Aromatherapy intelligent margin calculation method based on dynamic perception of vehicle body data

Abstract

The present invention discloses an intelligent fragrance remaining calculation method based on dynamic vehicle body data perception, comprising the following steps: S1. Generating an initial dataset; S2. Inputting the initial dataset into a multi-channel variational encoder to output a hierarchical feature representation; S3. Inputting the hierarchical feature representation into a spatiotemporal context perception module to generate a spatiotemporal feature representation; S4. Inputting the spatiotemporal feature representation into a heterogeneous feature mapping layer to generate a unified data representation; S5. Inputting the unified data representation into a dynamic prediction and control module to generate an optimized feature representation; S6. Inputting the optimized feature representation into a fragrance remaining prediction model to perform remaining prediction using a mixture of Gaussian processes; S7. Inputting the fragrance remaining prediction results into a feedback integrator for self-supervised learning; S8. Integrating the method into an on-board system to display the fragrance remaining status and issue personalized notifications. The present invention utilizes multi-channel variational coding, spatiotemporal context perception, and Gaussian processes to achieve accurate prediction and calculation of vehicle fragrance remaining.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent vehicle-mounted systems, and in particular to a method for calculating an intelligent aromatherapy remainder based on dynamic perception of vehicle body data. Background Art

[0002] Among existing in-vehicle environmental control systems, in-vehicle air quality management and comfort adjustment have become crucial components in the development of intelligent vehicles. Aromatherapy systems are widely used in vehicles to enhance driver and passenger comfort. However, traditional aromatherapy systems have significant limitations in management and control. They primarily rely on preset usage patterns and simple time or quantitative control strategies, and are unable to dynamically adjust and accurately predict based on real-time environmental data and user behavior. This traditional approach struggles to cope with complex environmental changes both inside and outside the vehicle, as well as varying user preferences. Consequently, aromatherapy systems suffer from issues such as insufficient adjustment precision, inflexible response, and inaccurate management of remaining fragrance levels.

[0003] In current research and technological development, some systems have introduced sensors to detect in-vehicle environmental data, such as parameters such as temperature, humidity, and air quality. However, these systems often lack data processing and multi-source information fusion capabilities, resulting in limited correlation analysis and feature extraction capabilities between data, and the inability to achieve in-depth real-time intelligent analysis. This makes it difficult for existing systems to accurately model user usage behavior and external environmental changes in complex environments when predicting and optimizing fragrance remaining. In addition, traditional data processing models typically rely on fixed algorithms and simple models when dealing with dynamically changing complex features. They are unable to adjust parameters and structures based on the real-time environment or user feedback, making it difficult for the system to maintain long-term predictive performance and adaptability.

[0004] When regulating aromatherapy systems, existing methods often lack the ability to comprehensively analyze multidimensional features, especially when simultaneously processing different features such as high-frequency, low-frequency, and sudden events. Existing technologies often use simple single-channel analysis methods, which cannot effectively cope with complex and changing environmental inputs. This is especially true when the external environment changes rapidly or when user behavior in the vehicle is abnormal. Prediction accuracy and system stability will decrease significantly. In addition, due to the inability to dynamically adjust the loss function and model structure to optimize system performance, traditional methods have weak adaptability and robustness, making it impossible to achieve efficient margin management for aromatherapy systems in different scenarios.

[0005] While existing technologies have attempted to improve the intelligence of systems by introducing machine learning and simple neural networks, the complexity of these models and the computational power limit their performance when processing real-time data and multimodal inputs. For example, some systems may attempt to use variational autoencoders to handle nonlinear features, but these systems often fail to incorporate multi-channel structures and adaptive loss adjustment strategies, making them unable to address the uncertainty and feature coupling issues in high-dimensional data. Furthermore, existing time series data analysis methods, such as traditional sliding windows and unidirectional attention mechanisms, struggle to fully capture the spatiotemporal characteristics of the data and cannot fully reflect the relationship between historical and real-time context, thereby limiting the model's predictive capabilities and contextual adaptability.

[0006] In terms of feedback and self-learning, most current aromatherapy systems lack effective self-supervised learning and dynamic reconstruction mechanisms. Even though a few systems attempt to use feedback loops for simple parameter updates, they lack the ability to effectively detect and respond to long-term deviations and new features, making it difficult to improve system performance through adaptive adjustments. Existing learning mechanisms are mostly static and fail to achieve automated model reconstruction and optimization, limiting the scalability and durability of the models over long-term use.

[0007] Therefore, how to provide an intelligent fragrance remainder calculation method based on dynamic perception of vehicle body data is an urgent problem that needs to be solved by those skilled in the art. Summary of the Invention

[0008] One purpose of the present invention is to propose an intelligent aromatherapy remainder calculation method based on dynamic perception of vehicle body data. The present invention combines a multi-channel variational encoder, a spatiotemporal context perception module, and a mixed Gaussian process technology to achieve efficient collection and analysis of in-vehicle and out-of-vehicle environmental data and user behavior data. Feature fusion is performed through a bidirectional spatiotemporal attention mechanism and a heterogeneous feature mapping layer to ensure that the system has excellent feature recognition and dynamic response capabilities in different environments. With the help of real-time feedback and self-supervised learning mechanisms, the present invention can continuously optimize the model structure and parameters, improve the prediction accuracy and adaptive performance of the system in complex and changing environments, and has the significant advantages of high precision, robustness, and fast response.

[0009] According to an embodiment of the present invention, a method for calculating the intelligent remaining amount of fragrance based on dynamic perception of vehicle body data includes the following steps:

[0010] S1. Collect environmental data and user behavior data through sensors inside and outside the vehicle, and perform noise filtering, outlier removal, and standardization on the data to generate an initial data set.

[0011] S2. Input the initial data set into a multi-channel variational encoder, classify and encode it according to high frequency, low frequency and sudden events, output a hierarchical feature representation, and realize feature information exchange through the interactive network between channels, and dynamically adjust the variational loss function of the encoder;

[0012] S3. Input the hierarchical feature representation into the spatiotemporal context perception module, capture the time series data features through the spatiotemporal sliding window, combine the bidirectional spatiotemporal attention mechanism to generate spatiotemporal feature representation, and inject historical and real-time context weights;

[0013] S4. Input the spatiotemporal feature representation into the heterogeneous feature mapping layer to perform spatiotemporal feature mapping and conversion, and map to generate a unified data representation;

[0014] S5. Input the unified data representation into the dynamic prediction and control module, dynamically adjust the input feature weights based on real-time feedback, and generate an optimized feature representation;

[0015] S6. Inputting the optimized feature representation into the aromatherapy remainder prediction model, performing remainder prediction using a mixed Gaussian process, and updating the parameters of the aromatherapy remainder prediction model in combination with variational inference;

[0016] S7: Input the fragrance remaining amount prediction result into the feedback integrator, collect user feedback and in-vehicle and out-of-vehicle data for self-supervised learning. When long-term deviations or new data features are detected, the dynamic intelligent reconstruction and optimization learning mechanism is triggered to update the structure and optimize the parameters of the fragrance remaining amount prediction model;

[0017] S8. Integrate the updated fragrance remaining prediction model into the vehicle system to display the fragrance remaining status in real time, and issue personalized notifications and usage suggestions based on user behavior and fragrance remaining prediction results.

