A side window advertisement display management system and method for a vehicle

Abstract

The present invention discloses a side window advertisement display management system and method for a vehicle, specifically related to the field of intelligent bus advertisement systems, which is used to solve the problems of precise advertisement matching and seamless display in the context of dynamic driving. By means of multi-source data fusion and Kalman filtering technology, vehicle driving state data is accurately extracted, and a deep learning model is used for feature analysis to generate driving context tags in real time. Based on these tags, the system intelligently selects advertisements that highly match the current context, analyzes the duration and robustness in combination with historical driving data, and uses a reinforcement learning algorithm to optimize the advertisement selection strategy. Finally, through a pre-caching mechanism and a seamless switching algorithm, it is ensured that the optimal advertisement content can be dynamically loaded and stably displayed on the bus side window display, realizing the continuity and smoothness of advertisement playback. It significantly improves the relevance and efficiency of advertisement placement, enhances the riding experience of passengers and the advertisement dissemination effect.

Description

Technical Field

[0001] The present invention relates to the field of intelligent bus advertising systems, and more particularly, to a vehicle side window advertising display management system and method. Background Art

[0002] With the popularization of public transportation and the development of information technology, buses, subways, high-speed rails, etc. have gradually become important tools for passengers to travel. To improve the riding experience, the side window advertising system of buses has emerged as a new type of information dissemination method. Side window advertising not only brings commercial value to buses but also provides convenient information services for passengers. Through in-vehicle displays, buses can show dynamic information such as station information, traffic conditions, and weather warnings to passengers. At the same time, the development of intelligent technology enables the real-time perception of the vehicle's driving state and changes in the in-vehicle environment, providing new opportunities for the precise push of advertising content. However, most existing bus side window advertising systems rely on fixed advertising libraries, with preset advertising content that does not change with the driving state and lacks real-time linkage, resulting in a low fit between the advertisement and the driving situation, and a decline in advertising effectiveness and passenger attention. Therefore, intelligently adjusting the side window advertising content based on the real-time driving state of buses has become a key technical requirement for improving advertising benefits and enhancing the passenger experience.

[0003] To solve the above problems, a technical solution is provided as follows. Summary of the Invention

[0004] To overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a vehicle side window advertising display management system and method. By integrating multi-source data fusion and Kalman filtering technology, the vehicle driving state data is accurately captured, and feature extraction is performed in combination with a deep learning model to realize real-time driving situation recognition; based on the generated situation tags, candidate advertisements with highly matching semantics are intelligently selected from the tagged advertising content library, and at the same time, the duration and robustness of historical driving data are analyzed, and the advertising selection strategy is optimized through a reinforcement learning algorithm to ensure that the selected advertisement is the most suitable in the current situation; the selected optimal advertisement is then dynamically loaded and stably displayed on the bus side window display through an advanced pre-caching mechanism and seamless switching algorithm, ensuring the continuity and smoothness of advertisement playback; realizing precise data processing, intelligent situation perception and optimized advertisement selection, as well as efficient dynamic loading and display, forming a highly coordinated and adaptable advertising delivery system; not only improving the relevance and display effect of advertisements, but also ensuring the stable presentation of advertisements in a dynamic driving environment, significantly enhancing the visual experience of passengers and the effectiveness of advertisement dissemination, so as to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] S1. Based on the multi-source data fusion technology, extract the comprehensive data stream of the vehicle, and ensure the accuracy of the driving state data through Kalman filter denoising processing.

[0007] S2. Use the deep learning model to extract features from the driving state data, establish a driving situation recognition model by combining acceleration and steering features, and generate real-time situation labels for the vehicle.

[0008] S3. Based on the driving situation labels, screen out candidate advertisements with multi-dimensional semantic matching from the labeled advertisement content library, and optimize the advertisement selection strategy through the reinforcement learning model in combination with the duration robustness of the driving state, so as to achieve situation-driven advertisement selection and obtain the optimal advertisement.

[0009] S4. Adopt the pre-caching mechanism and seamless switching algorithm to dynamically load in real time and display the optimal advertisement content on the side window display of the bus.

[0010] In a preferred embodiment, step S1 includes the following:

[0011] S1.1. First, integrate the data from different sensors to form a comprehensive data stream describing the driving state of the vehicle; for each data source, perform standardization processing first to eliminate the influence between different dimensions and magnitudes.

