Vehicle-mounted personalized massage anti-carsickness control system based on bumping mode prediction

Abstract

The invention discloses a vehicle-mounted personalized massage anti-carsickness control system based on bumping mode prediction, and aims to effectively relieve the problem of passenger carsickness induced by bumping of a vehicle. The system comprises a road traffic bump database, a first detection module, a bump prediction module, a control module and a vehicle-mounted massager, the road traffic bump database is used for storing standard bump data; the first detection module is used for acquiring vehicle driving data; the bumping prediction module is used for calling corresponding standard bumping data according to the vehicle driving data and predicting future bumping data of the current vehicle; the control module is used for generating preliminary massage parameters according to the current state of the user, adjusting the preliminary massage parameters in real time according to the future bumping data, and acting the adjusted massage parameters on the vehicle-mounted massager. The carsickness risk in the vehicle-mounted massage process can be reduced through accurate prediction of the dizzy bumping and dynamic adaptation of the massage parameters, and the carsickness prevention and relieving effect is greatly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent control, in particular to a vehicle-mounted personalized massage anti-car-sickness control technology based on bumping mode prediction, and is especially suitable for relieving car sickness caused by vehicle bumping through dynamic adjustment of massage parameters. BACKGROUND

[0002] Car sickness is becoming more and more prominent today. Bumping during vehicle driving stimulates the human vestibular balance system, causing nervous system disorders, such as dizziness, nausea, vomiting, and other car sickness symptoms, which directly affects the comfort and safety of travel.

[0003] Existing methods for relieving car sickness mainly include drugs and non-drugs. Drugs such as car sickness drugs and car sickness patches can inhibit the sensitivity of the vestibular nerve, but they need to be taken in advance and have side effects such as drowsiness and dry mouth. Among the non-drug methods, adjusting the riding posture and opening the window for ventilation are limited by the environment, and distracting attention has little effect on severe car sickness.

[0004] In recent years, physical therapies such as massage have gained attention due to their lack of drug side effects. For example, massaging the Neiguan acupoint can relieve car sickness symptoms, but traditional vehicle-mounted massage is disturbed by bumping, resulting in unstable force, which cannot accurately act on the acupoint or muscle group, making it difficult to achieve a sustained and effective car sickness prevention effect.

[0005] Therefore, how to reduce the impact of bumping during vehicle-mounted massage and ensure the anti-car-sickness effect of massage is a problem that needs to be solved by those skilled in the art. SUMMARY

[0006] Therefore, the present application provides a vehicle-mounted personalized massage anti-car-sickness control system based on bumping mode prediction, which accurately predicts the bumping mode that causes dizziness and dynamically adapts the massage parameters to actively offset the stimulation of the vestibular system by bumping, thereby reducing the risk of car sickness from the root cause and significantly improving the prevention and real-time relief effect of car sickness.

[0007] To achieve the above purpose, the present application adopts the following technical solutions:

[0008] A vehicle-mounted personalized massage anti-car-sickness control system based on bumping mode prediction, comprising a road traffic bumping database, a first detection module, a bumping prediction module, a control module, and a vehicle-mounted massager.

[0009] The road traffic bumping database is used to store standard bumping data; the first detection module is used to obtain vehicle driving data; the bumping prediction module is used to retrieve the corresponding standard bumping data according to the vehicle driving data and predict the future bumping data of the current vehicle.

[0010] The control module is configured to generate preliminary massage parameters, and adjust the preliminary massage parameters in real time according to the future jolt data, and apply the adjusted massage parameters to the vehicle-mounted massager.

[0011] The control module generates preliminary massage parameters based on the current physiological state of the user, dynamically adjusts the massage intensity, frequency and action site in real time in combination with the predicted dizziness-causing jolt data, actively compensates for the jolt interference, makes the massage stimulation accurately offset the adverse stimulation of the vestibular system, and maximizes the anti-car-sickness effect.

[0012] Preferably, the second detection module is further configured to collect physiological data of the user, input the collected physiological data into the trained first recognition model, confirm the car-sickness type through the first recognition model, and map the car-sickness type to the corresponding preliminary massage parameters for output.

