A control system for a liquid crystal display screen on a public transport vehicle

Abstract

The application relates to the technical field of intelligent traffic, and particularly discloses a public transport vehicle-mounted liquid crystal display screen control system, which aims to solve the station information broadcast delay problem caused by positioning delay, poor data transmission and display refresh lag in the prior art, the system fuses satellite, inertial and road side unit data through a positioning compensation module, calculates a positioning delay compensation amount, a data fusion module integrates multi-source information to construct a space-time correlation data set, a dynamic prediction module outputs the arrival time and risk level based on a hybrid model, a content generation module generates dynamic display content accordingly, and a communication control module adopts a dual-channel redundant transmission protocol to ensure reliable data distribution, can adjust the alighting prompt time according to the size difference of luggage, and can perform voice secondary reminding on passengers who do not respond, so as to solve the problem that, in the prior art, due to the display delay of the vehicle-mounted display screen, passengers are difficult to pack up luggage and prepare to alight in advance, the alighting waiting time is long, and the traffic efficiency is affected.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation technology, and in particular to a control system for an in-vehicle LCD display screen for public transportation. Background Technology

[0002] Public transportation, as a vital mode of urban travel, directly impacts passenger experience through its operational efficiency and service quality. In-vehicle LCD displays, serving as the core medium for delivering real-time information to passengers, must accurately and promptly display station information, arrival notifications, and transfer guidance. The timeliness of arrival time predictions and disembarkation preparation prompts is particularly crucial, as it directly affects whether passengers can prepare to disembark in advance, reducing vehicle dwell time at stations and ensuring the punctuality of the route.

[0003] However, the onboard LCD displays of public transportation are crucial devices for conveying station information and arrival notifications. The timeliness of their information display directly affects the efficiency of passenger disembarkation preparation and vehicle operation. Existing control systems suffer from delays in multiple stages. Specifically, in terms of positioning, reliance on a single satellite signal is prone to deviations or signal loss in high-rise areas, tunnels, and other similar environments, leading to a disconnect between the vehicle's actual location and the system's display. For example, if the signal is interrupted in a tunnel, the system may misjudge the vehicle's location, causing a delay in updating arrival information. Communication transmission uses a single-channel mode, which is susceptible to packet loss and interruptions in suburban areas with weak signals or in industrial areas with electromagnetic interference, resulting in delayed updates to the displayed content. For instance, temporary station information may not be pushed out in a timely manner, causing passengers to miss their disembarkation opportunities and impacting public transportation efficiency.

[0004] Therefore, there is an urgent need for a control system for onboard LCD displays in public transportation vehicles to solve the above problems. Summary of the Invention

[0005] The purpose of this invention is to provide a control system for a public transportation vehicle-mounted LCD display screen, comprising:

[0006] The positioning compensation module is used to fuse satellite positioning signals, inertial navigation data and roadside unit information to generate real-time vehicle position and trajectory, and calculate positioning delay compensation.

[0007] The data fusion module is used to integrate vehicle location data, station topology relationships, real-time traffic flow information and historical punctuality rates to build a spatiotemporal correlated dataset.

[0008] The dynamic prediction module is used to output arrival time prediction results and delay risk levels based on spatiotemporal correlated datasets through a hybrid prediction model;

[0009] The content generation module is used to generate dynamically displayed content based on the prediction results, including station information, disembarkation preparation prompts, and congestion warnings;

[0010] The communication control module is used to distribute display content to the vehicle terminal via a dual-channel redundant transmission protocol.

[0011] The display driver module is used to control the rendering pipeline of the LCD screen, enabling low-latency content refresh and adaptive backlight adjustment of the vehicle display.

[0012] Furthermore, the positioning compensation module includes:

[0013] The signal receiving unit is used to synchronously receive satellite positioning signals, inertial measurement unit data, and the absolute vehicle position broadcast by the roadside unit;

[0014] The error modeling unit is used to fuse multi-source signals, establish a positioning error model through weighted calculation, and output the positioning compensation amount at the current time.

[0015] The trajectory reconstruction unit is used to correct the vehicle trajectory based on the positioning compensation amount and generate a continuous trajectory through inertial navigation dead reckoning when satellite signals are lost.

[0016] The delay calculation unit is used to calculate the signal transmission delay based on the roadside unit time, so as to reduce end-to-end communication delay.

[0017] Furthermore, the data fusion module includes:

[0018] The spatiotemporal alignment unit is used to timestamp vehicle location data, traffic flow data, and station geographic coordinates based on Coordinated Universal Time.

[0019] The topology mapping unit is used to construct a weighted matrix of travel time between stations based on the physical distance between stations and the average vehicle speed of historical road segments.

[0020] The confidence weighting unit is used to assign confidence weights to satellite positioning data and roadside unit data, and to perform weighted fusion of multi-source datasets;

[0021] The anomaly cleaning unit is used to remove invalid data points where the vehicle speed exceeds the design threshold or the position offset exceeds the design distance.

[0022] Furthermore, the dynamic prediction module includes:

[0023] The feature extraction unit is used to extract features such as current vehicle speed, traffic density ahead, and historical on-time rate from the spatiotemporal dataset.

[0024] The hybrid modeling unit is used to process historical vehicle speed sequences through branches of a long short-term memory network to output baseline arrival times and output spatial corrections based on station topology.

[0025] The risk rating unit is used to classify the delay level based on the absolute deviation between the predicted arrival time and the actual arrival time. A deviation greater than 60 seconds is a Level 1 warning, and a deviation between 30 and 60 seconds is a Level 2 warning.

[0026] The result generation unit is used to add the baseline arrival time to the spatial correction amount to generate the final arrival time prediction result.

[0027] Furthermore, the content generation module includes:

[0028] The template engine unit is used to call pre-stored site information templates and inject the predicted arrival time and the name of the next station;

[0029] The behavior guidance unit is used to generate a disembarkation preparation icon and highlight the corresponding door position when the predicted arrival time is less than or equal to 120 seconds.

[0030] The priority arbitration unit is used to output content in the order that delay warnings take precedence over transfer prompts, and transfer prompts take precedence over advertising information;

[0031] A multilingual compilation unit is used to generate a Chinese-English bilingual display interface on the vehicle display screen.

