anomaly assessment is as follows: acquire vehicle dynamic signals as a quantitative standard for driving scenario risk; vehicle dynamic signals include the average vehicle speed monitored by wheel speed sensors, the average acceleration monitored by IMU (Inertial Measurement Unit), and the average yaw rate monitored by IMU during a preset vehicle monitoring period; count the number of vehicle dynamic signals that are greater than the upper limit of the corresponding vehicle dynamic acceptable range based on a counter, record this number as the upper limit anomaly number, and determine whether the upper limit anomaly number is greater than the specified upper limit state type number. For example, when only the average vehicle speed is detected to exceed the upper limit of the corresponding vehicle dynamic acceptable range, while the average acceleration and average yaw rate do not exceed the upper limit of the corresponding vehicle dynamic acceptable range, the upper limit anomaly number is recorded as 1. The specified upper limit state type number is represented by the average upper limit anomaly number over a historical time period.

[0039] To trigger the dispatch center's emergency response mechanism immediately, enabling dispatchers to monitor the comprehensive anomaly status of the vehicle perception system in real time, quickly identify the root cause of faults (such as sensor failures, data transmission link interruptions, etc.), prevent the continuous accumulation of anomalies from escalating risks, and provide accurate basis for the dispatch center to issue emergency response instructions, ensuring the safety of the driving process and the effectiveness of dispatch execution, an initial dynamic qualification judgment of the vehicle is performed: if the upper limit number of anomalies exceeds the upper limit number of status types, an alarm is immediately sent to the dispatch center; if the upper limit number of anomalies does not exceed the upper limit number of status types, the risk level of the driving scenario is quantified. The specific process of driving scenario risk level quantification is as follows: based on the accurate identification data reflecting the vehicle's perception status, a value used to quantify the risk level of the driving scenario is obtained, i.e., the driving scenario risk quantification value; by inputting the vehicle's dynamic signals into a preset machine learning model, the predicted driving scenario risk quantification value is output.

[0040] It should be added that by randomly dividing the dataset, which consists of vehicle dynamic signals collected within a specific historical interaction period and preset predicted driving scenario risk quantification values, into a training set, and inputting the training set into a preset machine learning model for training, a mature model is obtained. Taking the random forest model as an example, the accuracy and stability of the model are improved by constructing multiple decision trees and combining their prediction results. During the training process, the model will repeatedly learn and analyze the data in the training set, and continuously adjust its parameters and structure based on the inherent correlation between data features and predicted driving scenario risk quantification values ​​to minimize prediction errors, thereby providing strong data support and decision-making basis for safe driving and intelligent interaction of vehicles. The reacquired vehicle dynamic signals can be input into the corresponding trained model to output predicted driving scenario risk quantification values.

[0041] The risk quantification value for driving scenarios is obtained through the following method:

[0042] ;

[0043] Wherein, FX represents the quantification value of driving scenario risk, λ represents the state adjustment factor, which is set by preset personnel based on historical experience, P represents the probability of driver hand movements (based on the prediction output of a recurrent neural network according to the current driving scenario), A represents the recognition accuracy under the current vehicle perception state (such as the ratio of the number of times the vehicle display terminal successfully responds to the dispatch center's instructions monitored by the counter to the total number of responses), and A0 represents the recognition accuracy under the ideal state, which is set by preset personnel.

[0044] It should be added that the training model is obtained by inputting a training set consisting of driving scenario states (such as congestion, parking, emergency braking, etc.) from historical time periods and preset hand action probabilities into a recurrent neural network. The driving scenario states re-collected through the in-vehicle intelligent terminal are then input into the training model, and the corresponding probability of the driver's hand actions is output.

[0045] It should be explained that existing risk quantification formulas typically use the classic structure of "risk = probability × severity". The probability part of this formula is represented by P. The larger the ratio of the recognition accuracy under the current vehicle perception state to the recognition accuracy under the ideal state, the closer the current perception performance is to the ideal state, the lower the degree of performance loss, and the lower the severity. The smaller the ratio, the higher the severity and the higher the risk level. The state adjustment factor is essentially equivalent to assigning a severity adjustment weight to different levels and combinations of abnormal system states based on a predefined "risk matrix".

[0046] To reduce latency bottlenecks in intermediate transmission links, the high-speed read / write characteristics of the shared memory (a physical memory area directly accessed by multiple processor cores) of the in-vehicle intelligent terminal are utilized to achieve millisecond-level transmission and processing of push information. The system determines whether the quantified risk value of the driving scenario is greater than the predicted risk value. If so, the corresponding push information is directly written to the shared memory inside the in-vehicle intelligent terminal; otherwise, a wait-for-push command is sent to the dispatch center. The system obtains the quantified risk value and the predicted risk value of the driving scenario by quantifying the risk level of the driving scenario. When the quantified risk value of the driving scenario is greater than the predicted risk value, the corresponding push information is directly written to the shared memory inside the in-vehicle intelligent terminal. The shared memory of the in-vehicle intelligent terminal is a multi-processor core, such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit). The system directly accesses physical memory areas such as the Unit (Graphics Processing Unit). After push information is written, it does not need to go through intermediate links such as bus transmission, protocol parsing, and data copying. This can minimize the time difference between receiving and processing push information, solve the latency problem in high-urgency scenarios, and help achieve millisecond-level transmission of critical information in high-risk scenarios, ensuring the instant delivery of emergency dispatch instructions. By sending a waiting push instruction to the dispatch center when the risk quantification value of the driving scenario is not greater than the predicted risk quantification value of the driving scenario, it helps to achieve orderly caching and staggered push of non-urgent information, avoid low-priority information occupying high-urgency interaction channels, and optimize the dynamic allocation efficiency of push resources.