[0018] Optionally, the vehicle body status data specifically includes the vehicle start / stop status, whether the windows are open, and whether the air conditioner in the vehicle is turned on.

[0019] Optionally, the S2 specifically includes:

[0020] S21, input the environmental data and user behavior data in the initial dataset into the multi-channel variational encoder;

[0021] S22. In a multi-channel variational encoder, hierarchically classify the input data into three categories: high frequency, low frequency, and sudden events. The high frequency channel extracts rapidly changing environmental features in the input data, the low frequency channel extracts long-term stable features in the input data, and the sudden event channel extracts unpredictable abnormal changes in the input data.

[0022] S23, feature encoding is performed on the data in each channel, and the high-frequency feature encoder generates a high-frequency feature vector V high, the low-frequency feature encoder generates a low-frequency feature vector V low , the emergency feature encoder generates the emergency feature vector V event ;

[0023] S24, feature information is exchanged through the interactive network between channels, and the high-frequency feature vector V high , low-frequency eigenvector V low and the emergency feature vector V event Perform interactive processing and output comprehensive feature matrix M exchange :

[0024]

[0025] Among them, k represents the dimension of the feature vector, and α represents the normalization coefficient;

[0026] S25. During the feature information exchange process, the variational loss function of the encoder is dynamically adjusted using co-correlation analysis to optimize the encoding quality of each feature channel:

[0027]

[0028] Among them, L var represents the loss function, n represents the total number of features in the input dataset, KL represents the divergence value, γ i represents the weight coefficient, M represents the number of additional features, q(z|x i ) represents the encoded posterior distribution, p(z represents the prior distribution, represents the standard deviation of the input feature, ∈ represents the smoothing term, θ represents the adjustment factor, μ represents the offset, β and δ represent the adjustment parameters, T represents the time step, and λ represents the scaling parameter;

[0029] S26. Output the adjusted hierarchical feature representation.

[0030] Optionally, the S3 specifically includes:

[0031] S31. Inputting the hierarchical feature representation into a spatiotemporal context perception module, wherein the spatiotemporal context perception module performs a temporal analysis on the input multi-channel features to capture the dynamic changes of environmental data and user behavior in the temporal and spatial dimensions;

[0032] S32. Apply a spatiotemporal sliding window strategy to divide the input feature data into different time windows for processing. The spatiotemporal sliding window dynamically adjusts its size and step size according to vehicle speed, environmental conditions, and data changes to capture both short-term and long-term time series features.

[0033] S33. Within each spatiotemporal sliding window, a bidirectional spatiotemporal attention mechanism is used to perform feature analysis. The bidirectional spatiotemporal attention mechanism captures the association between the current time step and past features through a forward attention network, and captures the association between the current time step and future features through a backward attention network, generating an attention weight that covers global temporal information.

[0034] S34, based on the attention weights covering global temporal information generated by the bidirectional spatiotemporal attention mechanism, weights the features within the window, highlights the attention of important features related to the current task, suppresses noise and irrelevant features, and forms a feature representation with temporal correlation and context association;

[0035] S35. Injecting historical context information and real-time context information into the spatiotemporal feature representation, and dynamically combining the weights of historical data and real-time data through context weight injection technology;

[0036] S36. Output the spatiotemporal feature representation that combines historical and real-time context information as input for unified data representation.

[0037] Optionally, the S4 specifically includes:

[0038] S41, inputting the spatiotemporal feature representation output by the spatiotemporal context perception module into a heterogeneous feature mapping layer, wherein the heterogeneous feature mapping layer is composed of multiple parallel processing units, each of which performs preliminary analysis and segmentation on features of different sources and time scales;

[0039] S42. Through the double normalization strategy and adaptive filtering algorithm, the weight distribution of high-frequency features and low-frequency features is gradually balanced, and dynamic range adjustment is adopted to maintain a stable distribution of features from each source, reducing the conversion bias caused by inconsistent data scales;

[0040] S43. In the heterogeneous feature mapping conversion network, a multi-layer convolution combined with a deformable convolution is used to map and reconstruct features layer by layer, gradually transforming the original spatiotemporal features into a comprehensive representation with a unified feature space. The deformable convolution is used to automatically adjust the shape of the convolution kernel to adapt to changes in features in different time periods.

[0041] S44. During the feature mapping process, contextual features are embedded in combination with the output of the context-aware module. The historical features are organically integrated with the current real-time features through the contextual attention mechanism to form a real-time context weight mapping. Through the dynamic adjustment of weight priorities, fine-grained weight assignment is performed on the input features.

[0042] S45. Adopt a multi-channel feature aggregation strategy to perform hierarchical aggregation processing on the mapped high-frequency, low-frequency and emergency event features, introduce an interaction weight matrix between channels, and achieve interactive enhancement of features and complementarity of multi-dimensional information;

[0043] S46. Perform stability check on the final unified data representation before output, and perform output correction in combination with the feature smoothing mechanism.

[0044] Optionally, the S5 specifically includes:

[0045] S51. Input the generated unified data representation into a dynamic prediction and control module, wherein the dynamic prediction and control module includes a feature weighting unit and an adaptive feedback unit, which is used to adjust the weights of input features in real time to adapt to environmental changes;

[0046] S52. In the feature weighting unit, a multi-layer adaptive weight assignment algorithm is used to perform weighted processing on the input unified data representation. The multi-layer adaptive weight assignment algorithm automatically assigns weights based on the rate of change and correlation of features at different time points to generate a preliminary optimized feature matrix. The weights are updated through an iterative process to adapt to dynamic changes in the environment and user behavior.