[0012] S1.2. Adopt a multi-dimensional adaptive fusion matrix, and dynamically adjust the weights of each data source according to the changes in the real-time environment and vehicle state, including the following steps:

[0013] First, perform standardization processing on the spatial positioning data, attitude angle data, and acceleration data.

[0014] Then, set initial weights for each data source.

[0015] Next, calculate the correlation between each data source through the state correlation function, and the evaluation of the state correlation function is based on the mutual information in information theory.

[0016] Finally, quantify the correlation degree between each data source through the calculation of mutual information; update the weight matrix according to the correlation of each data source.

[0017] S1.3. Use the updated weight matrix to perform weighted combination on the standardized data to obtain a comprehensive driving state vector, which reflects the current driving state of the vehicle.

[0018] In a preferred embodiment, step S1 further includes the following:

[0019] S1.4. Use non-linear Kalman filter to optimize and filter the fused driving state vector, including the following two stages:

[0020] Prediction stage: Based on the state at the previous moment, use the state transition matrix to make a prediction: ; where is the predicted vehicle driving state, is the state transition matrix, describing the state dynamics of the vehicle.

[0021] Update stage: Combine the observed data at the current moment , and optimize the prediction result through the update formula: ; ; where is the Kalman gain, is the prediction error covariance, is the observation matrix, is the observation noise covariance, is the observed data.

[0022] In a preferred embodiment, step S2 includes the following:

[0023] S2.1, Based on the driving state vector, use a convolutional neural network to extract short-term feature fluctuations and capture local features in the vehicle driving state.

[0024] S2.2, Each feature in the driving state vector has temporal correlation. Use a long short-term memory network to mine the temporal features of the driving state.

[0025] S2.3, After obtaining the fused feature vector, map the corresponding features to a high-dimensional context space, capture the feature relationship through non-linear mapping, and finally generate a real-time driving context label.

[0026] In a preferred embodiment, step S3 includes the following:

[0027] S3.1, In the labeled advertisement content library, retrieve all advertisements related to the current driving context label to form a preliminary advertisement set; use a deep semantic embedding model to map the driving context label and the advertisement label vector to a high-dimensional semantic space to obtain semantic embedding vectors and ; Similarity calculation function: ; where represents the inner product of vectors, represents the norm of the vector; represents the semantic similarity between the th advertisement and the current context label; Sort the candidate advertisements in descending order according to the similarity value, and select the advertisements with similarity higher than the similarity threshold to form a candidate advertisement set.

[0028] In a preferred embodiment, step S3 further includes the following:

[0029] S3.2. The bus travels on a fixed route, and the duration of the driving state in different sections is regular; the playing duration of the advertisement needs to match the expected duration in the current driving scenario to ensure that the advertisement can be played completely without affecting the passenger experience; collect the historical driving data of the bus under the same route, the same section, and the same driving scenario label to obtain a duration sequence ; Use information entropy to evaluate the robustness of the duration; if the information entropy is less than or equal to the corresponding threshold, it is considered that the duration has sufficient robustness; use kernel density estimation to obtain the probability density function of the duration; determine the expected duration as the value corresponding to the peak of the probability density function , that is, the duration of the maximum likelihood estimation.

[0030] In a preferred embodiment, step S3 further includes the following:

[0031] S3.3. For each advertisement in the candidate advertisement set, the corresponding playing duration is ; Define an advertisement duration matching degree function: ; is the time scale parameter of the duration sequence , calculated according to the mutual information of the data; the function is defined as: ; Ensure that the advertisement duration does not exceed the expected duration; sort the advertisements in descending order according to the advertisement duration matching degree, and select the advertisements with a matching degree higher than the corresponding threshold to form the final candidate advertisement set.

[0032] In a preferred embodiment, step S3 further includes the following:

[0033] S3.4. Use reinforcement learning to drive the optimization of advertisement selection; select the optimal advertisement from the final candidate advertisement set for placement; through the reinforcement learning model, in each state, select the action with the maximum Q value according to the Q value table, that is, select the advertisement that best matches the state in the current candidate advertisement set, and mark it as the optimal advertisement.

[0034] A vehicle side window advertisement display management system includes: a data fusion module, a context recognition module, an advertisement selection module, and a dynamic loading module.

