Message reminding method and system based on smart watch

Abstract

The invention relates to the technical field of intelligent equipment, in particular to a message reminding method and system based on a smart watch, and the method comprises the steps: obtaining historical riding data, real-time riding data and physiological data of a user through setting a riding reminding message mapping set; constructing a user fatigue recognition model, calculating a first riding fatigue coefficient according to the historical riding data and the real-time riding data, and if the first riding fatigue coefficient exceeds a preset fatigue coefficient threshold value, sending a first fatigue message prompt to the mobile phone; constructing a user physiological state recognition model, analyzing the physiological data, calculating a second physiological abnormality coefficient, and if the second physiological abnormality coefficient exceeds a preset physiological abnormality coefficient threshold value, sending a second physiological state abnormality message prompt to the mobile phone; and if the two coefficients are both in a threshold range, calculating a first recommended speed based on the two coefficients, and if an error between the riding speed and the first recommended speed exceeds a preset speed threshold, sending a third speed mismatching message prompt to the mobile phone. The state of a riding user can be monitored in real time, and riding safety is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of smart devices, and in particular to a message reminder method and system based on a smart watch. Background Art

[0002] Smartwatches are multifunctional wearable devices that connect to smartphones to provide notifications, health monitoring, and fitness tracking. They can receive calls and text messages, monitor heart rate, count steps, and track sleep quality. With technological advancements, the functionality of smartwatches continues to expand, making them an indispensable assistant in our daily lives.

[0003] With the rise of a nationwide fitness trend, cycling is gaining popularity as a healthy, environmentally friendly, and enjoyable form of exercise. Cycling not only enhances cardiopulmonary function and improves muscle strength, but also helps people relax and relieve stress. More and more people are choosing cycling as a daily leisure activity.

[0004] However, cyclists face numerous challenges during cycling, such as fatigue management, speed control, and environmental fluctuations. Excessive fatigue can lead to accidents, while excessive speed increases the likelihood of injury. Furthermore, weather changes and road conditions can also impact cyclist safety. Therefore, timely notifications are crucial for improving cycling safety and the overall cycling experience. Smartwatches' real-time monitoring and feedback capabilities can help cyclists stay informed of their own status and the external environment, enabling them to make better cycling decisions.

[0005] To this end, a message reminder method and system based on a smart watch are proposed. Summary of the Invention

[0006] The present invention aims to provide a smartwatch-based message reminder method and system. This invention relates to the field of smart device technology, and specifically to a smartwatch-based message reminder method and system. The present invention establishes a cycling reminder message mapping set to obtain a user's historical cycling data, real-time cycling data, and physiological data. A user fatigue recognition model is constructed to calculate a first cycling fatigue coefficient based on the historical and real-time cycling data. If the coefficient exceeds a preset fatigue coefficient threshold, a first fatigue reminder message is sent to the mobile phone. A user physiological state recognition model is constructed to analyze the physiological data and calculate a second physiological abnormality coefficient. If the coefficient exceeds a preset physiological abnormality coefficient threshold, a second physiological state abnormality reminder message is sent to the mobile phone. If both coefficients are within a threshold range, a first recommended speed is calculated based on the two coefficients. If the cycling speed deviates from the first recommended speed by more than a threshold, a third speed mismatch reminder message is sent to the mobile phone. This invention can monitor the cyclist's status in real time and ensure cycling safety.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A message reminder method based on a smart watch, comprising:

[0009] S1. Set a riding reminder message mapping set, the set includes each message reminder key-value pair, the key is used to represent the riding event number that triggers the message reminder, and the value is used to represent the message that needs to be transmitted to the first medium when the riding event occurs. The first medium is a mobile phone connected to the smart watch;

[0010] S2. Obtain the user's historical riding data, the first riding data, and the second physiological data;

[0011] S3. Build a user fatigue recognition model, identify the historical riding data and the first riding data, obtain a first riding fatigue coefficient, determine whether the first riding fatigue coefficient exceeds a preset fatigue coefficient threshold, and if so, the smartwatch triggers a first fatigue event and sends a first fatigue message reminder to the user's mobile phone;