[0047] S53: Receive, through the adaptive feedback unit, changes in external environment data and user behavior data in real time, compare the feedback results with the current weight distribution, and adjust the weight distribution strategy to gradually learn and adapt to changes in input features under different environmental conditions based on an incremental learning mechanism;

[0048] S54, screening the preliminarily optimized feature matrix, using a feature selection algorithm to eliminate redundant and low-correlation features, retaining high-correlation features, and ensuring high information density, wherein the feature selection algorithm uses a weight pruning technology;

[0049] S55. The filtered feature matrix is ​​passed to the dynamic control strategy module, and the feature matrix is ​​finally optimized in combination with the context association model. The context association model adjusts the time and environment priorities of the features based on historical data and real-time feedback to generate a comprehensive optimized feature representation;

[0050] S56. Output the optimized feature representation as the input of the aromatherapy remainder prediction module.

[0051] Optionally, the S6 specifically includes:

[0052] S61, inputting the optimized feature representation into an aromatherapy residual quantity prediction model, wherein the aromatherapy residual quantity prediction model adopts a mixed Gaussian process structure and combines multi-dimensional feature input to achieve modeling of complex nonlinear relationships between features;

[0053] S62, the aromatherapy remaining quantity prediction model consists of a two-layer feature processing unit. The first layer is the input feature mapping layer, which embeds the optimized feature representation into a high-dimensional feature space. The second layer adopts a multi-kernel Gaussian process structure, combining linear kernel, radial basis kernel and polynomial kernel functions to capture the complex interactive relationship between high-frequency, low-frequency and emergency event features.

[0054] S63. In the Gaussian process core layer, define the margin prediction formula and implement adaptive adjustment of margin prediction by dynamically weighing different feature combinations:

[0055]

[0056] Among them, y 余量 (t) represents the prediction result of aromatherapy residue, μ 高频 (t) represents the predicted mean of high-frequency features, μ 低频 (t) represents the predicted mean of low-frequency features, μ 突发 (t) represents the predicted mean of the emergency event characteristics, σ 低频 (t represents the prediction standard deviation of low-frequency features, σ 高频 (t) represents the prediction standard deviation of high-frequency features, σ 突发 (t) represents the predicted standard deviation of the emergency event characteristics, α 1,j , α 2,l and λ r represents the weight adjustment coefficient, β 1,i and β 2,p Indicates the integral adjustment coefficient, γ 1,i and γ 2,p represents the scaling parameter, δ3 represents the global adjustment coefficient, f j (t) represents the input feature gradient function at time t, h l (t) represents the input feature correction function at time t, g r (t) represents the correction function of the sudden event, k represents the number of weight coefficients in the high-frequency feature, m represents the number of adjustment parameters of the high-frequency feature, n represents the number of adjustment coefficients of the low-frequency feature, q represents the number of adjustment parameters for time series integration in the low-frequency feature, s represents the number of correction functions in the sudden event, u represents the number of time steps used for cumulative standard deviation calculation in the sudden event part, t v represents the low-frequency features at time step v, θ1 and θ2 represent threshold parameters;

[0057] S64, using variational inference technology to update model parameters in real time, optimize the posterior distribution of the Gaussian process, and ensure that the model can adapt to changes in input;

[0058] S65. During the residual forecast process, combine the uncertainty quantitative analysis of the model to estimate the confidence interval of the forecast results to identify and address potential forecast biases;

[0059] S66: Output aromatherapy remaining amount prediction result y 余量 (t), combining multidimensional features and uncertainty information to generate aromatherapy remainder prediction results with high confidence.

[0060] Optionally, the S7 specifically includes:

[0061] S71. Inputting the prediction result of the fragrance remaining amount prediction model into a feedback integrator, wherein the feedback integrator continuously collects data from the vehicle's internal and external environments, including temperature, humidity, vehicle speed, and user operation behavior information, and combines it with user feedback to form a multi-dimensional data set;

[0062] S72. Based on the multi-dimensional data set, compare the actual environmental data with the predicted results, generate a deviation analysis report, identify the deviation between the predicted value and the actual value, and determine whether the deviation exceeds the preset long-term threshold;

[0063] S73. When the deviation analysis results show that the set long-term threshold is exceeded or a new feature appears, the self-supervised learning module is activated to generate pseudo labels from unlabeled data, expand the training data set, automatically learn new environmental features, and improve the adaptive ability of the fragrance remaining prediction model to cope with the environment;

[0064] S74, inputting the expanded training data set into the incremental learning unit of the fragrance remaining prediction model, wherein the incremental learning unit performs a progressive parameter update on the fragrance remaining prediction model to maintain efficient operation of the fragrance remaining prediction model in a real-time environment and optimize prediction performance;

[0065] S75. When frequent long-term deviations or new data features are detected, a dynamic intelligent reconstruction mechanism is triggered to automatically adjust the structure of the aromatherapy remainder prediction model, including optimizing the model hierarchy, reconfiguring the weight distribution, and adjusting the feature path;

[0066] S76. After completing the structural adjustment and parameter optimization of the fragrance remaining amount prediction model, the improved fragrance remaining amount prediction model is applied to the vehicle-mounted system, and the updated version information and adjustment details of the fragrance remaining amount prediction model are recorded.

[0067] The beneficial effects of the present invention are:

[0068] First, by introducing a multi-channel variational encoder, the present invention is able to efficiently and hierarchically encode and analyze in- and out-of-vehicle environmental data and user behavior data, fully capturing multidimensional features such as high-frequency, low-frequency, and sudden events. The application of this multi-channel structure improves the ability to process complex data, enabling more accurate identification and response to different feature changes in a real-time environment, effectively compensating for the limitations of traditional single-channel analysis methods when processing multi-source data. Combined with the interactive network between channels, the present invention enables dynamic information exchange and collaborative analysis between features, improving the quality and robustness of the overall feature encoding.

[0069] Secondly, by introducing a spatiotemporal context-aware module and a bidirectional spatiotemporal attention mechanism, the system captures the bidirectional temporal characteristics of historical and real-time data, achieving comprehensive awareness of the dynamic changes in the vehicle's interior and exterior environments. The infusion of historical and real-time contextual weights ensures that the system accurately reflects user behavior and environmental changes across different time spans, thereby improving prediction accuracy and the system's adaptability to the environment. Compared to existing unidirectional time series analysis methods, this bidirectional spatiotemporal analysis method significantly enhances the system's spatiotemporal awareness and addresses the issues of prediction lag and missing contextual information.