[0035] Data fusion module: Based on multi-source data fusion technology, extract the comprehensive data stream of the vehicle, and ensure the accuracy of the driving state data through Kalman filter denoising processing, and then transfer the processed data to the context recognition module.

[0036] Situation recognition module: Use a deep learning model to extract features from the fused driving state data, establish a driving situation recognition model by combining acceleration and steering features, generate real-time situation labels of the vehicle, and pass the labels to the advertisement selection module.

[0037] Advertisement selection module: Screen out candidate advertisements with multi-dimensional semantic matching from the labeled advertisement content library based on the real-time situation labels. Combine the duration robustness of the driving state, optimize the advertisement selection strategy through a reinforcement learning model, and finally obtain the optimal advertisement and pass it to the dynamic loading module.

[0038] Dynamic loading module: Adopt a pre-caching mechanism and a seamless switching algorithm to dynamically load in real-time and accurately display the optimal advertisement content on the bus side window display, ensuring the continuity and smoothness of the advertisement playback.

[0039] Technical effects and advantages of a side window advertisement display management system and method for a means of transportation according to the present invention:

[0040] By integrating multi-source data fusion technology and a deep learning model, the present invention extracts and processes vehicle driving state data, constructs a real-time driving situation recognition system, and thus generates dynamic situation labels. Based on these situation labels, advertisements highly matching the current driving state can be intelligently screened out from the labeled advertisement content library. By combining historical driving state data to analyze the duration and robustness of the driving state, the advertisement selection strategy is continuously optimized through a reinforcement learning algorithm to achieve situation-driven precise advertisement placement. Finally, a pre-caching mechanism and a seamless switching algorithm are adopted to ensure that the optimal advertisement content can be dynamically loaded in real-time and stably displayed on the bus side window display, guaranteeing the continuity and smoothness of the advertisement playback. It not only improves the relevance and efficiency of advertisement placement, but also realizes a high degree of matching between the advertisement content and the driving situation through intelligent data processing and optimization algorithms, significantly enhancing the visual experience of passengers and the advertisement dissemination effect, reflecting the technological innovation and practical value of the system in the field of intelligent advertisement placement. Brief Description of the Drawings

[0041] Figure 1 It is a schematic flow chart of a side window advertisement display management method for a means of transportation according to the present invention.

[0042] Figure 2 It is a schematic structural diagram of a side window advertisement display management system for a means of transportation according to the present invention. Detailed Embodiments

[0043] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0044] Embodiment 1: Figure 1 A method for managing the display of side window advertisements of a vehicle according to the present invention is provided, including:

[0045] S1. Based on multi-source data fusion technology, extract the comprehensive data stream of the vehicle, and ensure the accuracy of the driving state data through Kalman filter denoising processing.

[0046] S2. Extract features from the driving state data through a deep learning model, establish a driving situation recognition model by combining acceleration and steering features, and generate real-time situation labels of the vehicle.

[0047] S3. Based on the driving situation labels, screen out candidate advertisements with multi-dimensional semantic matching from the labeled advertisement content library, and optimize the advertisement selection strategy through a reinforcement learning model in combination with the duration robustness of the driving state, so as to achieve situation-driven advertisement selection and obtain the optimal advertisement.

[0048] S4. Adopt a pre-caching mechanism and a seamless switching algorithm to dynamically load in real time and display the optimal advertisement content on the side window display of the bus.

[0049] During the operation of the bus, accurately obtaining and parsing the spatial positioning, attitude angle, and motion acceleration information of the vehicle is of great significance, especially in the context of realizing dynamic switching of advertisement content. Step S1 ensures the accuracy and stability of the vehicle driving state data through data fusion and processing methods, providing a high-quality data basis for subsequent situation recognition and advertisement matching.

[0050] Step S1 includes the following:

[0051] S1.1. First, integrate the data from different sensors to form a comprehensive data stream that comprehensively describes the driving state of the vehicle. Specifically, the fused data includes but is not limited to:

[0052] Spatial positioning information: reflecting the geographical location and trajectory of the vehicle.

[0053] Attitude angle information: describing the pitch angle, roll angle, etc. of the vehicle, reflecting the tilt and steering state of the vehicle.

[0054] Motion acceleration information: reflecting the linear acceleration and angular acceleration of the vehicle, indicating the rate of change of the vehicle's motion.

[0055] For each data source, perform standardization first to eliminate the influence between different dimensions and magnitudes, and ensure the effective fusion of each data source under the same dimension.