[0013] Preferably, the third detection module is further configured to acquire a user image; the first recognition model extracts features according to the physiological data and the user image, and finally identifies the car-sickness type according to the physiological features and the gender of the user.

[0014] Preferably, the traffic database is configured to interact with the control module through a vehicle networking module, store standard jolt data in a classified manner, and each piece of data contains a road section ID and a corresponding jolt signal; the jolt signal represents multi-directional acceleration at different positions of the road section.

[0015] Preferably, the jolt prediction module includes a positioning submodule, a data calling submodule and a data analysis submodule.

[0016] The positioning submodule is configured to confirm current position information through the vehicle driving data, and the position information includes a road section currently driven and a position on the current road section.

[0017] The data calling submodule is configured to generate a sql retrieval statement according to the current position information, call corresponding standard jolt data in the traffic database, divide the standard jolt data based on the current position, and obtain first jolt data and second jolt data.

[0018] The data analysis submodule is configured to analyze the difference mode of the jolt according to the first jolt data and the vehicle driving data, confirm a jolt mode conversion parameter, perform a conversion operation on the second jolt data in combination with the conversion parameter, and obtain the future jolt data.

[0019] Preferably, the control module includes a state recognition submodule and a parameter adjustment submodule.

[0020] The state recognition submodule is used to determine the type of car sickness according to the physiological state of the user, and generate corresponding level control parameters.

[0021] The parameter adjustment submodule is used to analyze the current vehicle form of the jolt mode according to the standard jolt data, and adjust the control parameters.

[0022] A vehicle-mounted personalized massage anti-car sickness control system based on jolt mode prediction, comprising the following steps:

[0023] S1: Construct a traffic database to store the full-range jolt data standards of different road sections through the traffic database;

[0024] S2: Obtain vehicle driving data, and retrieve corresponding standard jolt data in the traffic database according to the vehicle driving data;

[0025] S3: Identify the jolt mode difference by comparing the standard jolt data and the vehicle driving data, and adjust the standard jolt data of the subsequent road section according to the jolt mode difference to obtain future jolt data;

[0026] S4: Generate preliminary massage parameters and adjust the preliminary massage parameters in real time based on the future jolt data.

[0027] Preferably, S3 specifically includes:

[0028] S31: According to the positioning information in the vehicle driving data, determine the current driving road section and the relative position in the road section.

[0029] S32: Divide the standard jolt data into first jolt data of the passed road section and second jolt data of the unpassed road section with the current position as the dividing point.

[0030] S33: Analyze the difference mode of the jolt according to the first jolt data and the vehicle driving data, and confirm the conversion parameters of the jolt mode.

[0031] S34: Perform conversion operation combining the conversion parameters and the second jolt data to obtain the future jolt data.

[0032] Preferably, S4 includes:

[0033] S41: Collect user heart rate variability and galvanic skin response data through a physiological sensor.

[0034] S42: Input the HRV and GSR data into a pre-trained car sickness level classification model to output a car sickness severity score.

[0035] S43: Select a basic massage parameter set in a preset parameter mapping table according to the car sickness severity score.

[0036] S44: analyze the time-frequency characteristics of the future jolt data to generate a jolt offset parameter matrix.

[0037] S45: convolve the basic massage parameter set with the jolt offset parameter matrix to generate a final massage control instruction.

[0038] S46: pose-compensate and correct the massage control instruction according to real-time vehicle pose data.

[0039] According to the technical solution described above, compared with the prior art, the present disclosure provides a vehicle-mounted personalized massage anti-car-sickness control system based on jolt pattern prediction, which can predict future jolts, thereby adjusting massage parameters in a timely manner before jolts occur, so that the actual massage intensity is more accurately implemented on the user, and the massage effect is guaranteed. BRIEF DESCRIPTION OF DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to the provided drawings without creative labor.

[0041] Figure 1 A structure diagram of a vehicle-mounted personalized massage anti-car-sickness control system based on jolt pattern prediction is provided for the embodiments of the present application.

[0042] Figure 2 A structure diagram of a jolt prediction module is provided for the embodiments of the present application.