[0070] Furthermore, through the heterogeneous feature mapping layer, the present invention achieves new progress in multimodal data processing. This layer enables seamless mapping and conversion of features from different sources and timescales, enabling comprehensive analysis of multi-source data within a unified feature space. This not only improves the efficiency of feature fusion but also ensures the stability and consistency of the system's processing of complex inputs in dynamic environments. Furthermore, the dynamic prediction and control module adjusts input feature weights through real-time feedback, enabling the system to rapidly adapt to changes in the environment or user behavior, ensuring the flexibility and continuous optimization capabilities of the prediction model.

[0071] The aromatherapy remaining quantity prediction model combines mixed Gaussian processes and variational inference techniques, providing the ability to model complex nonlinear feature relationships and dynamically update parameters. By incorporating real-time uncertainty quantification and variational inference, the system can adaptively adjust when long-term deviations or new features are detected, improving prediction accuracy and the model's robustness in changing environments. This capability enables the system to not only effectively predict current data but also maintain its long-term prediction performance through self-learning.

[0072] Finally, through a feedback integrator and a dynamic intelligent reconstruction mechanism, this invention achieves continuous self-supervised learning and model optimization. The system automatically updates and reconstructs the model when deviations exceed thresholds or new data features are detected, ensuring long-term stability and scalability. Compared to traditional approaches, this invention offers a higher level of adaptability and intelligence, improving user experience and system performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0074] Figure 1 This is a flow chart of the intelligent aromatherapy remainder calculation method based on dynamic perception of vehicle body data proposed by the present invention;

[0075] Figure 2 This is a structural diagram of the intelligent fragrance remainder calculation method based on dynamic perception of vehicle body data proposed in the present invention. DETAILED DESCRIPTION

[0076] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0077] refer to Figure 1 and Figure 2 The intelligent aromatherapy reserve calculation method based on dynamic perception of vehicle body data includes the following steps:

[0078] S1. Collect environmental data and user behavior data through sensors inside and outside the vehicle, and perform noise filtering, outlier removal, and standardization on the data to generate an initial data set.

[0079] S2. Input the initial data set into a multi-channel variational encoder, classify and encode it according to high frequency, low frequency and sudden events, output a hierarchical feature representation, and realize feature information exchange through the interactive network between channels, and dynamically adjust the variational loss function of the encoder;

[0080] S3. Input the hierarchical feature representation into the spatiotemporal context perception module, capture the time series data features through the spatiotemporal sliding window, combine the bidirectional spatiotemporal attention mechanism to generate spatiotemporal feature representation, and inject historical and real-time context weights;

[0081] S4. Input the spatiotemporal feature representation into the heterogeneous feature mapping layer to perform spatiotemporal feature mapping and conversion, and map to generate a unified data representation;

[0082] S5. Input the unified data representation into the dynamic prediction and control module, dynamically adjust the input feature weights based on real-time feedback, and generate an optimized feature representation;

[0083] S6. Inputting the optimized feature representation into the aromatherapy remainder prediction model, performing remainder prediction using a mixed Gaussian process, and updating the parameters of the aromatherapy remainder prediction model in combination with variational inference;

[0084] S7: Input the fragrance remaining amount prediction result into the feedback integrator, collect user feedback and in-vehicle and out-of-vehicle data for self-supervised learning. When long-term deviations or new data features are detected, the dynamic intelligent reconstruction and optimization learning mechanism is triggered to update the structure and optimize the parameters of the fragrance remaining amount prediction model;

[0085] S8. Integrate the updated fragrance remaining prediction model into the vehicle system to display the fragrance remaining status in real time, and issue personalized notifications and usage suggestions based on user behavior and fragrance remaining prediction results.

[0086] In this embodiment, the vehicle body status data specifically includes the vehicle start / stop status, whether the windows are open, and whether the air conditioner in the vehicle is turned on.

[0087] In this embodiment, S2 specifically includes:

[0088] S21, input the environmental data and user behavior data in the initial dataset into the multi-channel variational encoder;

[0089] S22. In a multi-channel variational encoder, hierarchically classify the input data into three categories: high frequency, low frequency, and sudden events. The high frequency channel extracts rapidly changing environmental features in the input data, the low frequency channel extracts long-term stable features in the input data, and the sudden event channel extracts unpredictable abnormal changes in the input data.

[0090] S23, feature encoding is performed on the data in each channel, and the high-frequency feature encoder generates a high-frequency feature vector V high , the low-frequency feature encoder generates a low-frequency feature vector V low , the emergency feature encoder generates the emergency feature vector V event ;

[0091] S24, feature information is exchanged through the interactive network between channels, and the high-frequency feature vector V high , low-frequency eigenvector V low and the emergency feature vector V event Perform interactive processing and output comprehensive feature matrix M exchange :

[0092]

[0093] Among them, k represents the dimension of the feature vector, and α represents the normalization coefficient;

[0094] S25. During the feature information exchange process, the variational loss function of the encoder is dynamically adjusted using co-correlation analysis to optimize the encoding quality of each feature channel:

[0095]

[0096] Among them, L var represents the loss function, n represents the total number of features in the input dataset, KL represents the divergence value, γ i represents the weight coefficient, M represents the number of additional features, q(z|x i ) represents the encoded posterior distribution, p(z represents the prior distribution, represents the standard deviation of the input feature, ∈ represents the smoothing term, θ represents the adjustment factor, μ represents the offset, β and δ represent the adjustment parameters, T represents the time step, and λ represents the scaling parameter;

[0097] S26. Output the adjusted hierarchical feature representation.

[0098] In this embodiment, S3 specifically includes:

[0099] S31. Inputting the hierarchical feature representation into a spatiotemporal context perception module, wherein the spatiotemporal context perception module performs a temporal analysis on the input multi-channel features to capture the dynamic changes of environmental data and user behavior in the temporal and spatial dimensions;

[0100] S32. Apply a spatiotemporal sliding window strategy to divide the input feature data into different time windows for processing. The spatiotemporal sliding window dynamically adjusts its size and step size according to vehicle speed, environmental conditions, and data changes to capture both short-term and long-term time series features.

[0101] S33. Within each spatiotemporal sliding window, a bidirectional spatiotemporal attention mechanism is used to perform feature analysis. The bidirectional spatiotemporal attention mechanism captures the association between the current time step and past features through a forward attention network, and captures the association between the current time step and future features through a backward attention network, generating an attention weight that covers global temporal information.

[0102] S34, based on the attention weights covering global temporal information generated by the bidirectional spatiotemporal attention mechanism, weights the features within the window, highlights the attention of important features related to the current task, suppresses noise and irrelevant features, and forms a feature representation with temporal correlation and context association;

[0103] S35. Injecting historical context information and real-time context information into the spatiotemporal feature representation, and dynamically combining the weights of historical data and real-time data through context weight injection technology;

[0104] S36. Output the spatiotemporal feature representation that combines historical and real-time context information as input for unified data representation.