[0047] To further enhance the adaptability of interaction and display during driving and reduce the impact of visual interference on driving safety, an additional function is provided: adaptive display terminal screen brightness adjustment. This function can configure the corresponding parameters of the vehicle display terminal based on the output screen brightness, which helps to achieve real-time dynamic adaptation of screen brightness, ambient light and driving scenario. This helps to ensure display clarity in all-weather driving scenarios and indirectly improves driving and operation safety.

[0048] The state perception anomaly assessment also includes adaptive display terminal screen brightness adjustment; the specific process of adaptive display terminal screen brightness adjustment is as follows: adaptive display terminal screen brightness adjustment is used to promote a dynamic balance between the effectiveness of information transmission and the safety of visual experience while ensuring driving safety; the light intensity (monitored by the vehicle ambient light sensor) is input into the pre-created screen brightness allocation table, the screen brightness is output, and the corresponding parameters of the vehicle display terminal are configured based on the output screen brightness.

[0049] It needs to be explained in detail that in the intelligent interaction process of the vehicle display terminal, a data architecture with multi-dimensional mapping characteristics and supporting two different forms of mapping relationships is first established: one-to-one precise mapping and many-to-one composite mapping. This ensures accurate conversion and efficient processing of data between different dimensions. Personnel are pre-selected to conduct comprehensive and systematic collection of diverse information within a specific historical interaction period. This information covers a wide range, including combinations of light intensity, data volume corresponding to outliers and interaction information, combinations of interaction matching values ​​and data volume of information pushed when the vehicle intelligent terminal starts pushing information, parameter combinations of initial speech rate and initial speech volume, parameter combinations of initial push content display area width and initial push content display font size, and other key environmental parameters that have a potential impact on the performance of the vehicle display terminal. This collected and pre-processed historical key information is input into a machine learning model with feature importance assessment capabilities. Among these, the random forest model is widely used due to its powerful feature selection and classification prediction capabilities.

[0050] The random forest model uses an information gain-based feature splitting algorithm to deeply analyze the input data. Based on the pre-defined mapping criteria, the model accurately calculates the contribution of each feature to the target variable during each layer of decomposition. By comprehensively considering these contributions, the model can automatically select key features that have a significant statistical correlation with the target variable. During the model training phase, based on the selected core features, a data hierarchical clustering algorithm is used to group the data.

[0051] By continuously iterating and optimizing cluster centers and boundaries, data within the same cluster exhibits higher similarity, while data between different clusters shows greater diversity. After this series of optimization operations, it is possible to obtain rigorously verified acceptable values ​​for screen brightness, buffer, and time slices, along with corresponding combinations of push priority, speech rate and volume, and push content display area width and font size.

[0052] Subsequently, the system correlates and matches the raw data from historical monitoring with the iterative optimization parameters output by the model, constructing a screen brightness allocation table, a buffer allocation table, a time slice partitioning table, a voice broadcast parameter allocation table, and a display parameter allocation table. When monitoring information such as the combination of light intensity, write anomalies, and the amount of data corresponding to interactive information; the combination of the interaction matching partition value and the amount of data pushed when the vehicle intelligent terminal starts pushing; the parameter combination of the initial broadcast speed and initial broadcast volume; and the parameter combination of the initial push content display area width and the initial push content display font size, this information only needs to be input into the screen brightness allocation table. The system utilizes tables or sets such as buffer allocation tables, time slice allocation tables, voice broadcast parameter allocation tables, and display parameter allocation tables. Based on the mapping relationships recorded in these tables, the system can quickly and accurately obtain the corresponding screen brightness, buffer, time slice qualification values, and corresponding push priority combinations, broadcast speed and volume combinations, and push content display area width and push content font size combinations. This process enables dynamic adaptation of environmental information and display parameters, effectively improving the visibility and user experience of the vehicle display terminal under different environmental conditions, and providing a solid technical guarantee for the stable operation of the vehicle intelligent interaction system.

[0053] In this embodiment, vehicle dynamic signals and the upper limit number of anomalies are obtained by performing state perception anomaly assessment. When the upper limit number of anomalies exceeds the specified upper limit number of state types, an alarm is sent to the dispatch center; otherwise, the risk level of the driving scenario is quantified. This helps to achieve accurate early warning and closed-loop management of vehicle state perception anomalies, while providing data-driven quantitative basis for driving scenario risks, avoiding subjectivity and one-sidedness in risk assessment. Through the correlation between state perception anomaly assessment and driving scenario risk level quantification, it helps to further strengthen the coupling analysis of abnormal situations and driving risks, making subsequent interaction strategy adjustments more in line with actual scenario needs, thereby improving the responsiveness of the in-vehicle intelligent terminal to dynamic driving scenarios and reducing interaction delays caused by perception deviations.

[0054] Example 2, as an additional technical solution to Example 1, without changing the basic method of Example 1; when there are high operational load scenarios (such as congested or accelerating road sections), and the driver's hand gestures are not captured in real time, resulting in low efficiency of bus dispatch instruction interaction and causing service and dispatch deviations, in order to accurately identify the lack of state perception in high operational load scenarios, overcome the interaction bottleneck caused by incomplete perception coordination, and improve the timeliness and accuracy of bus dispatch instruction transmission, it is necessary to conduct state perception anomaly assessment; the specific process of state perception anomaly assessment is as follows: when the frequency of driver's hand gesture monitoring acquired by the on-board intelligent terminal is greater than that preset by a pre-set person... When the hand movement monitoring frequency is set, the average vehicle speed within the corresponding preset vehicle monitoring time period is monitored by wheel speed sensors and marked as the vehicle abnormal state perception value. This serves as a quantitative standard for the risk of driving scenarios under abnormal perception state. (If the driver's hand movement monitoring frequency within the preset vehicle monitoring time period is not greater than the preset hand movement monitoring frequency, it is considered that the perception state is qualified and information is pushed normally.) The higher the average vehicle speed, the higher the probability of perception data distortion (such as hand movement capture delay, misjudgment, etc.) and interaction command mismatch. Moreover, the risk transmission speed is faster under high-speed driving conditions, and the driving risk formed by the superposition of high operating load and abnormal vehicle speed is higher.