[0012] S4. Build a user physiological state recognition model, identify the second physiological data, obtain the second user physiological abnormality coefficient, determine whether the second user physiological abnormality coefficient exceeds the preset physiological abnormality coefficient threshold, and if so, the smart watch triggers a second physiological state abnormality event and sends a second physiological state abnormality message to the user's mobile phone to remind;

[0013] S5. If both the first cycling fatigue coefficient and the second user physiological abnormality coefficient are less than corresponding thresholds, a first recommended speed is obtained based on the first cycling fatigue coefficient, the basic recommended speed, and the second user physiological abnormality coefficient. If the error between the first recommended speed and the cycling speed exceeds a preset speed threshold, the smartwatch triggers a third speed mismatch event and sends a third speed mismatch message reminder to the user's mobile phone.

[0014] Preferably, the historical riding data includes time series data of the user in various stages of past riding, including historical weather data, historical riding mileage, historical riding time, historical riding speed, historical riding slope, historical riding acceleration, historical riding road slope and historical riding road traffic volume; the first riding data includes time series data of the user during real-time riding, including weather data, riding speed, riding acceleration, riding time, riding distance, riding road slope and riding road traffic volume; the second physiological data includes heart rate, respiratory rate and body temperature.

[0015] Preferably, the user fatigue recognition model includes a first input layer, a first preprocessing layer, a first feature extraction layer, a fatigue coefficient recognition layer and a first output layer;

[0016] The first input layer is used to input the historical riding data and the first riding data into a user fatigue recognition model;

[0017] The first preprocessing layer is used to preprocess the historical riding data and the first riding data, including denoising and abnormal data deletion, to obtain historical riding identification data and first riding identification data;

[0018] The first feature extraction layer is used to extract features from the historical riding identification data and the first riding identification data through an LSTM network to obtain a first fatigue feature, where the first fatigue feature includes tolerable riding intensity, tolerable riding load, real-time riding intensity, and real-time riding load;

[0019] The fatigue coefficient recognition layer is used to analyze the first fatigue feature to obtain a first riding fatigue coefficient;

[0020] The first output layer is used to output the first cycling fatigue coefficient.

[0021] Preferably, the first riding fatigue coefficient is:

[0022]

[0023] Wherein, CFC represents the first cycling fatigue coefficient; represents the first fatigue influence coefficient; WCI represents the real-time cycling intensity; CWCI represents the tolerable cycling intensity; WCL represents the real-time cycling load; CWCL represents the tolerable cycling load; γ represents the second fatigue influence coefficient; exp represents the exponential function with e as the base.

[0024] Preferably, the user physiological state recognition model includes a second input layer, a second preprocessing layer, a second feature extraction layer, a physiological abnormality coefficient recognition layer and a second output layer;

[0025] The second input layer is used to input the second physiological data into the user physiological state recognition model;

[0026] The second preprocessing layer is used to perform data cleaning and denoising on the second physiological data to obtain second physiological recognition data respectively;

[0027] The second feature extraction layer is used to extract features from the second physiological recognition data through a convolutional neural network integrated with an attention mechanism; obtain second physiological state features, wherein the physiological state features include heart rate fluctuation amplitude, heart rate peak value, heart rate trough value, average heart rate, respiratory rate change rate, average respiratory rate, average body temperature and body temperature change amplitude;

[0028] The physiological abnormality coefficient identification layer obtains the second user's physiological abnormality coefficient by analyzing the second physiological state characteristics;

[0029] The second output layer is used to output the physiological abnormality coefficient of the second user.

[0030] Preferably, the second user's physiological abnormality coefficient is:

[0031]

[0032] Wherein, PHC represents the physiological abnormality coefficient of the second user; ψ i represents the attention mechanism weight of the i-th second physiological state feature; PSC i represents the i-th second physiological state characteristic value.

[0033] Preferably, the first recommended speed is:

[0034] RSped=MeanSped*(1+k1*CFC+k2*PHC);

[0035] Among them, RSped represents the first recommended speed; MeanSped represents the basic recommended speed, which is used to represent the average speed of the user during the riding process; k1 represents the first speed adjustment coefficient; CFC represents the first riding fatigue coefficient; k2 represents the second speed adjustment coefficient; PHC represents the second user physiological abnormality coefficient.