[0105] In this embodiment, the S4 specifically includes:

[0106] S41, inputting the spatiotemporal feature representation output by the spatiotemporal context perception module into a heterogeneous feature mapping layer, wherein the heterogeneous feature mapping layer is composed of multiple parallel processing units, each of which performs preliminary analysis and segmentation on features of different sources and time scales;

[0107] S42. Through the double normalization strategy and adaptive filtering algorithm, the weight distribution of high-frequency features and low-frequency features is gradually balanced, and dynamic range adjustment is adopted to maintain a stable distribution of features from each source, reducing the conversion bias caused by inconsistent data scales;

[0108] S43. In the heterogeneous feature mapping conversion network, a multi-layer convolution combined with a deformable convolution is used to map and reconstruct features layer by layer, gradually transforming the original spatiotemporal features into a comprehensive representation with a unified feature space. The deformable convolution is used to automatically adjust the shape of the convolution kernel to adapt to changes in features in different time periods.

[0109] S44. During the feature mapping process, contextual features are embedded in combination with the output of the context-aware module. The historical features are organically integrated with the current real-time features through the contextual attention mechanism to form a real-time context weight mapping. Through the dynamic adjustment of weight priorities, fine-grained weight assignment is performed on the input features.

[0110] S45. Adopt a multi-channel feature aggregation strategy to perform hierarchical aggregation processing on the mapped high-frequency, low-frequency and emergency event features, introduce an interaction weight matrix between channels, and achieve interactive enhancement of features and complementarity of multi-dimensional information;

[0111] S46. Perform stability check on the final unified data representation before output, and perform output correction in combination with the feature smoothing mechanism.

[0112] In this embodiment, the S5 specifically includes:

[0113] S51. Input the generated unified data representation into a dynamic prediction and control module, wherein the dynamic prediction and control module includes a feature weighting unit and an adaptive feedback unit, which is used to adjust the weights of input features in real time to adapt to environmental changes;

[0114] S52. In the feature weighting unit, a multi-layer adaptive weight assignment algorithm is used to perform weighted processing on the input unified data representation. The multi-layer adaptive weight assignment algorithm automatically assigns weights based on the rate of change and correlation of features at different time points to generate a preliminary optimized feature matrix. The weights are updated through an iterative process to adapt to dynamic changes in the environment and user behavior.

[0115] S53: Receive, through the adaptive feedback unit, changes in external environment data and user behavior data in real time, compare the feedback results with the current weight distribution, and adjust the weight distribution strategy to gradually learn and adapt to changes in input features under different environmental conditions based on an incremental learning mechanism;

[0116] S54, screening the preliminarily optimized feature matrix, using a feature selection algorithm to eliminate redundant and low-correlation features, retaining high-correlation features, and ensuring high information density, wherein the feature selection algorithm uses a weight pruning technology;

[0117] S55. The filtered feature matrix is ​​passed to the dynamic control strategy module, and the feature matrix is ​​finally optimized in combination with the context association model. The context association model adjusts the time and environment priorities of the features based on historical data and real-time feedback to generate a comprehensive optimized feature representation;

[0118] S56. Output the optimized feature representation as the input of the aromatherapy remainder prediction module.

[0119] In this embodiment, S6 specifically includes:

[0120] S61, inputting the optimized feature representation into an aromatherapy residual quantity prediction model, wherein the aromatherapy residual quantity prediction model adopts a mixed Gaussian process structure and combines multi-dimensional feature input to achieve modeling of complex nonlinear relationships between features;

[0121] S62, the aromatherapy remaining quantity prediction model consists of a two-layer feature processing unit. The first layer is the input feature mapping layer, which embeds the optimized feature representation into a high-dimensional feature space. The second layer adopts a multi-kernel Gaussian process structure, combining linear kernel, radial basis kernel and polynomial kernel functions to capture the complex interactive relationship between high-frequency, low-frequency and emergency event features.

[0122] S63. In the Gaussian process core layer, define the margin prediction formula and implement adaptive adjustment of margin prediction by dynamically weighing different feature combinations:

[0123]

[0124] Among them, y 余量 (t) represents the prediction result of aromatherapy residue, μ 高频 (t) represents the predicted mean of high-frequency features, μ 低频 (t) represents the predicted mean of low-frequency features, μ 突发 (t) represents the predicted mean of the emergency event characteristics, σ 低频 (t represents the prediction standard deviation of low-frequency features, σ 高频 (t) represents the prediction standard deviation of high-frequency features, σ 突发(t) represents the predicted standard deviation of the emergency event characteristics, α 1,j , α 2,l and λ r represents the weight adjustment coefficient, β 1,i and β 2,p Indicates the integral adjustment coefficient, γ 1,i and γ 2,p represents the scaling parameter, δ3 represents the global adjustment coefficient, f j (t) represents the input feature gradient function at time t, h l (t) represents the input feature correction function at time t, g r (t) represents the correction function of the sudden event, k represents the number of weight coefficients in the high-frequency feature, m represents the number of adjustment parameters of the high-frequency feature, n represents the number of adjustment coefficients of the low-frequency feature, q represents the number of adjustment parameters for time series integration in the low-frequency feature, s represents the number of correction functions in the sudden event, u represents the number of time steps used for cumulative standard deviation calculation in the sudden event part, t v represents the low-frequency features at time step v, θ1 and θ2 represent threshold parameters;

[0125] S64, using variational inference technology to update model parameters in real time, optimize the posterior distribution of the Gaussian process, and ensure that the model can adapt to changes in input;

[0126] S65. During the residual forecast process, combine the uncertainty quantitative analysis of the model to estimate the confidence interval of the forecast results to identify and address potential forecast biases;

[0127] S66: Output aromatherapy remaining amount prediction result y 余量 (t), combining multidimensional features and uncertainty information to generate aromatherapy remainder prediction results with high confidence.