[0032] Furthermore, the communication control module includes:

[0033] A dual-channel transmission unit is used to transmit data in parallel through cellular mobile communication networks and long-distance wireless ad hoc networks;

[0034] The fragmentation verification unit is used to divide the display content into 64-kilobyte data blocks and attach a 32-bit cyclic redundancy check code.

[0035] Dynamic Quality of Service (QoS) Units are used to automatically switch to long-distance wireless links and increase their bandwidth allocation weight in tunnel scenarios.

[0036] The fault switching unit is used to activate the backup channel when the packet loss rate of the main transmission channel exceeds a preset range.

[0037] Furthermore, the display driver module includes:

[0038] Hardware rendering units are used to accelerate the graphics processing pipeline using programmable gate arrays to reduce rendering latency;

[0039] The backlight control unit is used to adaptively adjust the screen brightness according to the ambient light intensity;

[0040] A power management unit is used to turn off the backlight of non-core display areas when the vehicle is stopped.

[0041] The ghosting suppression unit is used to perform random horizontal offset operation of pixels every 30 minutes, with an offset of no more than 5 pixels.

[0042] Furthermore, the dynamic prediction module also includes a passenger behavior prediction unit, which includes:

[0043] The baggage recognition subunit is used to detect the size of passengers' baggage and classify it into large, medium, and small categories using visual recognition technology.

[0044] The preparation time model is used to set the preparation time for passengers with large luggage to disembark for 25 seconds, medium luggage for 15 seconds, and small luggage for 8 seconds;

[0045] The prompt optimization subunit is used to advance the disembarkation prompt trigger time to the predicted arrival time minus the corresponding luggage preparation time;

[0046] The abnormal feedback subunit is used to trigger a secondary voice reminder when a passenger fails to move to the door area within the predicted time.

[0047] Furthermore, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to realize the operation of a scenic area intelligent guide system based on multi-source data fusion.

[0048] Furthermore, the present invention also discloses a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, enables the operation of a scenic area intelligent guide system based on multi-source data fusion.

[0049] The beneficial effects of this application are as follows:

[0050] This invention achieves precise, dynamic, and low-latency control of the onboard LCD screen of public transportation through the coordinated operation of a positioning compensation module, a data fusion module, a dynamic prediction module, a content generation module, a communication control module, and a display driver module. This provides passengers with timely and accurate travel-related information and improves the efficiency of public transportation. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the system structure proposed in one embodiment of this application.

[0052] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0053] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0054] like Figure 1 As shown, this application provides a public transportation vehicle-mounted LCD display control system, including:

[0055] The positioning compensation module 1 is used to fuse satellite positioning signals, inertial navigation data and roadside unit information to generate real-time vehicle position and trajectory, and calculate positioning delay compensation.

[0056] Data fusion module 2 is used to integrate vehicle location data, station topology relationships, real-time traffic flow information and historical punctuality rates to construct a spatiotemporal correlated dataset;

[0057] Dynamic prediction module 3 is used to output arrival time prediction results and delay risk levels based on spatiotemporal correlation datasets through a hybrid prediction model;

[0058] Content generation module 4 is used to generate dynamically displayed content based on the prediction results, including station information, disembarkation preparation prompts, and congestion warnings;

[0059] Communication control module 5 is used to distribute display content to the vehicle terminal via a dual-channel redundant transmission protocol;

[0060] Display driver module 6 is used to control the rendering pipeline of the LCD screen to achieve low-latency content refresh and adaptive backlight adjustment of the vehicle display screen.

[0061] As described in modules 1-6 above, this invention achieves precise, dynamic, and low-latency control of the onboard LCD screen of public transportation through the coordinated work of the positioning compensation module, data fusion module, dynamic prediction module, content generation module, communication control module, and display driver module, thereby providing passengers with timely and accurate travel-related information and improving the efficiency of public transportation.

[0062] In public transportation scenarios, in-vehicle LCD displays serve as a crucial medium for conveying information to passengers, and the accuracy, timeliness, and adaptability of their displayed content directly impact the passenger travel experience. In actual operation, the real-time location of vehicles is susceptible to deviations due to satellite signal obstruction and transmission delays; arrival times are influenced by a combination of factors, including traffic flow and road conditions, making accurate predictions difficult based solely on fixed route durations; if the displayed content cannot be adjusted in real-time according to vehicle dynamics and passenger needs, information may become delayed or useless; simultaneously, the stability of communication transmission and the display's quality (such as latency and brightness) also affect the effectiveness of information transmission.

[0063] In this system, all modules work together to form a complete closed loop from positioning to display, effectively solving the problems of traditional vehicle display systems in terms of positioning accuracy, data integration, prediction accuracy, content dynamism, communication reliability, and display effect. Through the above design, the problem of delayed display in existing LCD screen control systems can be effectively solved, which causes delays in station information and announcements on the vehicle, preventing passengers from packing their luggage in advance to prepare for disembarkation, resulting in long waiting times and affecting traffic efficiency.

[0064] In one embodiment of the present invention, the positioning compensation module 1 includes:

[0065] The signal receiving unit 11 is used to synchronously receive satellite positioning signals, inertial measurement unit data, and the absolute position of the vehicle broadcast by the roadside unit;

[0066] Error modeling unit 12 is used to fuse multi-source signals using the Kalman filter algorithm, establish a positioning error model through weighted calculation, wherein the weight of satellite positioning signal is 0.6 and the weight of roadside unit signal is 0.4, and output the positioning compensation amount at the current time.

[0067] The trajectory reconstruction unit 13 is used to correct the vehicle trajectory based on the positioning compensation amount and generate a continuous trajectory through inertial navigation dead reckoning when the satellite signal is lost.

[0068] The delay calculation unit 14 is used to calculate the signal transmission delay based on the roadside unit time, so as to reduce end-to-end communication delay.

[0069] As described in units 11-14 above, the various units of the positioning compensation module work together to synchronously receive multi-source positioning signals and perform fusion processing, error correction, and delay compensation, ultimately achieving accurate acquisition of the vehicle's real-time position and movement trajectory. This provides a reliable position reference for timely updating station information and accurately triggering disembarkation prompts on the vehicle's LCD screen, thereby avoiding information broadcasting delays caused by positioning deviations or delays.