[0056] S1.2. To improve the accuracy of the fused data, use a multi-dimensional adaptive fusion matrix. This algorithm dynamically adjusts the weights of each data source according to the changes in the real-time environment and vehicle state, so as to ensure that the data sources with higher correlation account for a larger proportion in the fusion process. This process includes the following steps:

[0057] Initial data standardization: First, standardize the spatial positioning data , attitude angle data and acceleration data .

[0058] Initial weight setting: Set an initial weight for each data source , which represents the importance of spatial positioning, attitude angle, and motion acceleration data in the preliminary fusion stage.

[0059] State correlation evaluation: Calculate the correlation between each data source through the state correlation function. The evaluation of the state correlation function is based on the mutual information in information theory ; , where and represent the observed values of different data sources at time respectively. Through the calculation of mutual information, the correlation degree between each data source can be quantified.

[0060] Weight update: Dynamically update the weight matrix according to the correlation of each data source. The specific weight update formula is: ; where is the correlation vector at the current moment, is the initial weight, is the number of data sources, represents the index or number of different data sources.

[0061] S1.3. Use the updated weight matrix to perform weighted combination on the standardized data to obtain a comprehensive driving state vector , which contains information such as the spatial positioning, attitude angle, and motion acceleration of the vehicle, and can accurately reflect the current driving state of the vehicle. The specific calculation formula is: ; where, is the comprehensive vehicle driving state vector, which contains the results of multiple data sources after weighting.

[0062] S1.4. To further improve the stability and accuracy of the data, nonlinear Kalman filtering is used to optimize and filter the fused driving state vector. This process includes the following two stages:

[0063] Prediction stage: Based on the state at the previous moment, use the state transition matrix to make a prediction: ; where is the predicted vehicle driving state, is the state transition matrix, which describes the state dynamics of the vehicle.

[0064] Update stage: Combine the observation data at the current moment , and optimize the prediction result through the update formula: ; ; where is the Kalman gain, is the prediction error covariance, is the observation matrix, is the observation noise covariance, is the observation data.

[0065] Through this process, Kalman filtering can effectively minimize noise and errors, and further improve the accuracy of the driving state data.

[0066] In step S1, by constructing a multi-source data fusion framework, adopting the multi-dimensional adaptive fusion matrix (MDAFM) algorithm and the nonlinear Kalman filtering optimization method, the high-precision real-time extraction of the vehicle driving state is achieved, providing reliable data support for the bus advertising system. The multi-dimensional adaptive fusion matrix dynamically adjusts the fusion weights according to the real-time state correlation of each data source. Through correlation evaluation and adaptive weight adjustment, the positioning, attitude angle, and motion acceleration data can be optimally fused in a changing environment to form an accurate driving state vector. Combined with nonlinear Kalman filtering, the fusion result is predicted and corrected, effectively eliminating sensing noise and errors, thereby further improving the stability and robustness of the driving state data. The combination of multi-source fusion and dynamic filtering ensures the dynamic and accurate perception of the vehicle driving state, not only enabling the advertising content to be precisely and intelligently adjusted based on the real-time state of the vehicle, but also laying a solid technical foundation for subsequent driving scenario recognition and intelligent matching of advertising content.

[0067] Step S2 includes the following:

[0068] S2.1. Based on the driving state vector, a convolutional neural network is used to extract short-term feature fluctuations and capture local features in the vehicle driving state.

[0069] Convolution operation: Use a one-dimensional convolutional kernel to perform a sliding calculation on the driving state vector to extract local features within each time period, such as sudden acceleration changes or rapid changes in angular velocity within a short period of time.

[0070] Convolution layer output: The feature matrix generated by the convolution layer can capture the instantaneous feature changes of the driving state vector. For example, when the vehicle makes a sharp turn, the lateral acceleration will increase significantly, and the feature matrix output by the convolution layer will highlight this change.

[0071] S2.2. Each feature in the driving state vector has temporal correlation. State changes such as acceleration, deceleration, and sharp turns have a certain persistence and order. Therefore, a long short-term memory network is used to further mine the temporal features of the driving state.

[0072] Long short-term memory network input: Input the feature matrix output by the convolution layer into the long short-term memory network. The long short-term memory network will extract the dependencies in the time series based on the state changes at previous and subsequent moments. For example, the acceleration changes when the vehicle is going uphill and turning have an obvious dependence, and the long short-term memory network can capture this change pattern.