[0128] In this embodiment, the S7 specifically includes:

[0129] S71. Inputting the prediction result of the fragrance remaining amount prediction model into a feedback integrator, wherein the feedback integrator continuously collects data from the vehicle's internal and external environments, including temperature, humidity, vehicle speed, and user operation behavior information, and combines it with user feedback to form a multi-dimensional data set;

[0130] S72. Based on the multi-dimensional data set, compare the actual environmental data with the predicted results, generate a deviation analysis report, identify the deviation between the predicted value and the actual value, and determine whether the deviation exceeds the preset long-term threshold;

[0131] S73. When the deviation analysis results show that the set long-term threshold is exceeded or a new feature appears, the self-supervised learning module is activated to generate pseudo labels from unlabeled data, expand the training data set, automatically learn new environmental features, and improve the adaptive ability of the fragrance remaining prediction model to cope with the environment;

[0132] S74, inputting the expanded training data set into the incremental learning unit of the fragrance remaining prediction model, wherein the incremental learning unit performs a progressive parameter update on the fragrance remaining prediction model to maintain efficient operation of the fragrance remaining prediction model in a real-time environment and optimize prediction performance;

[0133] S75. When frequent long-term deviations or new data features are detected, a dynamic intelligent reconstruction mechanism is triggered to automatically adjust the structure of the aromatherapy remainder prediction model, including optimizing the model hierarchy, reconfiguring the weight distribution, and adjusting the feature path;

[0134] S76. After completing the structural adjustment and parameter optimization of the fragrance remaining amount prediction model, the improved fragrance remaining amount prediction model is applied to the vehicle-mounted system, and the updated version information and adjustment details of the fragrance remaining amount prediction model are recorded.

[0135] Example 1:

[0136] In order to verify the feasibility of the present invention in implementation, the present invention was applied to a high-end autonomous driving vehicle. During long-term use, the owner found that the traditional aromatherapy system showed inaccurate management and delayed response when responding to changes in the environment inside and outside the vehicle. For example, when the vehicle windows are open or the air conditioner is used for a long time, the airflow and temperature in the car change significantly. It is difficult for the traditional system to accurately adjust the aromatherapy remainder display according to these changes, resulting in aromatherapy waste or poor effect. In order to address this problem, the owner used the present invention's intelligent aromatherapy remainder calculation method based on dynamic perception of vehicle body data and tested its performance for three months.

[0137] In this embodiment, environmental data inside and outside the vehicle and user behavior data are collected in real time by a sensor network, including in-vehicle temperature, humidity, vehicle speed, air conditioning usage, and window opening status. After noise filtering and outlier removal, these data are input into a multi-channel variational encoder for hierarchical encoding of high frequency, low frequency, and sudden events. This step ensures comprehensive analysis and response to rapidly changing in-vehicle temperature and humidity, long-term stable air conditioning usage status, and sudden window opening operations. The spatiotemporal context perception module performs dynamic time series analysis on these encoded data through a spatiotemporal sliding window and a bidirectional spatiotemporal attention mechanism, enabling the system to identify and combine historical and real-time data for adaptive adjustment.

[0138] When a vehicle is stopped for an extended period, the temperature and humidity inside the vehicle gradually increase, especially in direct sunlight during the summer, where the temperature can quickly rise above 35°C. The system can detect temperature trends and adjust the fragrance remaining display data in real time through a dynamic prediction and control module and a fragrance remaining prediction model.

[0139] When driving with the windows open for extended periods, the temperature and humidity inside the vehicle fluctuate significantly, and frequent air movement affects the effectiveness of aromatherapy diffusion. Test data shows that when the windows are open for more than 20 minutes and the vehicle speed is maintained at 40 km / h, the system of the present invention can capture these dynamic changes through its spatiotemporal contextual awareness module and, combined with a contextual weighting mechanism, adjust the fragrance remaining display data in real time. Compared to traditional systems, this system reduces fragrance waste by 20% and improves the accuracy of remaining level management to over 85%.

[0140] When the air conditioner is on for extended periods, the humidity inside the vehicle decreases significantly. As the humidity gradually decreases from 50% to 30%, the system uses a multi-channel variational encoder and a fragrance level prediction model to analyze environmental data and user behavior, ensuring accurate display of the fragrance level even in the presence of large humidity fluctuations.

[0141] Table 1 Performance data of aromatherapy intelligent balance management in different scenarios

[0142]

[0143] Table 2 Analysis data of the feedback mechanism and self-learning effect of the system of the present invention

[0144]

[0145] The data analysis results in Tables 1 and 2 show that the proposed system significantly outperforms traditional systems in terms of margin management accuracy under various in-vehicle environments and usage scenarios. Specifically, in a long-term parking scenario, when the interior temperature rises to 35°C and the air conditioning is set to medium cooling, the proposed system reduces the margin error rate from 20% with the traditional system to 8%. This demonstrates that the system can flexibly adjust to real-time temperature and humidity changes, ensuring efficient margin management under static high-temperature conditions.

[0146] In scenarios where the vehicle was driving with the windows open for more than 20 minutes, at an interior temperature of 30°C and a humidity of 70%, without air conditioning, the system of the present invention maintained a margin error rate of 7%, compared to 18% for conventional systems, demonstrating superior adaptability and stability. This data demonstrates that the spatiotemporal contextual awareness module of the present invention can achieve more accurate real-time responses when detecting changes in in-vehicle air flow and user behavior.

[0147] In a test scenario with the air conditioner on for extended periods, with the air conditioner set to high cooling speed, the interior temperature maintained at 25°C, and the humidity reduced to 30%, the system's margin error rate significantly decreased to 6.5%, compared to 15% for the conventional system. This performance demonstrates the advantages of the present invention's adaptive optimization through feedback integration and self-supervised learning in the presence of large humidity fluctuations.

[0148] Data from nighttime driving scenarios with open windows further validated the effectiveness of the present invention. When the interior temperature was 28°C, the humidity was 50%, and the air conditioning was set to low-speed heating, the system maintained a margin error rate of 7.5%, compared to 17% for conventional systems. This demonstrates that the present invention, through contextual weight injection and real-time feature adjustment mechanisms, enables more refined management and optimization in complex and changing environments.

[0149] The user feedback results in Table 2 further support these conclusions. User satisfaction scores reached 8 and 9 in the scenarios of long-term parking and prolonged air conditioning operation, respectively, demonstrating a high level of user appreciation for the system's precise control in various situations. The system also excelled in self-learning adjustments, maintaining a self-learning frequency of 1 to 2 times per hour, demonstrating its ability to promptly adapt to changes in the vehicle's internal and external environments and user behavior. For example, in the "long-term parking" scenario, the system reduced its margin error from 10% to 7% through self-supervised learning, improving prediction accuracy by 30%.

[0150] Comprehensive analysis of these data shows that the present invention demonstrates excellent margin management performance in different in-vehicle scenarios, especially under complex conditions such as long-term window opening, air conditioning use, and temperature and humidity changes. It has excellent adaptability and optimization capabilities, significantly improving user experience and system efficiency.