[0043] Figure 3 A structure diagram of a vehicle-mounted personalized massage anti-car-sickness control system based on jolt pattern prediction is provided for the embodiments of the present application. DETAILED DESCRIPTION

[0044] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0045] Embodiment 1

[0046] The embodiments of the present application disclose a vehicle-mounted personalized massage anti-car-sickness control system based on jolt pattern prediction, which comprises a road traffic jolt database, a first detection module, a jolt prediction module, a control module, and a vehicle-mounted massager.

[0047] The road traffic bump database is used to store standard bump data; the first detection module is used to acquire vehicle driving data; the bump prediction module is used to acquire corresponding standard bump data according to the vehicle driving data, and predict future bump data of the current vehicle; the control module is used to generate preliminary massage parameters according to the current state of the user, and adjust the preliminary massage parameters in real time according to the future bump data, and apply the adjusted massage parameters to the vehicle-mounted massager.

[0048] In the embodiment, the road traffic bump database stores the dizziness-causing bump data collected by a road condition collection vehicle, contains multi-directional acceleration time domain features, and can accurately represent the stimulation intensity of bumps of different road sections on the human vestibular system, thereby providing a data basis for predictive adjustment of anti-car-sickness massage parameters.

[0049] The vehicle driving data acquired by the first detection module can be used as a basis to find corresponding standard bump data in the road traffic bump database. The extracted standard bump data includes the distance of the road that has been traveled and the distance of the road to be traveled in the future, so that the distance of the actual bump and the standard bump is judged by using the distance of the road that has been traveled, as a correction basis, to correct the future standard bump data and generate a prediction result that conforms to the current actual situation.

[0050] In order to further implement the above technical solutions, a second detection module is further included, which is used to acquire physiological data of the user.

[0051] Specifically, the second detection module includes a plurality of sensors for detecting physiological data of the user, and uses a wearable device to collect physiological data of the user. The wearable device contacts the user to collect heart rate and skin electrical response.

[0052] In the embodiment, the control module includes a state recognition submodule, which is used to determine the type of car sickness according to the physiological state of the user, and generate corresponding level control parameters. The physiological data collected is input into the trained first recognition model, and the type of car sickness is confirmed by the first recognition model and mapped to the corresponding preliminary massage parameters for output.

[0053] Specifically, the first identification model adopts a hybrid architecture of three-layer convolutional neural network (CNN) combined with long short-term memory network (LSTM). The input layer is used to receive the preprocessed physiological data sequence, with a dimension of [time window length x 2] including two feature dimensions of heart rate and skin electric response. The convolutional layer includes two convolutional blocks, each block consisting of 64 3x1 convolutional kernels + BatchNorm + ReLU activation function, for extracting local features of physiological signals. The LSTM layer is a bidirectional LSTM layer with 128 neurons, for capturing dynamic association features of heart rate and skin electric response over time. The fully connected layer includes a first output layer, which is a 5-node softmax classifier corresponding to the 0-4 levels of car sickness severity; and a second output layer, which is a 1-node regressor corresponding to the car sickness probability value.

[0054] Further, a fuzzy logic mapping engine is constructed, with the input being the car sickness severity and the user gender, and the output being the massage intensity and frequency as well as the part weight, and the fuzzy sets of input and output variables are defined by a triangular membership function.

[0055] Further, a third detection module is further included for acquiring a user image. The first identification model extracts features according to the physiological data and the user image, and finally identifies the car sickness type according to the physiological features and the user gender.

[0056] In the embodiment, through the identification of the user image, facial expressions and eye state are extracted to enrich the initial information, and the user gender is also identified.

[0057] In order to further implement the above technical solutions, the road traffic database interacts with the control module through the vehicle networking module, for classifying and storing the standard bump data, each data containing a road section ID and a corresponding bump signal; the bump signal represents the multi-directional acceleration at different positions of the road section.

[0058] In order to further implement the above technical solutions, the bump prediction module includes a positioning submodule, a data calling submodule and a data analysis submodule.