[0073] Feature fusion: The output hidden state vector of the long short-term memory network fuses multiple temporal features into a high-dimensional feature representation, enabling the state changes at different time points to be expressed in a continuous manner, providing a basis for subsequent situation recognition.

[0074] S2.3. After obtaining the fused feature vector, further map the corresponding features to a high-dimensional situation space, capture more complex feature relationships through non-linear mapping, and finally generate real-time driving situation labels.

[0075] Non-linear mapping: Based on the kernel function, map the fused feature vector to a high-dimensional space to reveal the non-linear relationships between features. For example, when the acceleration and steering angular velocity both increase significantly at the same time, the system may determine that the vehicle is in a "sharp turn" situation, and the non-linear mapping can capture these complex relationships.

[0076] Situation label generation function: Based on the mapped high-dimensional feature space, define a situation label generation function to classify the current state of the vehicle into specific situation labels such as "rapid acceleration", "rapid deceleration", "slow turn", "sharp turn", "uphill / downhill", etc.

[0077] Step S2 precisely constructs real-time driving situation labels for the vehicle through feature analysis based on the driving state vector and context recognition of the deep learning model, laying a foundation for subsequent intelligent advertisement selection. In this process, the system comprehensively utilizes the convolutional neural network to extract features of short-term state changes and the recurrent neural network to capture time series features, deeply integrating multi-dimensional data such as the vehicle's acceleration, steering angular velocity, and tilt angle to generate a high-dimensional context feature vector. Through the context label generation mechanism, the system realizes the transformation from dynamic and complex driving state data to clear context labels, providing accurate input for advertisement placement in different contexts.

[0078] Step S3 includes the following:

[0079] S3.1, In the labeled advertisement content library, retrieve all advertisements related to the current driving situation label to form a preliminary advertisement set.

[0080] Each advertisement in the advertisement content library is assigned a multi-dimensional attribute label vector, which includes various information such as theme, audience, applicable context, etc.

[0081] Using a deep semantic embedding model (such as Transformer or BERT), map the driving situation label and the advertisement label vector to a high-dimensional semantic space to obtain semantic embedding vectors and .

[0082] Similarity calculation function: ; where represents the inner product of vectors, represents the norm of the vector; represents the th advertisement's semantic similarity to the current context label. Sort the candidate advertisements in descending order according to the similarity value, and select the advertisements with similarity higher than the similarity threshold to form a candidate advertisement set.

[0083] S3.2, The bus travels on a fixed route, and the duration of the driving state in different sections has a certain regularity.

[0084] The playing duration of the advertisement needs to match the expected duration of the vehicle in the current driving situation to ensure that the advertisement can be played completely without affecting the passenger experience.

[0085] Collect the historical driving data of the bus on the same route, in the same section, and with the same driving situation label to obtain a duration time series .

[0086] Use information entropy to evaluate the robustness of the duration: ; where is the probability of the duration data within the discrete interval , that is , is the number of data within this interval, is the total amount of data.

[0087] The information entropy reflects the uncertainty of the duration. The smaller the information entropy value, the more concentrated the duration distribution and the higher the robustness.

[0088] If the information entropy is less than or equal to the corresponding threshold, it is considered that the duration has sufficient robustness.

[0089] The probability density function of the duration is obtained by kernel density estimation , using the Gaussian kernel function : ; is the bandwidth parameter, which affects the smoothness of the estimation.

[0090] Determine the expected duration is the value corresponding to the peak of the probability density function, that is, the duration of the maximum likelihood estimation.

[0091] S3.3. For each advertisement in the candidate advertisement set, the corresponding playback duration is .

[0092] Define the advertisement duration matching degree function: ; is the time scale parameter of the duration sequence , calculated according to the mutual information of the data. The function is defined as: ; Ensure that the advertisement duration does not exceed the expected duration.

[0093] Sort the advertisements in descending order according to the advertisement duration matching degree, select the advertisements with a matching degree higher than the corresponding threshold, and form the final candidate advertisement set.

[0094] S3.4. Use reinforcement learning to drive advertisement selection optimization; select the optimal advertisement from the final candidate advertisement set for placement to improve the advertisement effect and passenger experience.

[0095] State space: includes driving situation labels, expected duration, and attribute characteristics of candidate advertisements (such as playback duration).