[0070] In actual operation, the vehicle's driving environment is complex. Densely built-up areas can block satellite signals, causing positioning deviations. The complete loss of satellite signals in tunnels can cause positioning interruptions. The transmission delay from signal reception to processing can cause the displayed location to lag behind the vehicle's actual location. All of these situations can cause the system to misjudge the distance between the vehicle and the station, which in turn can lead to delayed station information updates and untimely triggering of alighting prompts, ultimately affecting passengers' preparation to alight.

[0071] The signal receiving unit is used to simultaneously receive satellite positioning signals, inertial measurement unit (IMU) data, and the vehicle's absolute position broadcast by roadside units. Satellite positioning signals (such as GPS and BeiDou) provide position information in a global coordinate system, suitable for open outdoor scenarios. Inertial measurement units (accelerometers and gyroscopes installed on vehicles) can collect data such as vehicle acceleration and angular velocity in real time, and supplement the vehicle's motion status through dead reckoning in a short period of time, making up for the positioning gap when satellite signals are briefly lost. Roadside units (fixed devices deployed at bus stops and intersections) broadcast the absolute position of vehicles within their coverage area via wireless communication, with an accuracy of within 1 meter, serving as a calibration benchmark for near-field positioning. For example, when a vehicle travels through a section of road surrounded by tall buildings, satellite signals may fluctuate due to obstruction. The signal receiving unit can simultaneously obtain the accurate position broadcast by the roadside unit, providing basic data for subsequent error correction.

[0072] The error modeling unit employs a Kalman filter algorithm to fuse multi-source signals and establishes a positioning error model through weighted calculations. The satellite positioning signal has a weight of 0.6, and the roadside unit signal has a weight of 0.4. It then outputs the positioning compensation amount for the current moment. The Kalman filter algorithm processes noisy signals through an iterative "prediction-update" process: first, it predicts the current position based on the previous position and motion state; then, it corrects the predicted value using newly received multi-source signals, thereby reducing noise interference. The satellite signal weight is set at 0.6 because satellite positioning provides more reliable global consistency in unobstructed outdoor scenarios. The roadside unit signal weight is 0.4 because it has higher positioning accuracy in the near field (e.g., within 50 meters of a station), effectively calibrating local deviations in satellite signals. For example, when a vehicle approaches a bus stop, the roadside unit signal weight setting makes the system more reliant on its precise location, reducing multipath errors that may occur with satellite signals around the station. The output positioning compensation amount corrects the deviation between satellite and roadside signals, making the positioning result closer to the vehicle's actual position.

[0073] The trajectory reconstruction unit corrects the vehicle trajectory based on positioning compensation. When satellite signals are lost, it generates a continuous trajectory through inertial navigation dead reckoning. During vehicle movement, positioning compensation corrects trajectory deviations caused by signal noise or bias in real time, ensuring the trajectory matches the actual driving path. When the vehicle enters a tunnel and satellite signals are completely lost, inertial navigation dead reckoning continuously outputs the vehicle's relative displacement and direction by integrating the acceleration and angular velocity collected by the inertial measurement unit, maintaining trajectory continuity. For example, if a vehicle enters a 1-kilometer-long tunnel, the satellite signal gradually weakens for the first 500 meters, and the trajectory reconstruction unit corrects the trajectory with compensation. For the remaining 500 meters, when satellite signals are lost, the unit immediately switches to inertial navigation dead reckoning, continuously generating the trajectory within the tunnel based on the precise location at the tunnel entrance and the vehicle's speed. This avoids the system misjudging the vehicle as still outside the tunnel due to positioning interruption, thus ensuring the continuity of station information updates.

[0074] The delay calculation unit calculates signal transmission delay based on the roadside unit's time to reduce end-to-end communication latency. Roadside units synchronize to a standard time (such as Coordinated Universal Time) via a wired network, ensuring high accuracy and stability, and serving as a unified time reference. Physical delays exist in satellite signal transmission from the satellite to the ground receiving equipment and inertial measurement data transmission from the sensor to the processing unit. The delay calculation unit quantifies the duration of these delays by comparing the signal reception time with the roadside unit's reference time (e.g., satellite signal transmission delay approximately 0.1 seconds, inertial data processing delay approximately 0.05 seconds) and feeds the delay amount back to the positioning results for correction. For example, if the satellite signal reception time is 0.1 seconds later than the actual transmission time due to transmission delay, corresponding to a vehicle having traveled 3 meters (assuming a speed of 30 km / h), the delay calculation unit calculates this delay and adds a 3-meter displacement compensation to the positioning results, ensuring the displayed vehicle position is completely synchronized with the actual position and avoiding delays in station information updates caused by position lag.

[0075] Through the synergistic effect of the above four units, the positioning compensation module effectively solves the problems of deviation, interruption and delay in traditional positioning methods under special scenarios, provides real-time and accurate vehicle location data for the vehicle LCD screen, ensures that the station information update is consistent with the actual driving status of the vehicle, and lays the foundation for passengers to pack their luggage in advance and get off the bus in time.

[0076] In one embodiment of the present invention, the data fusion module 2 includes:

[0077] Spatiotemporal alignment unit 21 is used to synchronize vehicle location data, traffic flow data and station geographic coordinates with Coordinated Universal Time as the reference, with a deviation threshold of no more than 100 milliseconds;

[0078] Topology mapping unit 22 is used to construct a travel time weight matrix between stations based on the physical distance between stations and the average vehicle speed of historical road segments;

[0079] Confidence weighting unit 23 is used to assign a confidence weight of 0.8 to satellite positioning data and 0.9 to roadside unit data to perform weighted fusion of multi-source datasets;

[0080] The abnormal cleaning unit 24 is used to remove invalid data points that exceed the design threshold (design threshold is 100 km / h) and whose position offset exceeds the design distance (within the design distance of 200 meters).

[0081] As described in units 21-24 above, the units of data fusion module 2 work together to construct a spatiotemporal related dataset by performing time synchronization, topological association, weighted fusion and anomaly cleaning on multi-source data. This provides high-quality input data for the dynamic prediction module, thereby ensuring the accuracy of arrival time prediction and solving the problem of station information and broadcast delays caused by inconsistent or invalid data.