[0059] The positioning submodule is used to confirm the current position information through the vehicle driving data, the position information including the current driving road section and the position on the current road section; the data calling submodule is used to generate a sql retrieval statement according to the current position information, to retrieve the corresponding standard bump data in the traffic database, and to divide the data based on the current position to obtain first bump data and second bump data; the data analysis submodule is used to analyze the bump pattern according to the first bump data, and to correct the second bump data using the same bump pattern to predict the future bump data.

[0060] Specifically, in the positioning sub-module, the latitude and longitude coordinates (accuracy 10 meters) are obtained by the vehicle-mounted GPS, combined with the three-axis acceleration and angular velocity data (sampling rate 100 Hz) collected by the vehicle-mounted inertial navigation system (INS), and the driving data such as vehicle speed and steering wheel angle are obtained by analyzing the CAN bus.

[0061] Then, the real-time trajectory points are matched with the high-precision electronic map road network by using the map matching technology based on the hidden Markov model (HMM), and the GPS positioning error is corrected to within 5 meters. The specific steps are as follows: generating a candidate road segment set: extracting all road segments within a radius of 50 meters centered on the current trajectory point calculating transition probability: determining the state transition probability according to the angle between the vehicle driving direction and the candidate road segment direction, distance and other parameters state decoding: solving the maximum probability path by the Viterbi algorithm to determine the current driving road segment. The driving distance of the vehicle on the current road segment is calculated by dead reckoning using the vehicle wheel speed pulse signal and steering wheel angle data.

[0062] Further, in order to ensure the positioning accuracy, when the vehicle passes through the road sign point (such as intersection, bridge), the absolute position is obtained by RFID tag or visual recognition technology, and the dead reckoning result is calibrated.

[0063] In the data calling process, the road segment to be extracted is confirmed by road segment matching, and then the range in the road segment is extracted according to the current position. For example: the bump data (historical data) of the 500-meter road segment that the vehicle has traveled is called to analyze the actual bump pattern of the current road segment; the standard bump data (predicted data) of the 1000-meter road segment in the future of the vehicle is called as the basis for bump prediction.

[0064] In the data analysis sub-module, the driving information of the current vehicle and the called standard bump data are input into the trained second neural network model, and the second neural network model analyzes the difference between the bump pattern of the current vehicle and the standard situation according to the first bump data and the driving data of the current vehicle, and adjusts the second bump data according to the difference, so that it fits the current vehicle operation.

[0065] Specifically, the second neural network model adopts the deep learning architecture of encoding-conversion-decoding, and its core goal is to learn and model the nonlinear conversion relationship (Transfer Pattern) between the actual driving bump pattern of the current vehicle and the standard bump pattern in the road traffic bump database, and apply this relationship to the standard bump data of the future road segment to generate the future bump prediction that fits the actual operation of the current vehicle. The model structure includes the following key parts and data processing processes:

[0066] The dual-channel feature extraction encoder inputs the actual data stream and the standard data stream into two channels respectively, and the two data streams are respectively input into two one-dimensional convolutional neural network modules (i.e., CNN modules) which have the same structure but independent parameters, for extracting local space-time features. After passing through multiple CNN modules, the two data streams are respectively encoded into high-dimensional feature vector sequences. In addition, the vehicle driving state information corresponding to the historical time period is embedded into a state feature vector as auxiliary input through a fully connected network; finally, the feature vector sequences output by the two channels (representing the actual and standard historical bump pattern features) and the embedded vehicle state feature vector are output.

[0067] The difference pattern learning and conversion relationship modeling module obtains the actual feature sequence, the standard feature sequence and the vehicle state feature vector from the encoder for processing. At the feature level, the difference feature sequence is obtained by calculating the difference vector between the actual and standard feature vectors at the corresponding time step. After the difference feature sequence and the vehicle state feature vector are spliced, they are input into a recurrent neural network layer, and then into a multi-layer perceptron to obtain the conversion parameter.