[0096] Action space: Select an advertisement from the final candidate advertisement set for placement.

[0097] State transition: Since the driving situation of the bus will change, the system will receive new driving situation labels at each time step, resulting in a change in the state.

[0098] Advertising Play Completion Rate : The ratio of the actual advertising play duration to the total advertising duration; when the advertisement is played in full, the advertising play completion rate is 1; if the advertisement is interrupted due to a change in the situation, the play completion rate is less than 1.

[0099] Advertising Switching Stability : Set a time window (such as 30 minutes) and count the number of advertising switches within this time window , the advertising switching stability score is: ; where is a tuning parameter representing the expected average number of switches.

[0100] Reward Function: ; where, and are weight coefficients that can be automatically adjusted according to the system goal and can be set to be equal initially.

[0101] The policy optimization algorithm uses deep reinforcement learning algorithms such as Deep Q-Network (DQN) or Proximal Policy Optimization (PPO).

[0102] Through the reinforcement learning model, at each state, an action with the maximum Q-value is selected according to the policy network or Q-value table, that is, the advertisement that best matches the state in the current candidate advertisement set is selected and marked as the optimal advertisement. The selection of this advertisement is not only based on the current driving situation label and advertisement duration matching degree, but also comprehensively considers historical feedback data such as advertising play completion rate and system stability to ensure that the advertising placement effect is globally optimal and consistent with the system goal.

[0103] Step S3 takes the driving situation label as the core and constructs a systematic decision-making process from advertisement screening to final placement. Through the high-dimensional semantic embedding model, the accurate matching of the situation label and the multi-dimensional attributes of the advertisement is realized to ensure that the initially screened advertisement content has high relevance; on this basis, combined with the duration analysis of historical driving data, the information entropy is used to evaluate the robustness and the kernel density estimation is used to determine the expected duration to ensure the matching of the advertisement play duration and the driving situation. The advertisement screening further relies on the duration matching degree function and strictly screens the suitable advertisement set in combination with the dynamic time scale parameter. Within this framework, the selection strategy is optimized through reinforcement learning, and both the state space and the reward function are based on objectively quantifiable system indicators, comprehensively considering the advertising play completion rate and switching stability, and an intelligent advertising placement decision-making mechanism adapted to the dynamic changes of the driving situation is established.

[0104] After completing the selection of the optimal advertisement based on driving scenario tags, it is necessary to efficiently and stably deliver the selected optimal advertisement content to the bus side window display. Since the bus travels on a fixed route, the driving states and their durations in different sections have certain regularity and dynamic variability, making the timely loading and switching of advertisement content particularly important. Step S4 aims to ensure the efficient display of continuity and smoothness of advertisements during vehicle operation through an advanced pre-caching mechanism and a seamless switching algorithm.

[0105] Step S4 includes the following:

[0106] Based on the obtained optimal advertisement, the selected advertisement content is then efficiently loaded into the side window display through a priority caching mechanism. First, the optimal advertisement is stored in the main cache area of the high-speed cache according to a predetermined priority, and at the same time, the relevant preparatory advertisements are stored in the auxiliary cache area according to the scenario matching degree. Using a memory management algorithm, the main cache area always stores the high-priority segments of the currently playing advertisement, and the auxiliary cache area is dynamically adjusted to store other high-matching advertisement segments. Adopting a dynamic buffer scheduling strategy based on timestamps, the driving scenario changes are monitored in real time, and through pre-loading and asynchronous loading technologies, seamless switching of advertisement content is achieved.

[0107] The specific processing includes segmented pre-buffering, splitting the advertisement content into multiple playing segments and loading them in parallel to ensure a rapid response during advertisement switching. Edge computing nodes are used to optimize the data transmission rate and reduce latency. Combining with adaptive bitrate streaming technology, the quality of the advertisement data stream is dynamically adjusted according to the real-time network conditions. In addition, the system integrates an intelligent frame alignment algorithm to ensure the smoothness of visual transition by precisely synchronizing the end frame of the current advertisement with the start frame of the new advertisement.

[0108] Adopting multi-channel data transmission and efficient compression coding technology significantly reduces the occupied space and transmission time of advertisement data in the cache. The combination of dynamic buffer scheduling and multi-channel transmission enables the advertisement content to be stably and smoothly displayed in various driving scenarios.