[0036] A message reminder system based on a smart watch, comprising:

[0037] a message reminder mechanism determination module, configured to set a cycling reminder message mapping set, the set comprising message reminder key-value pairs, where the key represents the cycling event number that triggers the message reminder, and the value represents the message that needs to be transmitted to a first medium when the cycling event occurs, the first medium being a mobile phone connected to the smartwatch;

[0038] A data acquisition module, configured to acquire the user's historical riding data, first riding data, and second physiological data;

[0039] a first judging module, configured to identify the historical riding data and the first riding data by building a user fatigue recognition model, obtain a first riding fatigue coefficient, and judge whether the first riding fatigue coefficient exceeds a preset fatigue coefficient threshold;

[0040] a second judgment module, configured to identify the second physiological data by building a user physiological state recognition model, obtain a second user physiological abnormality coefficient, and determine whether the second user physiological abnormality coefficient exceeds a preset physiological abnormality coefficient threshold;

[0041] a third judgment module, configured to determine whether the first cycling fatigue coefficient and the second user physiological abnormality coefficient are both less than corresponding preset thresholds; if so, deriving a first recommended speed based on the first cycling fatigue coefficient, the basic recommended speed, and the second user physiological abnormality coefficient; and determining whether the first recommended speed exceeds a preset speed threshold;

[0042] The execution module triggers a first fatigue event if the first cycling fatigue coefficient exceeds a preset fatigue coefficient threshold, and sends a first fatigue message reminder to the user's mobile phone; if it is determined that the second user physiological abnormality coefficient exceeds a preset physiological abnormality coefficient threshold, the smart watch triggers a second physiological state abnormality event, and sends a second physiological state abnormality message reminder to the user's mobile phone; if the error between the first recommended speed and the cycling speed exceeds a preset speed threshold, the smart watch triggers a third speed mismatch event, and sends a third speed mismatch message reminder to the user's mobile phone.

[0043] Preferably, the first riding fatigue coefficient is:

[0044]

[0045] Among them, CFC represents the first cycling fatigue coefficient; θ represents the first fatigue influence coefficient; WCI represents the real-time cycling intensity; CWCI represents the tolerable cycling intensity; WCL represents the real-time cycling load; CWCL represents the tolerable cycling load; γ represents the second fatigue influence coefficient; exp represents the exponential function with base e.

[0046] The second user's physiological abnormality coefficient is:

[0047]

[0048] Wherein, PHC represents the physiological abnormality coefficient of the second user; ψ i represents the attention mechanism weight of the i-th second physiological state feature; PSC i represents the i-th second physiological state characteristic value.

[0049] Preferably, the first recommended speed is:

[0050] RSped=MeanSped*(1+k1*CFC+k2*PHC);

[0051] Among them, RSped represents the first recommended speed; MeanSped represents the basic recommended speed, which is used to represent the average speed of the user during the riding process; k1 represents the first speed adjustment coefficient; CFC represents the first riding fatigue coefficient; k2 represents the second speed adjustment coefficient; PHC represents the second user physiological abnormality coefficient.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. The present invention constructs a user fatigue recognition model to comprehensively analyze the user's historical and real-time riding data to obtain the user's first riding fatigue coefficient. Based on the comparison and analysis of the first riding fatigue coefficient with a preset fatigue coefficient threshold, it can determine whether the user is overly fatigued, and thus send timely reminder messages to the user to help the user avoid excessive fatigue and improve riding safety and exercise efficiency.

[0054] 2. The present invention constructs a user physiological state recognition model. The model is based on the real-time physiological data of the user during riding. The convolutional neural network based on the fusion attention mechanism performs feature analysis on the physiological data, extracts the user's second physiological state characteristics, and obtains the user's second physiological abnormality coefficient. Based on the comparative analysis with the preset physiological abnormality coefficient threshold, it is determined whether the user has physiological abnormalities during riding, and a message reminder is sent to the user in time to ensure the safety of the user's riding.