[0151] As can be seen from the above examples, the present invention achieves continuous optimization and dynamic adjustment in complex and changing driving environments through self-supervised learning and feedback integration. The system maintains high levels of user satisfaction across various scenarios and significantly reduces the error rate in fragrance level prediction, demonstrating greater adaptability and prediction accuracy compared to traditional systems.

[0152] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. The intelligent aromatherapy reserve calculation method based on dynamic perception of vehicle body data is characterized by: The steps include: S1. Collect environmental data, user behavior data, and vehicle status data through sensors inside and outside the vehicle, and perform noise filtering, outlier removal, and standardization on the data to generate an initial data set. S2. Input the initial data set into a multi-channel variational encoder, classify and encode it according to high frequency, low frequency and sudden events, output a hierarchical feature representation, and realize feature information exchange through the interactive network between channels, and dynamically adjust the variational loss function of the encoder; S3. Input the hierarchical feature representation into the spatiotemporal context perception module, capture the time series data features through the spatiotemporal sliding window, combine the bidirectional spatiotemporal attention mechanism to generate spatiotemporal feature representation, and inject historical and real-time context weights; S4. Input the spatiotemporal feature representation into the heterogeneous feature mapping layer to perform spatiotemporal feature mapping and conversion, and map to generate a unified data representation; S5. Input the unified data representation into the dynamic prediction and control module, dynamically adjust the input feature weights based on real-time feedback, and generate an optimized feature representation; S6. Inputting the optimized feature representation into the aromatherapy remainder prediction model, performing remainder prediction using a mixed Gaussian process, and updating the parameters of the aromatherapy remainder prediction model in combination with variational inference; S7: Input the fragrance remaining amount prediction result into the feedback integrator, collect user feedback and in-vehicle and out-of-vehicle data for self-supervised learning. When long-term deviations or new data features are detected, the dynamic intelligent reconstruction and optimization learning mechanism is triggered to update the structure and optimize the parameters of the fragrance remaining amount prediction model; S8. Integrate the updated fragrance remaining prediction model into the vehicle system to display the fragrance remaining status in real time and issue personalized notifications and usage suggestions based on user behavior and fragrance remaining prediction results; The S2 specifically includes: S21, input the environmental data and user behavior data in the initial dataset into the multi-channel variational encoder; S22. In a multi-channel variational encoder, hierarchically classify the input data into three categories: high frequency, low frequency, and sudden events. The high frequency channel extracts rapidly changing environmental features in the input data, the low frequency channel extracts long-term stable features in the input data, and the sudden event channel extracts unpredictable abnormal changes in the input data. S23, feature encoding is performed on the data in each channel, and the high-frequency feature encoder generates a high-frequency feature vector V high , the low-frequency feature encoder generates a low-frequency feature vector V low , the emergency feature encoder generates the emergency feature vector V event ; S24, feature information is exchanged through the interactive network between channels, and the high-frequency feature vector V high , low-frequency eigenvector V low and the emergency feature vector V event Perform interactive processing and output comprehensive feature matrix M exchange : Among them, k represents the dimension of the feature vector, and α represents the normalization coefficient; S25. During the feature information exchange process, the variational loss function of the encoder is dynamically adjusted using co-correlation analysis to optimize the encoding quality of each feature channel: Among them, L var represents the loss function, n represents the total number of features in the input dataset, KL represents the divergence value, γ i represents the weight coefficient, M represents the number of additional features, q(z|x i ) represents the encoded posterior distribution, p(z) represents the prior distribution, represents the standard deviation of the input feature, ∈ represents the smoothing term, θ represents the adjustment factor, μ represents the offset, β and δ represent the adjustment parameters, T represents the time step, and λ represents the scaling parameter; S26. Output the adjusted hierarchical feature representation.

2. The method for calculating the intelligent aromatherapy remainder based on dynamic perception of vehicle body data according to claim 1 is characterized in that: The vehicle body status data specifically includes the vehicle start / stop status, whether the windows are open, and whether the air conditioner in the vehicle is turned on.

3. The method for calculating the intelligent aromatherapy remainder based on dynamic vehicle body data perception according to claim 1 is characterized in that: The S3 specifically includes: S31. Inputting the hierarchical feature representation into a spatiotemporal context awareness module, wherein the spatiotemporal context awareness module performs a temporal analysis on the input multi-channel features to capture the dynamic changes of environmental data and user behavior in the temporal and spatial dimensions; S32. Apply a spatiotemporal sliding window strategy to divide the input feature data into different time windows for processing. The spatiotemporal sliding window dynamically adjusts its size and step size according to vehicle speed, environmental conditions, and data changes to capture both short-term and long-term time series features. S33. Within each spatiotemporal sliding window, a bidirectional spatiotemporal attention mechanism is used to perform feature analysis. The bidirectional spatiotemporal attention mechanism captures the association between the current time step and past features through a forward attention network, and captures the association between the current time step and future features through a backward attention network, generating an attention weight that covers global temporal information. S34, based on the attention weights covering global temporal information generated by the bidirectional spatiotemporal attention mechanism, weights the features within the window, highlights the attention of important features related to the current task, suppresses noise and irrelevant features, and forms a feature representation with temporal correlation and context association; S35. Injecting historical context information and real-time context information into the spatiotemporal feature representation, and dynamically combining the weights of historical data and real-time data through context weight injection technology; S36. Output the spatiotemporal feature representation that combines historical and real-time context information as input for unified data representation.

4. The method for calculating the intelligent aromatherapy remainder based on dynamic vehicle body data perception according to claim 1 is characterized in that: The S4 specifically includes: S41, inputting the spatiotemporal feature representation output by the spatiotemporal context perception module into a heterogeneous feature mapping layer, wherein the heterogeneous feature mapping layer is composed of multiple parallel processing units, each of which performs preliminary analysis and segmentation on features of different sources and time scales; S42. Through the double normalization strategy and adaptive filtering algorithm, the weight distribution of high-frequency features and low-frequency features is gradually balanced, and dynamic range adjustment is adopted to maintain a stable distribution of features from each source, reducing the conversion bias caused by inconsistent data scales; S43. In the heterogeneous feature mapping conversion network, a multi-layer convolution combined with a deformable convolution is used to map and reconstruct features layer by layer, gradually transforming the original spatiotemporal features into a comprehensive representation with a unified feature space. The deformable convolution is used to automatically adjust the shape of the convolution kernel to adapt to changes in features in different time periods. S44. During the feature mapping process, contextual features are embedded in combination with the output of the context-aware module. The historical features are organically integrated with the current real-time features through the contextual attention mechanism to form a real-time context weight mapping. Through the dynamic adjustment of weight priorities, fine-grained weight assignment is performed on the input features. S45. Adopt a multi-channel feature aggregation strategy to perform hierarchical aggregation processing on the mapped high-frequency, low-frequency and emergency event features, introduce an interaction weight matrix between channels, and achieve interactive enhancement of features and complementarity of multi-dimensional information; S46. Perform stability check on the final unified data representation before output, and perform output correction in combination with the feature smoothing mechanism.