[0109] Finally, the optimal advertisement is efficiently and stably delivered to the side window display through an optimized caching and loading process, ensuring that the advertisement always maintains a high-quality display effect in a complex driving environment.

[0110] Step S4 preferentially stores the currently playing advertisement and highly matched alternative advertisements through a double-layer cache architecture, splits the advertisement into multiple playing segments using segmented pre-buffering technology and loads them in parallel to ensure the continuity of the playing process. The intelligent frame alignment algorithm realizes the smooth transition of the advertisement content and avoids visual interruption. Edge computing optimizes the data transmission rate and latency, and the adaptive bitrate streaming technology dynamically adjusts the advertisement quality to ensure stable playing in different network environments. Multi-channel data transmission and efficient compression coding further improve the loading efficiency and ensure that the advertisement is presented on the side window display with high resolution and smoothness. The overall design realizes the efficiency and stability of dynamic advertisement loading and switching, and ensures the continuity and visual effect of advertisement display in the dynamic driving scenario.

[0111] Embodiment 2: Figure 2 A side window advertisement display management system for a vehicle of the present invention is provided, including: a data fusion module, a situation recognition module, an advertisement selection module, and a dynamic loading module.

[0112] Data fusion module: Based on multi-source data fusion technology, extract the comprehensive data stream of the vehicle, and ensure the accuracy of the driving state data through Kalman filter denoising processing, and then transfer the processed data to the situation recognition module.

[0113] Situation recognition module: Use a deep learning model to extract features from the fused driving state data, establish a driving situation recognition model in combination with acceleration and steering features, generate real-time situation labels of the vehicle, and transfer the labels to the advertisement selection module.

[0114] Advertisement selection module: Screen out candidate advertisements with multi-dimensional semantic matching from the labeled advertisement content library based on the real-time situation label, and optimize the advertisement selection strategy through a reinforcement learning model in combination with the duration robustness of the driving state, and finally obtain the optimal advertisement and transfer it to the dynamic loading module.

[0115] Dynamic loading module: Adopt a pre-caching mechanism and a seamless switching algorithm to dynamically load and accurately display the optimal advertisement content on the side window display of the bus in real time to ensure the continuity and smoothness of the advertisement playing.

[0116] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula that is closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0117] Only some exemplary embodiments of the present invention have been described by way of illustration. Undoubtedly, for those of ordinary skill in the art, the described embodiments can be modified in various different ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0118] It should be noted that in this text, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0119] As described above, this is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for managing the display of advertisements on the side windows of vehicles, characterized in that: Includes steps: S1, based on multi-source data fusion technology, extracts the comprehensive data stream of the vehicle and uses Kalman filter denoising to ensure the accuracy of driving status data; S2, extracts features from driving status data through a deep learning model, builds a driving situation recognition model based on acceleration and steering features, and generates real-time situation labels for vehicles; S3, based on driving context labels, selects candidate ads with multi-dimensional semantic matching from the labeled advertising content library, combines the duration robustness of driving status, optimizes the advertising selection strategy through a reinforcement learning model, realizes context-driven advertising selection, and obtains the optimal advertisement; S4, using a pre-caching mechanism and seamless switching algorithm, dynamically loads and displays the best advertising content on the bus side window display in real time; Step S1 includes the following contents: S1.1, firstly integrate the data from different sensors to form a comprehensive data stream describing the driving status of the vehicle; for each data source, first perform standardization processing to eliminate the influence between different dimensions and magnitudes; S1.2 uses a multi-dimensional adaptive fusion matrix to dynamically adjust the weight of each data source according to the changes in the real-time environment and vehicle status, including the following steps: First, the spatial positioning data, attitude angle data, and acceleration data are standardized; Then, set the initial weight for each data source; Next, the correlation between the data sources is calculated through the state correlation function. The evaluation of the state correlation function is based on the mutual information in information theory. Finally, the correlation between the data sources is quantified by calculating the mutual information; the weight matrix is ​​updated according to the correlation of the data sources; S1.3, using the updated weight matrix to perform weighted combination on the standardized data to obtain a comprehensive driving state vector reflecting the current driving state of the vehicle; Step S1 also includes the following contents: S1.4, using nonlinear Kalman filtering to optimize and filter the fused driving state vector, including the following two stages: Prediction stage: Based on the state at the previous moment, use the state transfer matrix Make predictions: ;in, is the predicted vehicle driving state, is the state transfer matrix, describing the state dynamics of the vehicle; Update phase: Combine the current observation data , optimize the prediction results by updating the formula: ; ;in, is the Kalman gain, is the forecast error covariance, is the observation matrix, is the observation noise covariance, For observation data.