[0055] 3. The present invention obtains a first recommended speed based on the user's first fatigue coefficient and second physiological abnormality coefficient, thereby realizing personalized speed recommendations. This can effectively reduce the imbalance in riding status caused by riding too fast or too slow, and help users find the optimal speed range during exercise, especially during long-term or high-intensity riding, which can help ensure the safety and experience of users during riding. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A flowchart of a message reminder method based on a smart watch provided by an embodiment of the present invention;

[0057] Figure 2 A schematic diagram of the structure of a message reminder system based on a smart watch provided by an embodiment of the present invention;

[0058] Figure 3 A schematic diagram of a user fatigue recognition model provided by an embodiment of the present invention;

[0059] Figure 4 A schematic diagram of a user physiological state recognition model provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0061] Example 1

[0062] To improve the safety of cycling, user A applied a message reminder method based on a smartwatch.

[0063] Reference Figure 1 , is a flow chart of a message reminder method based on a smart watch, including:

[0064] S1. Set a cycling reminder message mapping set, the set including message reminder key-value pairs, where the key represents the cycling event number that triggers the message reminder, and the value represents the message that needs to be transmitted to a first medium when the cycling event occurs, the first medium being a mobile phone connected to the smartwatch; the cycling events include a first fatigue event, a second physiological state abnormality event, and a third speed mismatch event; the keys in the cycling reminder message mapping set include K1, K2, and K3, corresponding to the numbers of the first fatigue event, the second physiological state abnormality event, and the third speed mismatch event, respectively; the message data is shown in Table 1;

[0065] Table 1 Information data table

[0066] serial number event message K1 First fatigue event "Your riding condition is quite fatigued" K2 Second physiological state abnormal event "Your physiological state is abnormal" K3 Third speed mismatch event "Your riding speed does not match"

[0067] S2. Obtain the user's historical riding data, the first riding data, and the second physiological data;

[0068] Furthermore, the historical riding data includes time series data of the user in various stages of past riding, including historical weather data, historical riding mileage, historical riding time, historical riding speed, historical riding slope, historical riding acceleration, historical riding road slope and historical riding road traffic volume; the first riding data includes time series data of the user during real-time riding, including weather data, riding speed, riding acceleration, riding time, riding distance, riding road slope and riding road traffic volume; the second physiological data includes heart rate, respiratory rate and body temperature.

[0069] The historical riding data is obtained from the user's past riding data, the historical weather data includes temperature, humidity and wind speed, and the weather data is obtained through the weather application software in the mobile phone connected to the smart watch. The historical riding mileage, riding mileage, historical riding time, riding time, historical riding speed, riding speed, historical riding acceleration and riding acceleration are obtained through the keep software used by the user during each riding process. The historical riding road slope, riding road slope, riding road traffic volume and historical riding road traffic volume are obtained through the map software in the mobile phone connected to the smart watch. The heart rate, respiratory rate and body temperature are obtained through the sensors in the smart watch.

[0070] S3. Build a user fatigue recognition model, identify the historical riding data and the first riding data, obtain a first riding fatigue coefficient, determine whether the first riding fatigue coefficient exceeds a preset fatigue coefficient threshold, and if so, the smartwatch triggers a first fatigue event and sends a first fatigue message reminder to the user's mobile phone;

[0071] The user fatigue recognition model schematic diagram is as follows Figure 3 As shown;

[0072] The user fatigue recognition model includes a first input layer, a first preprocessing layer, a first feature extraction layer, a fatigue coefficient recognition layer and a first output layer;

[0073] The first input layer is used to input the historical riding data and the first riding data into a user fatigue recognition model;

[0074] The first preprocessing layer is used to preprocess the historical riding data and the first riding data, including denoising and abnormal data deletion, to obtain historical riding identification data and first riding identification data;

[0075] The first feature extraction layer is used to extract features from the historical riding identification data and the first riding identification data through an LSTM network to obtain a first fatigue feature, where the first fatigue feature includes tolerable riding intensity, tolerable riding load, real-time riding intensity, and real-time riding load;

[0076] The LSTM network analyzes the user's past riding data to obtain a time dependency between the user's historical riding data and the riding intensity, as well as a time dependency between the historical riding data and the riding load. The network then derives the user's riding intensity and riding load under different past riding conditions, obtains the maximum riding intensity and the maximum riding load, and uses the maximum riding intensity and the maximum riding load as the tolerable riding intensity and the tolerable riding load, respectively. Furthermore, the network obtains the real-time riding intensity and the real-time riding load based on the time dependency between the user's historical riding data and the riding intensity, the time dependency between the historical riding data and the riding load, and the first riding data.