[0055] To ensure the interaction strategy is accurately adapted to the current scenario and avoid interaction deviations caused by perceived anomalies, while also considering the effective transmission of dispatch instructions and driving safety, anomaly assessment is performed based on the vehicle anomaly perception value and a specified driving anomaly perception value: If the vehicle anomaly perception value is greater than the specified driving anomaly perception value, a prompt is sent to initiate a safety information pending interaction process, and after the safety information pending interaction process ends, the interaction information and instructions are dynamically matched; the specified driving anomaly perception value is represented by the average value of vehicle anomaly perception values ​​over a historical time period; if the vehicle anomaly perception value is not greater than the specified driving anomaly perception value, a prompt is sent to enter a safety information interaction mode; the safety information interaction mode specifically involves marking the interaction information corresponding to vehicle anomaly perception values ​​not greater than the specified driving anomaly perception value as non-urgent pending push information and sending a waiting push instruction; the safety information pending interaction process includes: marking the interaction information corresponding to vehicle anomaly perception values ​​greater than the specified driving anomaly perception value as interaction anomaly push information.

[0056] Based on interactive anomaly push information, state perception and interaction efficiency improvement correction are initiated to enhance the accuracy and reliability of anomaly identification: Interactive anomaly push information is input into a pre-set state anomaly prediction model, which outputs the perceived anomaly level corresponding to the interactive anomaly push information, including Level 1 perceived anomaly results and Level 2 perceived anomaly results, with the perceived anomaly level decreasing accordingly; when only Level 1 perceived anomaly results are detected, the corresponding interactive anomaly push information is directly written to the shared memory corresponding to the vehicle intelligent terminal, directly accessing the fixed physical memory address corresponding to the shared memory, thereby reducing the relay latency of push information between the vehicle intelligent terminal and the vehicle display terminal; when only Level 2 perceived anomaly results are detected, the corresponding interactive anomaly push information is asynchronously reported to the dispatch center; when both Level 1 and Level 2 perceived anomaly results are detected simultaneously, the interactive anomaly push information corresponding to the Level 1 perceived anomaly result is processed first.

[0057] Specifically, the model is trained by inputting the interaction anomaly push information collected within a specific historical interaction period and the preset perception anomaly level into a preset machine learning model (such as a random forest model) to obtain the trained model, namely the state anomaly prediction model; by inputting the reacquired interaction anomaly push information into the state anomaly prediction model, the corresponding perception anomaly level can be output.

[0058] By initiating a pending safety information interaction process to obtain first-level and second-level perception anomaly results, when only a first-level perception anomaly is detected, the corresponding interactive anomaly push information is directly written to the shared memory of the vehicle's intelligent terminal. This helps improve the transmission and processing efficiency of high-urgency anomaly information, shortens the response time of critical instructions, and thus enables the immediate fulfillment of core scheduling needs under high-load scenarios, reducing the safety risks caused by severe perception anomalies. When only a second-level perception anomaly is detected, the corresponding interactive anomaly push information is asynchronously reported to the dispatch center. This helps improve the reporting efficiency of non-urgent perception anomalies, avoids occupying the core interactive resources of the vehicle terminal, and thus helps the dispatch center to achieve comprehensive monitoring and batch optimization of vehicle perception status, continuously improving system adaptability. When both first-level and second-level perception anomaly results are detected simultaneously, the interactive anomaly push information corresponding to the first-level perception anomaly is processed first. This helps improve the priority adaptability of anomaly information processing, ensures that core risks are resolved first, and thus helps achieve a dynamic balance between interactive safety and dispatch efficiency under high operational load scenarios, minimizing operational and safety risks.

[0059] In this embodiment, an abnormal vehicle state perception value is obtained by performing a state perception anomaly assessment. When the abnormal vehicle state perception value is greater than a specified driving anomaly perception value, a safety information pending interaction process is initiated. Otherwise, the interaction information and instructions are dynamically matched. This helps to achieve differentiated adaptation of interaction strategies in high-operation-load scenarios and normal scenarios, avoids instruction transmission delays or mismatches caused by perception anomalies, and thus ensures accurate matching between bus dispatch instructions (such as route adjustment instructions) and the driver's real-time operation status, improves dispatch interaction efficiency, and reduces the risk of service and dispatch deviation.

[0060] like Figure 2 The figure shows a schematic diagram of dynamic conflict and anomaly evaluation of information priority in an intelligent interaction method for a vehicle display terminal provided by an embodiment of the present invention. As shown in the figure: the abnormal task interaction value is obtained, and if the abnormal task interaction value is not greater than the specified abnormal task interaction value, the interaction information and instructions are dynamically matched. Otherwise, the shared memory write anomaly of the vehicle intelligent terminal is monitored, and the write anomaly value is obtained. If the monitored write anomaly value is not greater than the pre-specified write anomaly value, the interaction information and instructions are dynamically matched. Otherwise, a corresponding buffer is allocated for the interaction information.

[0061] Preferably, the specific process of information priority dynamic conflict anomaly assessment is as follows: During the process of the vehicle-mounted intelligent terminal receiving and parsing the instructions from the dispatch center and pushing them to the vehicle-mounted display terminal for interaction, the vehicle-mounted intelligent terminal analyzes the abnormal task interaction situation. Based on the preset push time period, the total number of times the vehicle-mounted intelligent terminal receives and parses the instructions from the dispatch center is monitored by a counter, and the corresponding value is marked as the abnormal task interaction value.

[0062] The greater the total number of times the vehicle-mounted intelligent terminal receives and parses instructions from the dispatch center, the higher the concurrency of instructions within the preset time period. This also means that the system faces greater risks of dynamic priority conflicts and resource scheduling pressure. It implies that multiple types of instructions issued by the dispatch center (such as regular station notifications, emergency road condition warnings, and maintenance instructions) highly overlap within the time window. At this time, the probability of different priority instructions competing for the terminal's CPU computing power, communication bandwidth, and display resources increases exponentially, which can easily lead to dynamic priority conflict problems such as "high-priority warnings being blocked by low-priority notifications" and "multiple instructions being pushed at the same time causing display interface errors."