[0068] The future bump prediction decoder obtains the predicted future bump data according to the conversion parameter. Specifically, the future standard bump data is input into a 1D CNN feature extractor (similar in structure to the standard channel in the encoder, and part of the weights can be shared or independent). The CNN encodes it into a future standard bump feature sequence; using the learned conversion parameter (TP), each feature vector of the future standard bump feature sequence is subjected to a conversion operation; after the conversion operation, a predicted future actual bump feature sequence is obtained, which is input into a 1D transposed convolutional network (Deconvolutional Network) or a temporal upsampling network. The network is responsible for decoding / reconstructing the high-dimensional feature sequence back to the original multi-directional acceleration time sequence space to obtain the future bump data of the current vehicle.

[0069] In the embodiment, the control module further includes a parameter adjustment sub-module, which adjusts the massage parameters output by the first identification model in real time according to the future bump data on the basis of the first identification model.

[0070] Specifically, the feature sampling of the sliding window is performed on each acceleration axis and the combined acceleration in the future bump data to obtain a sliding window feature vector, including intensity features, impact features and frequency features. Then, the feature scoring is performed through a pre-set scoring rule engine or regressor to obtain a scoring sequence representing the bump influence size, and the massage intensity, massage frequency and massage site weight are adjusted according to the sequence.

[0071] Embodiment 2

[0072] As Figure 3, based on the same inventive concept, the embodiment of the present application discloses a kind of based on jolt mode prediction's vehicle individualized massage anti-car-sickness control system method, comprising the following steps:

[0073] S1: construct road traffic database, store the standard of whole jolt data of different road sections in traffic database by traffic database;

[0074] S2: obtain vehicle driving data, and according to vehicle driving data, corresponding standard jolt data is called in traffic database;

[0075] S3: by comparing standard jolt data and vehicle driving data, the jolt mode difference is identified, and the standard jolt data of subsequent road is adjusted according to jolt mode difference, and future jolt data is obtained;

[0076] S4: preliminary massage parameters are generated in combination with user car-sickness physiological characteristics, parameters are dynamically optimized based on predicted future jolt data, and through real-time confrontation of massage stimulation and jolt interference, early prevention and immediate relief of car-sickness symptoms are realized.

[0077] In order to further implement the above technical solutions, the S3 includes:

[0078] S31: according to the positioning information in vehicle driving data, the current driving road section and the relative position in road section are determined;

[0079] S32: with current position as demarcation point, standard jolt data is divided into first jolt data of road section passed and second jolt data of road section not passed;

[0080] S33: according to first jolt data and vehicle driving data, the difference mode of jolt is analyzed, and the conversion parameter of jolt mode is confirmed;

[0081] S34: the conversion parameter is combined with second jolt data to carry out conversion operation, and the future jolt data is obtained.

[0082] Further, the S4 includes:

[0083] S41: the heart rate variability and galvanic skin response data of user are collected by physiological sensor;

[0084] S42: the HRV and GSR data are input into pre-trained car-sickness grade classification model, and car-sickness severity score is output;

[0085] S43: according to the car-sickness severity score, the basic massage parameter set is selected in preset parameter mapping table;

[0086] S44: the time-frequency characteristics of future jolt data are analyzed, and jolt offset parameter matrix is generated;

[0087] S45: Convolution and fusion of the basic massage parameter set and the jolt cancellation parameter matrix to generate a final massage control instruction;

[0088] S46: Pose compensation correction of the massage control instruction according to real-time vehicle pose data.

[0089] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts between the various embodiments can be referred to each other. For the device disclosed by the embodiments, since it corresponds to the method disclosed by the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0090] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to the embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A personalized in-vehicle massage and anti-motion sickness control system based on bumpy pattern prediction, characterized in that: It includes a road traffic bump database, a first detection module, a bump prediction module, a control module and a vehicle massager; The road traffic bump database is used to store standard bump data; The first detection module is used to obtain vehicle driving data; The bump prediction module is used to retrieve corresponding standard bump data according to the vehicle driving data and predict future bump data of the current vehicle; The control module is used to generate preliminary massage parameters, adjust the preliminary massage parameters in real time according to the future bump data, and apply the adjusted massage parameters to the vehicle-mounted massager.