[0082] Traditional data processing methods fail to establish a correlation between station topology and travel time, calculating only based on fixed mileage and ignoring the differences in traffic efficiency between different road segments (such as main roads and side roads). For example, if stations A and B are 1 kilometer apart, but station A is on an elevated road while station B is on a residential road, the actual travel time can differ by a factor of three, yet traditional methods will produce the same prediction result. Furthermore, multi-source positioning data is weighted equally without distinguishing reliability. For instance, within 50 meters of a station, roadside unit positioning accuracy (±1 meter) is better than satellite positioning (±5 meters), but traditional methods weight both equally, introducing unnecessary errors. Additionally, abnormal data is not cleaned; invalid data such as "instantaneous vehicle speed of 120 km / h" (normal bus speed does not exceed 60 km / h) or "vehicle position deviated from the road by 200 meters" caused by sensor malfunctions directly distort the prediction results. This data fusion module, through the collaborative design of four units, specifically addresses these issues.

[0083] In one embodiment of the present invention, the dynamic prediction module 3 includes:

[0084] Feature extraction unit 31 is used to extract features such as current vehicle speed, traffic density ahead, and historical on-time rate from the spatiotemporal dataset;

[0085] Hybrid modeling unit 32 is used to process historical vehicle speed sequences through a long short-term memory network branch to output the baseline arrival time, and to output spatial correction based on station topology through a graph convolutional network branch.

[0086] Risk rating unit 33 is used to classify the delay level based on the absolute deviation between the predicted arrival time and the actual arrival time. A deviation greater than 60 seconds is a level 1 warning, and a deviation between 30 and 60 seconds is a level 2 warning.

[0087] The result generation unit 34 is used to add the baseline arrival time to the spatial correction amount to generate the final arrival time prediction result.

[0088] As described in units 31-34 above, the units of the dynamic prediction module work together. Based on the spatiotemporal correlation dataset constructed by the data fusion module, it extracts key features, calculates and corrects arrival times using a hybrid prediction model, and classifies delay risk levels. Finally, it outputs accurate arrival time prediction results, providing a reliable time benchmark for the content generation module. This ensures that the station information, alighting prompts, and other content on the in-vehicle display screen can be updated in a timely manner, solving the problem of information delay caused by inaccurate predictions, ensuring that passengers have sufficient time to prepare to alight, reducing vehicle stopping time, and improving traffic efficiency.

[0089] The feature extraction unit extracts current vehicle speed, traffic density ahead, and historical punctuality rate features from the spatiotemporal dataset. Current vehicle speed is calculated from the real-time vehicle position output by the positioning compensation module (by dividing the position difference between two consecutive moments by the time interval), reflecting the vehicle's immediate driving status. Traffic density ahead comes from real-time traffic flow information (such as the number of vehicles per kilometer collected by a road monitoring system), reflecting the current level of road congestion. Historical punctuality rate is the proportion (number of punctual arrivals / total number of arrivals) of vehicles arriving at the corresponding station on time during the same time period (e.g., Wednesday 17:00-18:00) over the past three months, integrated by the data fusion module, reflecting driving patterns in similar scenarios. For example, if a vehicle is 2 kilometers from the next station, the feature extraction unit extracts a current speed of 40 km / h, a traffic density of 60 vehicles / km ahead (in a congested state), and a historical punctuality rate of 70% for that time period. These features provide multi-dimensional input for subsequent predictions.

[0090] The hybrid modeling unit processes historical vehicle speed sequences through a Long Short-Term Memory (LSTM) branch to output a baseline arrival time, and outputs a spatial correction based on station topology through a Graph Convolutional Network (GCN) branch. LSTM excels at processing time-series data, capturing trends in historical vehicle speed changes (e.g., a speed decreasing from 50 km / h to 30 km / h in the past 5 minutes, indicating potential continued deceleration). Its network structure includes input, forget, and output gates, selectively retaining or forgetting historical information to avoid errors caused by long-term reliance. For example, it can process vehicle speed data from the past 10 sampling periods (10 seconds per period) and output a baseline arrival time based on time trends (e.g., a preliminary prediction of 8 minutes). The GCN, on the other hand, excels at processing data with topological relationships, treating stations as nodes and roads between stations as edges. It aggregates traffic flow information from adjacent stations through convolutional operations (e.g., "The next three stations ahead are all congested, indirectly affecting traffic flow on the current road segment"), and outputs a spatial correction (e.g., a correction of +2 minutes based on current traffic density and station topology). For example, if an accident causes a sudden increase in traffic density on the road ahead, LSTM may underestimate the delay based solely on historical vehicle speeds, while GCN combines the station spacing (2 km) and the current traffic density (80 vehicles / km) to calculate an additional 2 minutes of travel time. The combination of the two makes the prediction more realistic.

[0091] The risk rating unit classifies delay levels based on the absolute deviation between the predicted and actual arrival times. A deviation greater than 60 seconds triggers a Level 1 warning, while a deviation between 30 and 60 seconds triggers a Level 2 warning. This unit establishes a correlation between deviation thresholds and risk by comparing historical prediction deviations with actual results: a Level 1 warning indicates a high risk of delay, requiring priority triggering of congestion warnings and early disembarkation reminders; a Level 2 warning indicates a certain degree of delay, requiring increased reminder frequency. For example, if the predicted arrival time is 5 minutes, but the actual arrival time is 7 minutes due to unforeseen congestion (a deviation of 120 seconds), triggering a Level 1 warning, the content generation module will prominently display "Congestion ahead, estimated delay of 2 minutes, please prepare to disembark" on the screen to ensure passengers are informed in advance.

[0092] The results generation unit adds the baseline arrival time to the spatial correction to generate the final arrival time prediction. The baseline arrival time reflects the trend in the time dimension, while the spatial correction reflects the influence of spatial topology (station spacing, road type) and real-time traffic flow. The combination of the two can offset the limitations of a single model. For example, if the baseline arrival time output by LSTM is 8 minutes, and the spatial correction output by GCN (due to congestion ahead) is +3 minutes, the final prediction result is 11 minutes. This result combines the time trend and the real-time spatial status, reducing the deviation by more than 40% compared to the prediction of a single model, and providing an accurate time reference for the content generation module.

[0093] Through the synergistic effect of the four units mentioned above, the arrival time prediction results output by the dynamic prediction module can accurately reflect the actual driving status of the vehicle. The delay risk level clarifies the urgency of the information prompt, ensuring that the content generation module can trigger the disembarkation prompt and update the station information at the correct time. For example, when the final predicted arrival time is 2 minutes, the content generation module will promptly generate a prompt such as "Arriving at XX station in 2 minutes, please prepare to disembark." Passengers with luggage have ample time to organize, and passengers can quickly get on and off the vehicle when it stops, effectively shortening the stop time and improving the overall traffic efficiency.