[0063] In scenarios where regular station notifications and sudden road condition warnings are triggered at the same time, by monitoring abnormal task interaction values, the number of command parsing times of the vehicle intelligent terminal will increase sharply within a preset time period. If the abnormal task interaction value exceeds the normal baseline threshold, the system can directly capture the conflict signal of "multi-priority command overlapping push" by real-time monitoring of this value, and avoid high-priority warnings being blocked by low-priority notifications due to the lag in scheduling strategies.

[0064] To achieve reasonable scheduling of push resources and ensure that the timing and method of information push are precisely matched with the current scenario, a judgment is made based on the abnormal task interaction value and the specified abnormal task interaction value: if the abnormal task interaction value is not greater than the specified abnormal task interaction value, a waiting push instruction is sent, and the interaction information and instruction are dynamically matched; the specified abnormal task interaction value is represented by the average value of abnormal task interaction values ​​over a historical time period; if the abnormal task interaction value is greater than the specified abnormal task interaction value, the corresponding interaction information is directly written to the shared memory inside the vehicle intelligent terminal, and the write anomaly situation of the shared memory of the vehicle intelligent terminal is monitored from the response level of the vehicle intelligent terminal to obtain the write anomaly value; the write anomaly value is represented by the ratio of the number of requests corresponding to the interaction information that failed to be written within a preset write time after the write request is triggered (monitored by a counter) to the total number of write requests.

[0065] The judgment is based on the acquired write anomaly value. If the write anomaly value is not greater than the pre-defined write anomaly value, dynamic matching of interactive information and instructions is performed. The pre-defined write anomaly value is represented by the average value of write anomalies over a historical time period. If the write anomaly value is greater than the pre-defined write anomaly value, the current push information is marked as unqualified write information. In the next preset push time period, the write anomaly value and the data volume corresponding to the interactive information monitored by the communication protocol stack (such as CANoe / CANalyzer tool) are input into the preset buffer allocation table, the corresponding buffer is read, and the corresponding buffer is allocated for the interactive information. If the write anomaly value acquired again after buffer allocation is greater than the pre-defined write anomaly value, a write anomaly alarm is sent to the scheduling center. Otherwise, dynamic matching of interactive information and instructions is performed.

[0066] like Figure 3 The figure shows a schematic diagram of the dynamic matching of interactive information and instructions in an intelligent interaction method for a vehicle display terminal provided by an embodiment of the present invention. As shown in the figure: the interactive matching division value is obtained. If the monitored interactive matching division value belongs to the unqualified push range, the interactive information response is optimized. Otherwise, the push notification method is determined based on the output of the push information priority classification model. If the push information priority classification model outputs a high-level push result, the corresponding push notification is displayed on the vehicle display terminal and the corresponding push notification is broadcast by voice. If the output is a low-level push result, the corresponding push notification is displayed on the vehicle display terminal.

[0067] It should be added that the dynamic matching of interactive information and instructions is carried out as follows: From the perspective of the in-vehicle intelligent terminal pushing scheduling instructions, for the received priority push information, a delay assessment of the scheduling instruction push is performed based on the in-vehicle intelligent terminal. The duration corresponding to the time when the in-vehicle intelligent terminal initiates the push and the time when the in-vehicle display terminal responds, as monitored by a timer, is marked as the interaction matching threshold. It is then determined whether the obtained interaction matching threshold falls within the unqualified push range. If the interaction matching threshold falls within the unqualified push range, the current push information is marked as unqualified information, and the interaction information response is optimized in the next preset push time period. The range represents the area where the interaction matching threshold is greater than the specified threshold. The specified threshold is represented by the average value of interaction matching thresholds over a historical time period. The optimization of interaction information response is used to dynamically adapt to the allocation of push resources and improve the accuracy of interaction matching between the vehicle intelligent terminal and the vehicle display terminal. By dynamically matching interaction information and instructions to obtain the interaction matching threshold, and optimizing the interaction information response when the threshold falls within the unqualified range of the push, it helps to correct the matching deviation between information and instructions in a timely manner, improve the accuracy and effectiveness of the interaction response, and further ensure the driver's rapid execution of dispatch instructions.

[0068] Specifically, if the interaction matching value does not fall within the unqualified range for push notifications, the corresponding priority push information is marked as qualified push information, and it is determined whether to display it on the vehicle display terminal. The determination process is as follows: The corresponding qualified push information is input into a pre-created push information priority classification model, and the push information priority classification result is output. The push information priority classification result includes high-level push results and low-level push results, with the urgency of the corresponding qualified push information decreasing. If the output is a high-level push result, the corresponding push notification is displayed on the vehicle display terminal and broadcast via voice. If the output is a low-level push result, the corresponding push notification is displayed on the vehicle display terminal. When the interaction matching value does not fall within the unqualified range for push notifications, the push information priority classification result is obtained based on the qualified push information. When a high-level push result is detected, the corresponding push notification is displayed on the vehicle display terminal and broadcast via voice, which helps to achieve multimodal and accurate delivery of high-priority information and ensures that core content is quickly captured during driving. If a low-level push result is detected, the corresponding push notification is displayed on the vehicle display terminal, which helps to balance information acquisition and driving safety.

[0069] By randomly dividing the dataset of qualified push notifications (such as vehicle arrival notifications, road congestion notifications, road construction notifications, route adjustment notifications, etc.) collected within a specific historical interaction period and the pre-set push emergency results into a training set, the training set is input into a pre-set deep learning model, such as CNN (Convolutional Neural Network), for training to obtain the trained model, namely the push notification priority classification model. Qualified push notifications are then re-inputted into the trained model to output the push notification priority classification results.