2. The vehicle-mounted personalized massage and anti-motion sickness control system based on bumpy pattern prediction according to claim 1, characterized in that: It also includes a second detection module for acquiring physiological data of the user as the user's current status data; the control module generates the preliminary massage parameters according to the user's current status data.

3. The vehicle-mounted personalized massage and anti-motion sickness control system based on bumpy pattern prediction according to claim 2, characterized in that: It also includes a third detection module for acquiring a user image; when generating the preliminary massage parameters, the control module is also used to determine the user's gender based on the user image.

4. The vehicle-mounted personalized massage and anti-motion sickness control system based on bumpy pattern prediction according to claim 1, characterized in that: The road traffic database exchanges data with the control module through the vehicle networking module and is used to classify and store standard bump data. Each data contains a road section ID and a corresponding bump signal; the bump signal represents the multi-directional acceleration at different positions on the road section.

5. The vehicle-mounted personalized massage and anti-motion sickness control system based on bumpy pattern prediction according to claim 4 is characterized in that: The bump prediction module includes a positioning submodule, a data calling submodule and a data analysis submodule; The positioning submodule is used to confirm the current position information through the vehicle driving data, and the position information includes the current driving section and the position on the current section; The data calling submodule is used to generate an SQL search statement according to the current position information, retrieve the corresponding standard bump data from the traffic database, and divide it based on the current position to obtain the first bump data and the second bump data; The data analysis submodule is used to analyze the difference pattern of bumps according to the first bump data and the vehicle driving data, determine the conversion parameters of the bump pattern, and perform a conversion operation on the conversion parameters and the second bump data to obtain the future bump data.

6. The vehicle-mounted personalized massage and anti-motion sickness control system based on bumpy pattern prediction according to claim 1, characterized in that: The control module includes a state identification submodule and a parameter adjustment submodule; The state recognition submodule is used to determine the type of motion sickness according to the user's physiological state and generate control parameters of corresponding levels; The parameter adjustment submodule is used to analyze the bump pattern of the current vehicle type according to the standard bump data and adjust the control parameters.

7. A vehicle-mounted personalized massage anti-motion sickness control method based on bumpy pattern prediction, characterized in that: Take the following steps to proactively prevent and alleviate motion sickness caused by turbulence: S1: Building a road traffic database, storing full-course bumpy data standards of different road sections through the traffic database; S2: Acquire vehicle driving data, and retrieve corresponding standard bump data from the traffic database according to the vehicle driving data; S3: identifying a difference in bump patterns by comparing the standard bump data with the vehicle travel data, and adjusting the standard bump data for a subsequent journey according to the difference in bump patterns to obtain future bump data; S4: Generate preliminary massage parameters, and adjust the preliminary massage parameters in real time based on future turbulence data.

8. The vehicle-mounted personalized massage and anti-motion sickness control system based on bumpy pattern prediction according to claim 7, characterized in that: The S3 specifically includes: S31: Determine the current driving section and the relative position in the section based on the positioning information in the vehicle driving data; S32: using the current position as a dividing point, dividing the standard bump data into first bump data of the passed section and second bump data of the not passed section; S33: analyzing a difference pattern of bumps according to the first bump data and the vehicle driving data, and determining a conversion parameter of the bump pattern; S34: performing a conversion operation on the second turbulence data in combination with the conversion parameter to obtain the future turbulence data.

9. The vehicle-mounted personalized massage and anti-motion sickness control system based on bumpy pattern prediction according to claim 7 or 8, characterized in that: The S4 includes: S41: Collecting user's heart rate variability and skin galvanic response data through physiological sensors; S42: Inputting the user's heart rate variability and skin galvanic response data into a pre-trained motion sickness level classification model to output a motion sickness severity score; S43: selecting a basic massage parameter set in a preset parameter mapping table according to the motion sickness severity score; S44: analyzing the time-frequency characteristics of the future turbulence data to generate a turbulence cancellation parameter matrix; S45: performing convolution fusion on the basic massage parameter set and the turbulence compensation parameter matrix to generate a final massage control instruction; S46: Performing posture compensation correction on the massage control instruction according to the real-time posture data of the vehicle.

Images