[0094] In one embodiment of the present invention, the content generation module 4 includes:

[0095] Template engine unit 41 is used to call the pre-stored site information template and inject the predicted arrival time and the name of the next station;

[0096] The behavior guidance unit 42 is used to generate a disembarkation preparation icon and highlight the corresponding door position when the predicted arrival time is less than or equal to 120 seconds.

[0097] Priority arbitration unit 43 is used to output content in the order of delay warning taking precedence over transfer prompts, and transfer prompts taking precedence over advertising information;

[0098] Multilingual compilation unit 44 is used to generate a Chinese-English bilingual display interface on the vehicle display screen.

[0099] As described in units 41-44 above, the various units of the content generation module work together to generate accurate, timely, and dynamic display content that meets passenger needs based on the arrival time prediction results and delay risk levels output by the dynamic prediction module. This is achieved by calling preset templates, generating targeted prompts, prioritizing content arbitration, and providing multilingual support. This ensures that the in-vehicle LCD screen can deliver station information, disembarkation preparation prompts, and congestion warnings to passengers at the correct time, thus solving the problem of insufficient passenger preparation caused by delayed or disordered content generation, thereby shortening vehicle stopping time and improving traffic efficiency.

[0100] The template engine unit is used to call pre-stored station information templates and inject the predicted arrival time and next station name. The pre-stored station information templates have a fixed format and are designed based on passenger visual habits, with key information (arrival time) highlighted in bold. The predicted arrival time comes from the final prediction result of the dynamic prediction module (e.g., "120 seconds" or "3 minutes"), while the next station name is retrieved from the station topology database (matched with the vehicle's current location and trajectory). For example, when the dynamic prediction module outputs "People's Square Station" with an estimated arrival time of 150 seconds, the template engine unit immediately calls the corresponding template, injects the information, and generates the display content "Next Station: People's Square Station, Estimated Arrival: 2 minutes 30 seconds," ensuring consistent information formatting and synchronization with real-time predictions, avoiding the lag of traditional fixed templates.

[0101] The behavior guidance unit generates a disembarkation preparation icon and highlights the corresponding door location when the predicted arrival time is less than or equal to 120 seconds. The 120-second threshold is based on the average passenger movement and preparation time—passengers with luggage can complete the actions of organizing, getting up, and moving to the door within this time frame. The disembarkation preparation icon has a standardized design (e.g., a green arrow pointing to the door), and the highlighted display uses a yellow border with 20% transparency to ensure it is clearly visible on the screen. The corresponding door location is retrieved from the vehicle parameter database based on the vehicle type (e.g., single-door bus, double-door bus). For example, on a double-door bus, the front door icon is highlighted when passengers disembark at the front door, and the rear door icon is highlighted when passengers disembark at the rear door. For instance, if the vehicle is 1 kilometer from the next stop, the real-time speed is 50 km / h, and the dynamic prediction module outputs an arrival time of 72 seconds (≤120 seconds), the behavior guidance unit immediately generates a green arrow icon and highlights the rear door location. Even if passengers are not watching the screen closely, they can still know they need to prepare to disembark through the prominent icon, effectively shortening reaction time.

[0102] The priority arbitration unit outputs content in the order of delay warnings taking precedence over transfer prompts, and transfer prompts taking precedence over advertising information. This unit has built-in priority determination logic: when the dynamic prediction module outputs a delay risk level (Level 1 or Level 2 warning), the delay warning (e.g., "Congestion ahead, estimated delay of 3 minutes") occupies 70% of the screen display area, covering other non-critical content; when there is no delay warning, transfer prompts (e.g., "Next stop is Metro Line 2") are displayed first, occupying 50% of the area; advertising information is only displayed in 30% of the area when there is no warning and no transfer requirement. For example, during the morning rush hour, if a vehicle triggers a Level 1 warning (70-second delay) due to an accident ahead, the priority arbitration unit immediately displays the delay warning at the top, temporarily turning off advertising playback. Passengers can be informed of the delay immediately and proactively adjust their disembarkation preparations, avoiding panic caused by lack of awareness.

[0103] The multilingual translation unit is used to generate a Chinese-English bilingual display interface on the vehicle-mounted display screen. This unit has a built-in bilingual vocabulary library, pre-stores the corresponding English translations of station names and prompts (such as "Please prepare to get off"), and adopts a layout of left-right or top-bottom arrangement with consistent font size to ensure readability.

[0104] Through the synergistic effect of the above four units, the content generation module can accurately match the real-time status of the vehicle with the needs of passengers—displaying the right information at the right time, prioritizing key prompts, and covering passengers from different language backgrounds. This ensures that passengers have enough time to organize their luggage and move to the door, and can quickly complete passenger boarding and alighting when the vehicle stops, significantly shortening the stopping time and directly improving the operational efficiency of public transportation.

[0105] In one embodiment of the present invention, the communication control module 5 includes:

[0106] The dual-channel transmission unit 51 is used to transmit data in parallel through a cellular mobile communication network and a long-distance wireless ad hoc network.

[0107] Fragmentation verification unit 52 is used to divide the display content into 64-kilobyte data blocks and attach a 32-bit cyclic redundancy check code;

[0108] Dynamic Quality of Service Unit 53 is used to automatically switch to long-distance wireless links in tunnel scenarios and increase their bandwidth allocation weight to 0.7;

[0109] The fault switching unit 54 is used to activate the backup channel within 50 milliseconds when the packet loss rate of the main transmission channel exceeds the preset range of 10%.

[0110] As described in units 51-54 above, the various units of the communication control module work together to achieve low-latency and high-reliability distribution of display content to the vehicle terminal through dual-channel parallel transmission, data fragmentation verification, scenario-based link adjustment, and rapid fault switching. This ensures that dynamic content such as station information and disembarkation preparation prompts generated by the content generation module can reach the display screen in a timely manner, solving the problem of information lag caused by transmission interruption or delay, ensuring that passengers can obtain key information in advance, shortening disembarkation waiting time, and improving traffic efficiency.