[0070] like Figure 4 As shown, this is the push notification priority classification diagram provided by the present invention. As shown in the figure, the input layer receives qualified push notification text, which is converted into a numerical word vector matrix by the Embedding layer. Then, multiple convolutional kernels of size 2, 3, and 4 capture emergency keyword features of different lengths, such as "congestion" and "road closure for construction". The max pooling layer compresses and filters the convolutional features to retain the core emergency signals. The multi-dimensional features after pooling are concatenated and input into the fully connected layer to complete the classification calculation. Finally, the output layer directly gives the push notification priority result (high-level push result or low-level push result).

[0071] It needs to be added that the specific process of optimizing the interactive information response is as follows: The interaction matching division value and the data volume of the push information when the vehicle-mounted intelligent terminal starts pushing, monitored through the communication protocol stack (such as CANoe / CANalyzer tools), are input into a preset time-slot division table. The corresponding qualified time-slot value and the corresponding push priority are output. An adjustment prompt is sent to the preset personnel to adjust the initial time-slot corresponding to the interactive information to the qualified time-slot value, and the newly acquired interaction matching division value is re-monitored. If the new round of interaction matching division value still falls within the unqualified range for push, an alarm prompt is immediately sent. If the new round of interaction matching division value does not fall within the unqualified range for push, the push priority that matches it is used as the standard for push, and the push order of the corresponding push information is matched.

[0072] By adjusting the initial time slice corresponding to the interactive information to a qualified time slice value, it helps to improve the rationality of the processing resource allocation of the in-vehicle intelligent terminal for interactive information, avoid low-priority information from occupying core processing resources, thereby shortening the processing delay of high-priority information, ensuring the real-time performance of interactive response, improving the efficiency of driver-vehicle interaction management, and reducing the risk of scheduling command execution delays caused by improper allocation of processing resources.

[0073] In this embodiment, abnormal task interaction values ​​are obtained by performing dynamic conflict anomaly assessment of information priority. When the abnormal task interaction value is not greater than a specified abnormal task interaction value, dynamic matching of interaction information and instructions is performed. This helps improve the accuracy of information push and the adaptation to the urgency of driving scenarios and driver workload, avoids priority mismatch caused by static classification, and thus achieves high efficiency and accuracy of information transmission in normal scenarios, reducing the interference of unnecessary information on driving focus. When the abnormal task interaction value is greater than the specified abnormal task interaction value, the corresponding interaction information is directly written to the shared memory inside the vehicle intelligent terminal, and the write anomaly situation of the shared memory of the vehicle intelligent terminal continues to be monitored to obtain the write anomaly value. This helps to ensure the smooth rapid transmission channel of core push information in high-priority conflict scenarios, avoids information delay or loss due to resource competition, and thus improves the timeliness and processing efficiency of emergency dispatch instructions.

[0074] By dynamically matching interactive information with instructions when the write anomaly value is no greater than a predefined write anomaly value, it helps improve the adaptation and optimization capabilities of information and driving status under normal shared memory write scenarios, maximizes the use of system resources to improve the interaction effect, and thus ensures the stability and accuracy of information transmission. Conversely, by allocating corresponding buffers for interactive information based on the read buffer, it helps improve the fault tolerance of information under write anomaly scenarios, reduces information loss due to write failure, and thus ensures the complete transmission of core scheduling information. By monitoring the shared memory write anomaly of the vehicle intelligent terminal and the dynamic matching of interactive information and instructions, it helps to achieve closed-loop management of the entire process of information transmission and interaction adaptation, systematically avoiding the dual risks of transmission anomalies and adaptation imbalances.

[0075] Example 3, as an additional technical solution to Example 1 without changing the basic method, involves multiple information pushes, such as high-load driving scenarios in urban expressway confluence areas, and scenarios where regular station notifications and sudden road condition warnings are triggered at the same time. These scenarios require attention to the common needs of interaction efficiency, voice broadcasting, and screen display. The optimization of the interaction information response also includes: monitoring the interaction transmission efficiency when pushes information from the vehicle display terminal to the vehicle display terminal, and obtaining an interaction efficiency evaluation value. The interaction efficiency evaluation value is determined by the proportion of interaction information successfully transmitted from the vehicle intelligent terminal to the vehicle display terminal and confirmed by the vehicle display terminal within a preset interaction time period (e.g., the ratio of the number of times the screen pop-up confirmation is monitored to the total number of pushes via a counter).

[0076] In scenarios where regular station notifications and emergency road condition warnings are triggered at the same time, the core purpose of optimizing the interactive information response is to ensure that low-priority regular station notifications are delivered in an orderly manner without interfering with the transmission of key information, ultimately achieving both interactive efficiency and driving safety in scenarios with multiple information overlaps.

[0077] To improve the accuracy of information transmission without reducing interaction efficiency, the system determines whether the interaction efficiency within a preset interaction time period is qualified based on an interaction efficiency rating value: if the interaction efficiency rating value is greater than the specified interaction efficiency rating value, an interaction efficiency qualified prompt is sent, the corresponding interaction information is marked as qualified interaction information, and notification method matching is enabled; the specified interaction efficiency rating value is represented by the average of the interaction efficiency rating values ​​over historical time periods; if the interaction efficiency rating value is not greater than the specified interaction efficiency rating value, an interaction efficiency unqualified prompt is sent and feedback is sent to the dispatch center.

[0078] The matching of notification methods includes displaying corresponding push notifications on the vehicle display terminal and broadcasting corresponding push notifications via voice, as follows: In the voice channel, the speech rate and volume are dynamically adjusted based on the received interaction information, and voice broadcasts are performed according to the corresponding dispatch center instructions; by using the obtained speech rate and volume in the voice channel as the configuration values ​​for voice broadcasts, it helps to improve the recognizability and auditory comfort of voice information, avoids voice broadcast confusion when multiple pieces of information are superimposed, and thus helps to reduce information omissions caused by improper voice configuration.