2. A method for managing vehicle side window advertising display according to claim 1, characterized in that: Step S2 includes the following contents: S2.1, based on the driving state vector, a convolutional neural network is used to extract short-term feature fluctuations and capture local features in the vehicle driving state; S2.2, each feature in the driving state vector has temporal correlation, and the long short-term memory network is used to mine the temporal features of the driving state; S2.3, after obtaining the fused feature vector, the corresponding features are mapped to the high-dimensional situational space, and the feature relationship is captured through nonlinear mapping, and finally a real-time driving situation label is generated.

3. A method for managing vehicle side window advertising display according to claim 2, characterized in that: Step S3 includes the following contents: S3.1, in the labeled advertising content library, retrieve all advertisements related to the current driving situation label to form a preliminary advertising set; use the deep semantic embedding model to map the driving situation label and advertising label vector to a high-dimensional semantic space to obtain the semantic embedding vector and ; Similarity calculation function: ;in, represents the vector inner product, represents the norm of a vector; Indicates The semantic similarity between each advertisement and the current context label is calculated; the candidate advertisements are sorted in descending order according to the similarity value, and the advertisements with similarity higher than the similarity threshold are selected to form a candidate advertisement set.

4. A method for managing vehicle side window advertising display according to claim 3, characterized in that: Step S3 also includes the following: S3.2, buses travel on fixed routes, and the duration of driving status in different sections is regular; the duration of advertisements needs to match the estimated duration of the vehicle in the current driving situation to ensure that the advertisements can be played completely without affecting the passenger experience; collect historical driving data of buses on the same route, the same section, and the same driving situation label to obtain the duration sequence ; Use information entropy to evaluate the robustness of duration; If the information entropy is less than or equal to the corresponding threshold, the duration is considered to be sufficiently robust; Use kernel density estimation to obtain the probability density function of duration; Determine the expected duration The peak value of the probability density function corresponds to Value, i.e. duration of maximum likelihood estimation.

5. A method for managing vehicle side window advertising display according to claim 4, characterized in that: Step S3 also includes the following: S3.3, for each advertisement in the candidate advertisement set, the corresponding playing time is ; Define the ad duration matching function: ; For duration series The time scale parameter is calculated based on the mutual information of the data; the function Defined as: ; Make sure the ad does not exceed its estimated duration; The ads are sorted in descending order according to the ad duration matching degree, and the ads with matching degrees higher than the corresponding threshold are selected to form the final candidate ad set.

6. A method for managing vehicle side window advertising display according to claim 5, characterized in that: Step S3 also Includes the following: S3.4, use reinforcement learning to drive advertising selection optimization; select the best advertisement from the final candidate advertisement set for delivery; through the reinforcement learning model, in each state, select the action with the maximum Q value according to the Q value table, that is, select the advertisement in the current candidate advertisement set that best matches the state and mark it as the best advertisement.

7. A vehicle side window advertising display management system, used to implement a vehicle side window advertising display management method as claimed in any one of claims 1 to 6, characterized in that: include: Data fusion module, context recognition module, advertisement selection module and dynamic loading module; Data fusion module: Based on multi-source data fusion technology, it extracts the comprehensive data stream of the vehicle and uses Kalman filter denoising to ensure the accuracy of the driving status data. The processed data is then passed to the situation recognition module; Context recognition module: Use the deep learning model to extract features from the fused driving status data, combine acceleration and steering features to establish a driving context recognition model, generate real-time context labels for the vehicle, and pass the labels to the advertisement selection module; Advertisement selection module: Based on real-time context labels, candidate advertisements with multi-dimensional semantic matching are screened out from the labeled advertisement content library. Combined with the duration robustness of the driving state, the advertisement selection strategy is optimized through the reinforcement learning model, and the optimal advertisement is finally obtained and passed to the dynamic loading module. Dynamic loading module: adopts pre-caching mechanism and seamless switching algorithm to dynamically load and accurately display the optimal advertising content on the bus side window display in real time, ensuring the continuity and smoothness of advertising playback.

Images