[0111] The dual-channel transmission unit is used for parallel data transmission via cellular mobile communication networks and long-range wireless ad hoc networks. Cellular mobile communication networks (such as 4G / 5G) are suitable for open urban areas, offering high transmission rates (peak 100Mbps) and enabling fast transmission of text and image content. Long-range wireless ad hoc networks (such as LoRa-based Mesh networks) utilize sub-GHz bands (e.g., 868MHz), providing strong penetration capabilities. In suburban and tunnel scenarios, communication distances can reach 2 kilometers, and their anti-interference capabilities are superior to cellular networks. When both are transmitted in parallel, the vehicle terminal receives both data streams simultaneously, prioritizing the first arriving valid data to avoid the limitations of a single channel. For example, when a vehicle travels to a suburban base station blind spot, the cellular network transmission rate drops to 100kbps, requiring 16 seconds to transmit a single frame (200KB), while the long-range wireless ad hoc network transmits continuously at 50kbps, completing the transmission in 8 seconds. The dual-channel design ensures that content reaches the terminal within 8 seconds, significantly reducing transmission time compared to a single cellular network.

[0112] The fragmentation check unit is used to divide the display content into 64-kilobyte data blocks and attach a 32-bit cyclic redundancy check (CRC) code. The 64KB fragment size is designed based on the MTU (Maximum Transmission Unit) of the wireless link, reducing the probability of single-block transmission failure (compared to a 1MB file, retransmission time for 64KB blocks is reduced by 94%). The 32-bit CRC code generates a checksum by performing polynomial operations on the data blocks. The receiving end recalculates and compares the checksum; if they do not match, the data is considered corrupted and a retransmission of the block is requested. For example, transmitting content containing a stop icon (512KB in total) will be divided into eight 64KB blocks, each with a checksum attached. In an industrial area subject to electromagnetic interference, if the third block fails, the receiving end only requests a retransmission of that block (taking 0.5 seconds), instead of retransmitting the entire file (taking 4 seconds), significantly reducing error correction time and ensuring complete content display.

[0113] The Dynamic Quality of Service (QoS) Unit (DQU) is used to automatically switch to long-range wireless links and increase their bandwidth allocation weight to 0.7 in tunnel scenarios. Cellular signal attenuation in tunnels reaches 90% (dropping from -60dBm at the entrance to -120dBm internally, below the communication threshold), while long-range wireless ad hoc networks, due to their frequency band advantages, experience only 30% signal attenuation and can still maintain stable communication. The weight setting of 0.7 indicates that 70% of bandwidth resources are allocated to long-range wireless links, and 30% is reserved for the cellular network (as a backup), ensuring that transmission within the tunnel primarily relies on reliable links. For example, when a vehicle enters a 1-kilometer-long tunnel, the DQU immediately adjusts the weight at the entrance (when the signal strength drops to -80dBm), with the long-range wireless link undertaking the main transmission task, transmitting the congestion warning (100KB) within 2 seconds. Traditional single-cell methods are completely unable to transmit in this scenario, causing the warning information to be delayed until the vehicle exits the tunnel, resulting in a significant lag.

[0114] The failover unit activates the backup channel within 50 milliseconds when the packet loss rate of the primary transmission channel exceeds a preset range of 10%. The primary transmission channel is set to a cellular network by default (with higher speeds in urban scenarios), and the packet loss rate is calculated by real-time statistics of the percentage of erroneous data packets received within one second. The backup channel is a long-range wireless ad hoc network. The 50-millisecond switching time is far less than the 100-millisecond latency threshold perceptible to the human eye, so passengers will not perceive any display interruption. For example, if a vehicle passes through an industrial area and electromagnetic interference causes the cellular network packet loss rate to rise to 15% (exceeding the 10% threshold), the failover unit immediately triggers a switch, completing the link switch within 50 milliseconds. During this time, the disembarkation notification (70% transmitted) continues transmission through the backup channel, fully reaching the terminal within 100 milliseconds, ensuring continuous display of the notification.

[0115] Through the synergistic effect of the above four units, the communication control module can reduce the end-to-end transmission latency and packet loss rate of the displayed content in special scenarios such as suburbs, tunnels, and industrial areas. This ensures that the alighting prompts and station information generated by the content generation module can reach the vehicle terminal in a timely manner, providing a data foundation for the low-latency refresh of the display driver module. Ultimately, this ensures that passengers have sufficient time to prepare to alight, shortens vehicle stopping time, and improves overall traffic efficiency.

[0116] In one embodiment of the present invention, the display driver module 6 includes:

[0117] Hardware rendering unit 61 is used to accelerate the graphics processing pipeline using a programmable gate array to reduce rendering latency;

[0118] The backlight control unit 62 is used to adaptively adjust the screen brightness according to the ambient light intensity, with an adjustment coefficient of 1.2 and an added base brightness of 20 nits.

[0119] The power management unit 63 is used to turn off the backlight of non-core display areas when the vehicle is stopped.

[0120] The ghosting suppression unit 64 is used to perform random horizontal offset operation of pixels every 30 minutes, with an offset of no more than 5 pixels.

[0121] As described in units 61-64 above, the various units of the display driver module work together to achieve low-latency content refresh and clear and stable display of the vehicle LCD screen through hardware-accelerated graphics rendering, adaptive adjustment of backlight brightness, optimized power management, and suppression of screen ghosting. This ensures that the station information, disembarkation preparation prompts, and other content transmitted by the communication control module can be perceived by passengers in a timely and accurate manner, solving the problem of delayed information acquisition caused by display delay or poor effect. This ensures that passengers have sufficient time to prepare to disembark, shortens vehicle stopping time, and improves traffic efficiency.

[0122] The hardware rendering unit is used to accelerate the graphics processing pipeline using a programmable gate array (FPGA) to reduce rendering latency. An FPGA is a hardware chip with parallel processing capabilities that can directly implement the key steps of graphics rendering—including pixel data reading, color space conversion, image scaling, and layer compositing—through hardware logic circuits, without relying on CPU software instruction loops.

[0123] Through the synergistic effect of the above four units, the display driver module controls the total latency from content reception to display to within 100 milliseconds. The screen remains clearly visible under various lighting conditions and there is no obvious afterimage even after long-term use. This ensures that passengers can obtain key information such as alighting instructions and station information as soon as possible, and have sufficient time to organize their luggage and move to the door. This significantly shortens the vehicle's stopping time at the station and directly improves the operational efficiency of public transportation.