[0079] In the display channel, the screen layout is dynamically adjusted based on the received interactive information, and the information is displayed on the vehicle display terminal according to the corresponding dispatch center instructions. By using the width of the push content display area and the font of the push content as the configuration values ​​for display in the display channel, the readability and display rationality of visual information can be improved. This helps to ensure the effective transmission of routine information without interfering with the driver's field of vision, and achieves coordinated optimization of interaction efficiency, voice broadcast and screen display in scenarios with multiple information overlays.

[0080] Specifically, based on the received interaction information, the speech rate and volume are dynamically adjusted to more accurately match the voice broadcast with the needs of the interactive scenario. Specifically, the initial speech rate and initial volume parameters are combined and used as the parameters to be configured in a pre-built voice broadcast parameter allocation table. After importing the parameters to be configured into the pre-built voice broadcast parameter allocation table, the speech rate and volume are output. The obtained speech rate and volume are used as the configuration values ​​for voice broadcasting, and the corresponding adjustment instructions are fed back to the in-vehicle intelligent terminal, which then uploads them to the cloud. Based on the received interaction information, the speech rate and volume are dynamically adjusted. The full-screen layout is designed to adapt to dynamic interaction needs. Specifically, the initial push content display area width and initial push content font size are combined as parameters to be configured in a pre-built display parameter allocation table. After importing the parameters to be configured into the pre-built display parameter allocation table, the newly allocated push content display area width and push content font size are output. The obtained push content display area width and push content font size are used as the configuration values ​​for display, and the corresponding adjustment instructions are fed back to the in-vehicle intelligent terminal, which then uploads them to the cloud.

[0081] In this embodiment, an interaction efficiency rating is obtained by optimizing the interaction information response. When the interaction efficiency rating is greater than a predetermined value, notification method matching is enabled; otherwise, it is enabled. This helps to achieve dynamic adaptation between notification methods and interaction efficiency in scenarios with multiple information overlays. Regardless of the current interaction efficiency, accurate matching can balance the integrity of information transmission and driving safety. The interrelationship between the optimization of the interaction information response and the matching of notification methods helps to form a closed-loop collaborative mechanism of "optimization to evaluation to matching," thereby improving the interaction efficiency when multiple information is pushed, and ensuring driving safety and information transmission accuracy in high-load driving scenarios.

[0082] like Figure 5 The diagram shows a structural schematic of an intelligent interaction system for a vehicle display terminal provided in an embodiment of this application. The system, which includes an intelligent interaction method for a vehicle display terminal, comprises: a load anomaly monitoring module for performing a real-time operation load anomaly assessment based on abnormal situations perceived by the vehicle's state and dynamic conflicts in information priority; a push anomaly assessment module for determining whether to monitor abnormal situations during the push process of the in-vehicle intelligent terminal based on the assessment results; a notification method adaptation module for dynamically selecting a method to adapt to the urgency of push notifications after monitoring is completed, if monitoring is performed; and a delay processing module for reducing the push delay of push information in the in-vehicle intelligent terminal and the in-vehicle display terminal if no monitoring is performed.

[0083] This application provides a vehicle display terminal device, including: a display screen, input buttons, a communication interface, and a processor; the display screen is used to display the working status of the vehicle-mounted intelligent terminal, view dispatch center messages, and view local recordings and videos; the input buttons are used to provide physical interaction between the driver and the display screen; the communication interface is used to access a local storage device and read local data; the processor is used to coordinate the logical operations of the intelligent interaction system and manage the data transmission and reception and protocol parsing of the communication interface.

[0084] The operating principle of the vehicle display terminal device in this application embodiment is as follows: After the device is powered on, the processor completes system self-test and initialization, accesses the local storage device through the communication interface to read data (including local recordings, videos and messages sent by the dispatch center), and receives physical interaction commands from the driver (such as emergency call commands) transmitted by the input buttons; the processor, as the core, coordinates the logical operation of the intelligent interaction system, manages the data transmission and reception and protocol parsing of the communication interface, parses, classifies and executes the received commands, and outputs the processed visualization content (working status, dispatch messages, local audio and video, etc.) to the display screen for display.

[0085] Taking the driver display terminal P10 provided in this application as an example, the driver display terminal P10 is a human-to-human interaction platform between the driver and the vehicle-mounted intelligent terminal. The driver display terminal P10 displays the real-time operating status of the vehicle-mounted intelligent terminal (such as network status, sensor working status, shared memory usage rate, and interaction efficiency evaluation results) through the display screen. It is widely used in commercial vehicles such as buses, passenger vehicles, and long-distance trucks. Its communication interface, such as the M16-8 core aviation connector, is used to connect to local storage devices (such as solid-state drives) and vehicle-mounted intelligent terminals to read local data (such as local video recordings). Its display screen has a resolution of 1280*800 and can perform capacitive touch control. Its core processor may include a CPU.