5. The method for calculating the intelligent aromatherapy remainder based on dynamic perception of vehicle body data according to claim 1 is characterized in that: The S5 specifically includes: S51. Input the generated unified data representation into a dynamic prediction and control module, wherein the dynamic prediction and control module includes a feature weighting unit and an adaptive feedback unit, which is used to adjust the weights of input features in real time to adapt to environmental changes; S52. In the feature weighting unit, a multi-layer adaptive weight assignment algorithm is used to perform weighted processing on the input unified data representation. The multi-layer adaptive weight assignment algorithm automatically assigns weights based on the rate of change and correlation of features at different time points to generate a preliminary optimized feature matrix. The weights are updated through an iterative process to adapt to dynamic changes in the environment and user behavior. S53: Receive, through the adaptive feedback unit, changes in external environment data and user behavior data in real time, compare the feedback results with the current weight distribution, and adjust the weight distribution strategy to gradually learn and adapt to changes in input features under different environmental conditions based on an incremental learning mechanism; S54, screening the preliminarily optimized feature matrix, using a feature selection algorithm to eliminate redundant and low-correlation features, retaining high-correlation features, and ensuring high information density, wherein the feature selection algorithm uses a weight pruning technology; S55. The filtered feature matrix is ​​passed to the dynamic control strategy module, and the feature matrix is ​​finally optimized in combination with the context association model. The context association model adjusts the time and environment priorities of the features based on historical data and real-time feedback to generate a comprehensive optimized feature representation; S56. Output the optimized feature representation as the input of the aromatherapy remainder prediction module.

6. The method for calculating the intelligent aromatherapy remainder based on dynamic perception of vehicle body data according to claim 1 is characterized in that: The S6 specifically includes: S61, inputting the optimized feature representation into an aromatherapy residual quantity prediction model, wherein the aromatherapy residual quantity prediction model adopts a mixed Gaussian process structure and combines multi-dimensional feature input to achieve modeling of complex nonlinear relationships between features; S62, the aromatherapy remaining quantity prediction model consists of a two-layer feature processing unit. The first layer is the input feature mapping layer, which embeds the optimized feature representation into a high-dimensional feature space. The second layer adopts a multi-kernel Gaussian process structure, combining linear kernel, radial basis kernel and polynomial kernel functions to capture the complex interactive relationship between high-frequency, low-frequency and emergency event features. S63. In the Gaussian process core layer, define the margin prediction formula and implement adaptive adjustment of margin prediction by dynamically weighing different feature combinations: Among them, y 余量 (t) represents the prediction result of aromatherapy residue, μ 高频 (t) represents the predicted mean of high-frequency features, μ 低频 (t) represents the predicted mean of low-frequency features, μ 突发 (t) represents the predicted mean of the emergency event characteristics, σ 低频 (t) represents the prediction standard deviation of low-frequency features, σ 高频 (t) represents the prediction standard deviation of high-frequency features, σ 突发 (t) represents the predicted standard deviation of the emergency event characteristics, α 1，j , α 2，l and λ r represents the weight adjustment coefficient, β 1，i and β 2，p Indicates the integral adjustment coefficient, γ 1，i and γ 2，p represents the scaling parameter, δ3 represents the global adjustment coefficient, f j (t) represents the input feature gradient function at time t, h l (t) represents the input feature correction function at time t, g r (t) represents the correction function of the sudden event, k represents the number of weight coefficients in the high-frequency feature, m represents the number of adjustment parameters of the high-frequency feature, n represents the number of adjustment coefficients of the low-frequency feature, q represents the number of adjustment parameters for time series integration in the low-frequency feature, s represents the number of correction functions in the sudden event, u represents the number of time steps used for cumulative standard deviation calculation in the sudden event part, t v represents the low-frequency features at time step v, θ1 and θ2 represent threshold parameters; S64, using variational inference technology to update model parameters in real time, optimize the posterior distribution of the Gaussian process, and ensure that the model can adapt to changes in input; S65. During the residual forecast process, combine the uncertainty quantitative analysis of the model to estimate the confidence interval of the forecast results to identify and address potential forecast biases; S66: Output aromatherapy remaining amount prediction result y 余量 (t), combining multidimensional features and uncertainty information to generate aromatherapy remainder prediction results with high confidence.

7. The method for calculating the intelligent aromatherapy remainder based on dynamic vehicle body data perception according to claim 1 is characterized in that: The S7 specifically includes: S71. Inputting the prediction result of the fragrance remaining amount prediction model into a feedback integrator, wherein the feedback integrator continuously collects data from the vehicle's internal and external environments, including temperature, humidity, vehicle speed, and user operation behavior information, and combines it with user feedback to form a multi-dimensional data set; S72. Based on the multi-dimensional data set, compare the actual environmental data with the predicted results, generate a deviation analysis report, identify the deviation between the predicted value and the actual value, and determine whether the deviation exceeds the preset long-term threshold; S73. When the deviation analysis results show that the set long-term threshold is exceeded or a new feature appears, the self-supervised learning module is activated to generate pseudo labels from the unlabeled data, expand the training data set, automatically learn new environmental features, and improve the adaptive ability of the fragrance remaining prediction model to cope with the environment; S74, inputting the expanded training data set into the incremental learning unit of the fragrance remaining prediction model, wherein the incremental learning unit performs a progressive parameter update on the fragrance remaining prediction model to maintain efficient operation of the fragrance remaining prediction model in a real-time environment and optimize prediction performance; S75. When frequent long-term deviations or new data features are detected, a dynamic intelligent reconstruction mechanism is triggered to automatically adjust the structure of the aromatherapy remainder prediction model, including optimizing the model hierarchy, reconfiguring the weight distribution, and adjusting the feature path; S76. After completing the structural adjustment and parameter optimization of the fragrance remaining amount prediction model, the improved fragrance remaining amount prediction model is applied to the vehicle-mounted system, and the updated version information and adjustment details of the fragrance remaining amount prediction model are recorded.

Images