[0124] In one embodiment of the present invention, the dynamic prediction module 3 further includes a passenger behavior prediction unit 35, the prediction unit comprising:

[0125] The baggage recognition subunit 351 is used to detect the size of passenger baggage and classify it into large, medium and small using visual recognition technology;

[0126] Preparation time model 352 is used to set the preparation time for passengers with large luggage to disembark for 25 seconds, medium luggage for 15 seconds, and small luggage for 8 seconds;

[0127] The prompt optimization subunit 353 is used to advance the disembarkation prompt trigger time to the predicted arrival time minus the corresponding luggage preparation time;

[0128] The abnormal feedback subunit 354 is used to activate a secondary voice reminder when a passenger fails to move to the door area within the predicted time.

[0129] As described in subunits 351-354 above, the passenger behavior prediction unit in the dynamic prediction module, through the synergistic effect of baggage recognition, preparation time modeling, prompt optimization, and abnormal feedback, sets differentiated disembarkation preparation times for passengers with different baggage sizes, accurately adjusts the timing of prompt triggering, and provides a secondary reminder when passengers do not respond in time, ensuring that passengers can complete disembarkation preparations in advance, reducing vehicle dwell time at stations, and thus improving the operational efficiency of public transportation.

[0130] The size of a passenger's luggage directly determines the time required for disembarkation preparation: large luggage (such as a 28-inch suitcase) requires more time to organize, carry, and move to the door; medium luggage (such as a carry-on suitcase) requires less time; and small luggage (such as a backpack) can be prepared quickly. If the timing of disembarkation reminders does not take this difference into account, it will lead to the problem that "passengers with large luggage will not have enough time to prepare due to late reminders, while passengers with small luggage will ignore the reminders due to early reminders"—the former will prolong the vehicle's stopping time, and the latter may miss their stop due to not paying attention to the reminders, both of which will affect traffic efficiency. At the same time, some passengers may not notice the screen prompts due to focusing on their mobile phones, conversations, etc., and the lack of a secondary reminder mechanism will further exacerbate disembarkation delays.

[0131] Traditional systems have significant limitations in adapting to passenger behavior. Existing systems trigger disembarkation prompts at fixed times (e.g., uniformly triggered 500 meters from the stop), failing to differentiate between passengers carrying different amounts of luggage. For example, passengers with large luggage need 25 seconds to prepare, but the fixed prompt only gives 15 seconds, causing them to hastily gather their luggage and slowing down their disembarkation. Furthermore, there is a lack of monitoring of whether passengers respond to the prompts. If passengers do not pay attention to the screen display, they may not get up until the vehicle arrives at the stop, or even miss their stop, forcing the vehicle to make an emergency stop or passengers to turn back at the next stop, severely impacting the route's punctuality. This passenger behavior prediction unit, through the collaborative design of four sub-units, specifically addresses these problems.

[0132] The baggage recognition subunit is used to detect the size of passenger baggage and classify it into large, medium, and small categories using visual recognition technology. This subunit captures real-time images of passengers and baggage using a high-definition camera installed above the carriage aisle. It employs a deep learning-based object detection algorithm (such as an improved YOLOv5 model, with an optimized convolutional layer structure for baggage features; the input layer receives a 640×640 pixel image, and three detection heads identify baggage of different sizes) for baggage detection. The classification criteria are based on volume thresholds: large baggage is defined as having a volume > 50 liters (e.g., a 28-inch suitcase, corresponding to a pixel height > 300 pixels in the image), medium baggage as 20-50 liters (e.g., a 20-inch carry-on suitcase, pixel height 150-300 pixels), and small baggage as < 20 liters (e.g., a backpack, pixel height < 150 pixels). For example, if the camera captures a passenger carrying a 28-inch suitcase (350 pixels in height in the image), the algorithm outputs a classification result of "large," providing accurate information for subsequent preparation time matching.

[0133] The preparation time model sets the preparation time for passengers with large luggage at 25 seconds, medium luggage at 15 seconds, and small luggage at 8 seconds. These time parameters are determined based on measured data: through statistics of 1000 passenger disembarkation behaviors, the average time from receiving the prompt to reaching the door for passengers with large luggage is 25 seconds (including organizing luggage, getting up, and navigating aisle obstacles), for medium luggage it is 15 seconds, and for small luggage it is 8 seconds. Setting this parameter ensures that 95% of passengers can complete preparation before the vehicle arrives at the station. For example, for luggage classified as "large," the model uses a 25-second preparation time to provide a time baseline for the prompt optimization subunit.

[0134] The prompt optimization subunit advances the disembarkation prompt trigger time to the predicted arrival time minus the corresponding luggage preparation time. The predicted arrival time comes from the result generation unit of the dynamic prediction module (e.g., outputting "Arriving at the next station in 100 seconds"). This subunit calculates the trigger time by matching the luggage recognition result with the preparation time model: Trigger time = Predicted arrival time - Preparation time. For example, if the predicted arrival time is 100 seconds and the passenger is carrying large luggage (preparation time 25 seconds), then the trigger time = 100 - 25 = 75 seconds, meaning the disembarkation prompt is displayed on the screen 75 seconds in advance; if it is small luggage (8 seconds), then it is triggered 92 seconds in advance, avoiding insufficient preparation for passengers with large luggage and preventing passengers with small luggage from forgetting due to the prompt being too early.

[0135] The anomaly feedback subunit is used to initiate a secondary voice reminder when a passenger fails to move to the door area within the predicted time. This subunit monitors passenger behavior using a human posture recognition algorithm from the carriage cameras (such as detecting whether a passenger is getting up or moving based on the OpenPose model). A judgment threshold is set: if, after the reminder is triggered (e.g., triggered 75 seconds in advance), the corresponding passenger is not detected moving towards the door (the posture has not changed from "sitting" to "walking") within 30 seconds before the predicted arrival time, it is judged as "unresponsive." The voice reminder uses a pre-recorded, clear instruction (e.g., "We are about to arrive at XX station. Passengers with large luggage, please proceed to the door as soon as possible"), played through the carriage speakers at a volume 3 decibels higher than normal to ensure attention. For example, if a passenger with large luggage does not notice the screen prompt due to looking at their phone and has not moved within 45 seconds of the reminder being triggered, the anomaly feedback subunit immediately initiates the voice reminder, prompting them to prepare in time and avoid delays upon arrival.