[0086] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for intelligent interaction of a vehicle display terminal, characterized in that, The method comprises the following steps: Based on the abnormal situation and information priority dynamic conflict situation of vehicle state perception, the evaluation reflecting the real-time operation load situation is carried out, including state perception abnormality evaluation and information priority dynamic conflict abnormality evaluation; Based on the evaluation result, it is determined whether to monitor the abnormal situation of the vehicle intelligent terminal pushing process; If monitoring is carried out, after the abnormal situation monitoring is completed, the result of the abnormal situation monitoring is used to dynamically select the mode of adapting the pushing notification urgency; If monitoring is not carried out, the processing of reducing the pushing delay of the pushing information in the vehicle intelligent terminal and the vehicle display terminal is adopted; The specific process of the information priority dynamic conflict abnormality evaluation is as follows: In the process of receiving and analyzing the dispatch center instruction by the vehicle intelligent terminal and pushing it to the vehicle display terminal for interaction, based on the analysis of the abnormal task interaction situation of the vehicle intelligent terminal, based on the total number of times of receiving the dispatch center instruction by the vehicle intelligent terminal within the preset pushing time period, the abnormal task interaction value is marked; Based on the abnormal task interaction value and the specified abnormal task interaction value, it is judged: If the abnormal task interaction value is not greater than the specified abnormal task interaction value, the waiting pushing instruction is sent, and the interactive information and the instruction are dynamically matched; If the abnormal task interaction value is greater than the specified abnormal task interaction value, the corresponding interactive information is directly written into the shared memory inside the vehicle intelligent terminal, and the shared memory writing abnormality of the vehicle intelligent terminal is monitored from the response level of the vehicle intelligent terminal, and the writing abnormality value is obtained; The writing abnormality value is represented by the ratio of the number of requests of the interactive information that is not successfully written within the preset writing time after the writing request is triggered to the total number of writing requests; If the writing abnormality value is not greater than the pre-specified writing abnormality value, the interactive information and the instruction are dynamically matched; If the writing abnormality value is greater than the pre-specified writing abnormality value, the pushing information of this time is marked as unqualified writing information, and in the next round of preset pushing time period, the writing abnormality value and the data amount corresponding to the interactive information are input into the preset buffer allocation table, the corresponding buffer is read, and the corresponding buffer is allocated for the interactive information; The specific process of the interactive information and the instruction dynamic matching is as follows: From the vehicle intelligent terminal pushing dispatch instruction process level, for the received priority pushing information, the dispatch instruction pushing delay of the vehicle intelligent terminal is evaluated, and the time length corresponding to the time when the vehicle intelligent terminal starts pushing and the time when the vehicle display terminal responds is marked as the interactive matching division value; It is judged whether the obtained interactive matching division value belongs to the pushing unqualified range; If the interactive matching division value belongs to the pushing unqualified range, the pushing information of this time is marked as unqualified information, and in the next round of preset pushing time period, the interactive information response is optimized; The pushing unqualified range represents the range corresponding to the case that the interactive matching division value is greater than the specified interactive matching division value; If the interactive matching division value does not belong to the pushing unqualified range, the corresponding priority pushing information is marked as qualified pushing information, and it is judged whether to display the vehicle display terminal.

2. The intelligent interaction method for a vehicle display terminal according to claim 1, wherein, The evaluation reflecting the real-time operation load condition comprises state-aware abnormality evaluation and information priority dynamic conflict abnormality evaluation. The state-aware abnormality evaluation means obtaining data for evaluating vehicle state-aware collaborative abnormality, and determining whether to process corresponding interaction information. The information priority dynamic conflict abnormality evaluation means obtaining data for measuring task abnormality interaction, and determining whether to process corresponding interaction information.

3. The intelligent interaction method for a vehicle display terminal according to claim 2, wherein, The specific process of the state-aware abnormality evaluation is as follows: Obtain vehicle dynamic signals as quantitative standards of driving scene risks. The vehicle dynamic signals comprise average values of vehicle speeds, average values of accelerations, and average values of yaw rates in a preset vehicle monitoring time period. Count the number of vehicle dynamic signals greater than the upper limit of a corresponding vehicle dynamic qualified range, and mark the number as an upper limit abnormality number. If the upper limit abnormality number is greater than a specified upper limit state category number, immediately send an alarm prompt to a dispatch center. If the upper limit abnormality number is not greater than the upper limit state category number, perform driving scene risk level quantification. The specific process of the driving scene risk level quantification is as follows: Based on data reflecting recognition accuracy in a vehicle-aware state, obtain a value for quantifying a driving scene risk degree, that is, a driving scene risk quantification value. Input the vehicle dynamic signals into a preset machine learning model, and output a predicted driving scene risk quantification value. Determine whether the driving scene risk quantification value is greater than the predicted driving scene risk quantification value. If yes, directly write corresponding push information into shared memory in a vehicle-mounted intelligent terminal.

4. The intelligent interaction method for a vehicle display terminal of claim 2, wherein, If not, send a waiting push instruction to the dispatch center. The state-aware abnormality evaluation further comprises performing adaptive display terminal screen brightness adjustment, inputting illumination intensity into a pre-created screen brightness distribution table, outputting screen brightness, and performing corresponding parameter configuration on the vehicle-mounted display terminal based on the output screen brightness. The specific process of the state-aware abnormality evaluation is as follows: When the driver hand movement monitoring frequency obtained by the vehicle-mounted intelligent terminal is greater than a pre-set hand movement monitoring frequency, monitor the average value of vehicle speeds in a corresponding preset vehicle monitoring time period, mark as a vehicle abnormal state-aware value, and use it as a quantitative standard of driving scene risks in an awareness state abnormality scenario. Perform abnormality evaluation based on the vehicle abnormal state-aware value and a specified driving abnormality-aware value. If the vehicle abnormal state-aware value is greater than the specified driving abnormality-aware value, send a prompt to start a safety information pending interaction process, and after the safety information pending interaction process ends, perform dynamic matching of interaction information and instructions. If the vehicle abnormal state-aware value is not greater than the specified driving abnormality-aware value, send a prompt to enter a safety information interaction mode. The safety information interaction mode specifically comprises marking interaction information corresponding to the vehicle abnormal state-aware value not greater than the specified driving abnormality-aware value as non-urgent information to be pushed, and sending a waiting push instruction. The safety information pending interaction process comprises: Mark the interaction information corresponding to the vehicle abnormal state perception value greater than the specified driving abnormal perception value as interaction abnormal push information; Start state perception and interaction efficiency improvement correction based on interaction abnormal push information: Input the interaction abnormal push information into the pre-set state abnormality prediction model, output the perception abnormality degree level corresponding to the interaction abnormal push information, including the first-level perception abnormality result and the second-level perception abnormality result, and the corresponding perception abnormality degree decreases; When only the first-level perception abnormality result is monitored, the corresponding interaction abnormal push information is directly written into the shared memory corresponding to the vehicle intelligent terminal, and the fixed physical memory address corresponding to the shared memory is directly accessed; When only the second-level perception abnormality result is monitored, the corresponding interaction abnormal push information is reported to the dispatch center asynchronously; When the first-level perception abnormality result and the second-level perception abnormality result are monitored at the same time, the interaction abnormal push information corresponding to the first-level perception abnormality result is processed preferentially.