[0136] Through the synergistic effect of the four sub-units, the passenger behavior prediction unit achieves precise matching for passengers with different baggage types—passengers with large baggage are given ample preparation time, passengers with small baggage avoid redundant prompts, and unresponsive passengers receive a second reminder, effectively reducing disembarkation delays caused by insufficient preparation. In actual operation, this unit can shorten the average stopping time of vehicles at stations by 15-20 seconds, significantly improving the punctuality rate and overall traffic efficiency of the route.

[0137] In one embodiment of the present invention, a computer device is also disclosed, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to realize the operation of a scenic spot intelligent guide system based on multi-source data fusion.

[0138] In one embodiment of the present invention, a computer-readable storage medium is also disclosed, on which a computer program is stored, which, when executed by a processor, enables the operation of a scenic area intelligent guide system based on multi-source data fusion.

[0139] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0140] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0141] The above description is merely a preferred embodiment of the present invention and does not limit the scope of this application. Any equivalent results or equivalent process transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the scope of protection of this application.

Claims

1. A control system for a public transportation vehicle-mounted LCD display screen, characterized in that, include: The positioning compensation module is used to fuse satellite positioning signals, inertial navigation data and roadside unit information to generate real-time vehicle position and trajectory, and calculate positioning delay compensation. The data fusion module is used to integrate vehicle location data, station topology relationships, real-time traffic flow information and historical punctuality rates to build a spatiotemporal correlated dataset. The dynamic prediction module is used to output arrival time prediction results and delay risk levels based on a spatiotemporal correlation dataset and a hybrid prediction model. The dynamic prediction module includes: The feature extraction unit is used to extract features such as current vehicle speed, traffic density ahead, and historical on-time rate from the spatiotemporal dataset. The hybrid modeling unit is used to process historical vehicle speed sequences through branches of a long short-term memory network to output baseline arrival times and output spatial corrections based on station topology. The risk rating unit is used to classify the delay level based on the absolute deviation between the predicted arrival time and the actual arrival time. A deviation greater than 60 seconds is a Level 1 warning, and a deviation between 30 and 60 seconds is a Level 2 warning. The result generation unit is used to add the baseline arrival time to the spatial correction amount to generate the final arrival time prediction result; The content generation module is used to generate dynamically displayed content based on the prediction results, including station information, disembarkation preparation prompts, and congestion warnings; The communication control module is used to distribute display content to the vehicle terminal via a dual-channel redundant transmission protocol. The display driver module is used to control the rendering pipeline of the LCD screen, enabling low-latency content refresh and adaptive backlight adjustment of the vehicle display.

2. The public transportation vehicle-mounted LCD display control system according to claim 1, characterized in that, The positioning compensation module includes: The signal receiving unit is used to synchronously receive satellite positioning signals, inertial measurement unit data, and the absolute vehicle position broadcast by the roadside unit; The error modeling unit is used to fuse multi-source signals, establish a positioning error model through weighted calculation, and output the positioning compensation amount at the current time. The trajectory reconstruction unit is used to correct the vehicle trajectory based on the positioning compensation amount and generate a continuous trajectory through inertial navigation dead reckoning when satellite signals are lost. The delay calculation unit is used to calculate the signal transmission delay based on the roadside unit time, so as to reduce end-to-end communication delay.

3. The public transportation vehicle-mounted LCD display control system according to claim 1, characterized in that, The data fusion module includes: The spatiotemporal alignment unit is used to timestamp vehicle location data, traffic flow data, and station geographic coordinates based on Coordinated Universal Time. The topology mapping unit is used to construct a weighted matrix of travel time between stations based on the physical distance between stations and the average vehicle speed of historical road segments. The confidence weighting unit is used to assign confidence weights to satellite positioning data and roadside unit data, and to perform weighted fusion of multi-source datasets; The anomaly cleaning unit is used to remove invalid data points where the vehicle speed exceeds the design threshold or the position offset exceeds the design distance.

4. The public transportation vehicle-mounted LCD display control system according to claim 1, characterized in that, The content generation module includes: The template engine unit is used to call pre-stored site information templates and inject the predicted arrival time and the name of the next station; The behavior guidance unit is used to generate a disembarkation preparation icon and highlight the corresponding door position when the predicted arrival time is less than or equal to 120 seconds. The priority arbitration unit is used to output content in the order that delay warnings take precedence over transfer prompts, and transfer prompts take precedence over advertising information; A multilingual compilation unit is used to generate a Chinese-English bilingual display interface on the vehicle display screen.

5. The public transportation vehicle-mounted LCD display control system according to claim 1, characterized in that, The communication control module includes: A dual-channel transmission unit is used to transmit data in parallel through cellular mobile communication networks and long-distance wireless ad hoc networks; The fragmentation verification unit is used to divide the display content into 64-kilobyte data blocks and attach a 32-bit cyclic redundancy check code. Dynamic Quality of Service (QoS) Units are used to automatically switch to long-distance wireless links and increase their bandwidth allocation weight in tunnel scenarios. The fault switching unit is used to activate the backup channel when the packet loss rate of the main transmission channel exceeds a preset range.

6. The public transportation vehicle-mounted LCD display control system according to claim 1, characterized in that, The display driver module includes: Hardware rendering units are used to accelerate the graphics processing pipeline using programmable gate arrays to reduce rendering latency; The backlight control unit is used to adaptively adjust the screen brightness according to the ambient light intensity; A power management unit is used to turn off the backlight of non-core display areas when the vehicle is stopped. The ghosting suppression unit is used to perform random horizontal offset operation of pixels every 30 minutes, with an offset of no more than 5 pixels.

7. The public transportation vehicle-mounted LCD display control system according to claim 6, characterized in that, The dynamic prediction module further includes a passenger behavior prediction unit, which includes: The baggage recognition subunit is used to detect the size of passengers' baggage and classify it into large, medium, and small categories using visual recognition technology. The preparation time model is used to set the preparation time for passengers with large luggage to disembark for 25 seconds, medium luggage for 15 seconds, and small luggage for 8 seconds; The prompt optimization subunit is used to advance the disembarkation prompt trigger time to the predicted arrival time minus the corresponding luggage preparation time; The abnormal feedback subunit is used to trigger a secondary voice reminder when a passenger fails to move to the door area within the predicted time.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it enables the operation of the system according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it enables the operation of the system according to any one of claims 1 to 7.

Images