5. The intelligent interaction method for a vehicle display terminal of claim 1, wherein, Determine whether to perform vehicle display terminal display, and the determination process is as follows: Input the corresponding qualified push information into the pre-created push information priority classification model, output the push information priority classification result, and the push information priority classification result includes high-level push result and low-level push result, and the emergency degree of the corresponding qualified push information decreases; If the output is a high-level push result, display the corresponding push notification through the vehicle display terminal, and broadcast the corresponding push notification through voice; If the output is a low-level push result, display the corresponding push notification through the vehicle display terminal.

6. The intelligent interaction method for a vehicle display terminal according to claim 5, wherein, The specific process of the interaction information response optimization processing is as follows: Input the combination of the interaction matching division value and the data volume of the push information when the vehicle intelligent terminal starts pushing into the pre-set time slice division table, output the corresponding time slice qualified value and the corresponding push priority; Send an adjustment prompt to the pre-set personnel to adjust the initial time slice corresponding to the interaction information to the time slice qualified value, and re-monitor the interaction matching division value obtained in the new round, if the interaction matching division value in the new round still belongs to the push unqualified range, immediately send an alarm prompt, if the interaction matching division value in the new round does not belong to the push unqualified range, the push priority matched therewith is taken as the standard for pushing, and the push sequence of the corresponding push information is matched.

7. The intelligent interaction method for a vehicle display terminal of claim 5, wherein, The interaction information response optimization processing further includes: From the level of vehicle intelligent terminal push information response of the vehicle display terminal, monitor the interaction transmission efficiency when the interaction information is pushed to the vehicle display terminal, and obtain the interaction efficiency evaluation value; The interaction efficiency evaluation value is the proportion of the interaction information pushed by the vehicle intelligent terminal that is successfully transmitted to the vehicle display terminal and confirmed by the vehicle display terminal within the pre-set interaction time period; Determine whether the interaction efficiency within the pre-set interaction time period is qualified according to the interaction efficiency evaluation value: If the interaction efficiency evaluation value is greater than the specified interaction efficiency evaluation value, send an interaction efficiency qualified prompt, mark the corresponding interaction information as qualified interaction information, and start the matching of the notification mode; If the interaction efficiency evaluation value is not greater than the specified interaction efficiency evaluation value, send an interaction efficiency unqualified prompt and feedback to the dispatch center.

8. The intelligent interaction method for a vehicle display terminal of claim 7, wherein, The matching of the notification mode includes displaying the corresponding push notification through the vehicle display terminal and broadcasting the corresponding push notification through voice, specifically as follows: In the voice channel, the speech speed and volume are dynamically adjusted according to the received interaction information, and voice broadcasting is performed according to the corresponding dispatch center instruction; In the display channel, the screen layout is dynamically adjusted according to the received interaction information, and display is performed on the vehicle display terminal according to the corresponding dispatch center instruction; The dynamic adjustment of the speech speed and volume according to the received interaction information is specifically, The parameter combination of the initial broadcasting speed and the initial broadcasting volume is used as the to-be-configured parameter of the pre-constructed voice broadcasting parameter distribution table, the to-be-configured parameter is imported into the pre-constructed voice broadcasting parameter distribution table, and then the broadcasting speed and the broadcasting volume are output; The obtained broadcasting speed and broadcasting volume are used as the configuration value when voice broadcasting is performed, and the corresponding adjustment instruction is fed back to the vehicle-mounted intelligent terminal, which is then uploaded to the cloud by the vehicle-mounted intelligent terminal; The dynamic adjustment of the screen layout according to the received interaction information is specifically, The parameter combination of the initial push content display area width and the initial push content display font size is used as the to-be-configured parameter of the pre-constructed display terminal display parameter distribution table, the to-be-configured parameter is imported into the pre-constructed display terminal display parameter distribution table, and then the newly allocated push content display area width and push content display font size are output; The obtained push content display area width and push content display font size are used as the configuration value when display is performed, and the corresponding adjustment instruction is fed back to the vehicle-mounted intelligent terminal, which is then uploaded to the cloud by the vehicle-mounted intelligent terminal.

9. The system for applying the intelligent interaction method for a vehicle display terminal according to any one of claims 1 to 8, characterized in that, It includes: The load abnormality monitoring module is used to evaluate the real-time operation load condition based on abnormal conditions of vehicle state perception and dynamic conflict conditions of information priority; The push abnormality evaluation module is used to determine whether to monitor the abnormal conditions of the push process of the vehicle-mounted intelligent terminal based on the evaluation result; The notification mode adaptation module is used to dynamically select the mode of adapting the urgency of the push notification according to the result of the abnormal condition monitoring if the monitoring is performed after the abnormal condition monitoring is ended; The delay processing module is used to reduce the push delay degree of the push information in the vehicle-mounted intelligent terminal and the vehicle display terminal if the monitoring is not performed.

10. A vehicle display terminal device, applying the intelligent interaction method for vehicle display terminal as claimed in any one of claims 1-8, characterized in that, It includes a display screen, input keys, a communication interface, and a processor. The display screen is used to display the working state of the vehicle-mounted intelligent terminal, view the dispatch center messages, and view local videos and videos. The input keys are used for physical interaction between the driver and the display screen. The communication interface is used to access the local storage device and read local data. The processor is used to plan the logical operation of the intelligent interaction system and manage the data transmission and protocol analysis of the communication interface.

Images