[0077] The fatigue coefficient recognition layer is used to analyze the first fatigue feature to obtain a first riding fatigue coefficient;

[0078] The first output layer is used to output the first cycling fatigue coefficient.

[0079] The first riding fatigue coefficient is:

[0080]

[0081] Among them, CFC represents the first cycling fatigue coefficient; θ represents the first fatigue influence coefficient; WCI represents the real-time cycling intensity; CWCI represents the tolerable cycling intensity; WCL represents the real-time cycling load; CWCL represents the tolerable cycling load; γ represents the second fatigue influence coefficient; exp represents the exponential function with base e.

[0082] This embodiment constructs a user fatigue recognition model to conduct a comprehensive analysis of the user's historical riding data and real-time riding data, thereby obtaining the user's first riding fatigue coefficient. By comparing and analyzing the first riding fatigue coefficient with a preset fatigue coefficient threshold, it is possible to determine whether the user is overly fatigued, thereby sending a message reminder to the user in a timely manner, helping the user avoid excessive fatigue and improving riding safety and exercise efficiency.

[0083] S4. Build a user physiological state recognition model, identify the second physiological data, obtain the second user physiological abnormality coefficient, determine whether the second user physiological abnormality coefficient exceeds the preset physiological abnormality coefficient threshold, and if so, the smart watch triggers a second physiological state abnormality event and sends a second physiological state abnormality message to the user's mobile phone to remind;

[0084] The schematic diagram of the user physiological state recognition model is as follows Figure 4 As shown;

[0085] The user physiological state recognition model includes a second input layer, a second preprocessing layer, a second feature extraction layer, a physiological abnormality coefficient recognition layer and a second output layer;

[0086] The second input layer is used to input the second physiological data into the user physiological state recognition model;

[0087] The second preprocessing layer is used to perform data cleaning and denoising on the second physiological data to obtain second physiological recognition data respectively;

[0088] The second feature extraction layer is used to extract features from the second physiological recognition data through a convolutional neural network integrated with an attention mechanism; obtain second physiological state features, wherein the physiological state features include heart rate fluctuation amplitude, heart rate peak value, heart rate trough value, average heart rate, respiratory rate change rate, average respiratory rate, average body temperature and body temperature change amplitude;

[0089] The physiological abnormality coefficient identification layer obtains the second user's physiological abnormality coefficient by analyzing the second physiological state characteristics;

[0090] The second output layer is used to output the physiological abnormality coefficient of the second user.

[0091] Furthermore, the second user's physiological abnormality coefficient is:

[0092]

[0093] Wherein, PHC represents the physiological abnormality coefficient of the second user; ψ i represents the attention mechanism weight of the i-th second physiological state feature; PSC i represents the i-th second physiological state characteristic value.

[0094] This embodiment constructs a user physiological state recognition model. The model is based on the real-time physiological data of the user while riding. The model performs feature analysis on the physiological data based on a convolutional neural network with a fusion attention mechanism, extracts the user's second physiological state characteristics, and obtains the user's second physiological abnormality coefficient. Based on a comparative analysis with a preset physiological abnormality coefficient threshold, it is determined whether the user has a physiological abnormality during riding, and a message reminder is sent to the user in a timely manner to ensure the user's riding safety.

[0095] S5. If both the first cycling fatigue coefficient and the second user physiological abnormality coefficient are less than corresponding preset thresholds, a first recommended speed is obtained based on the first cycling fatigue coefficient, the basic recommended speed, and the second user physiological abnormality coefficient. If the error between the first recommended speed and the cycling speed exceeds a preset speed threshold, the smartwatch triggers a third speed mismatch event and sends a third speed mismatch message reminder to the user's mobile phone.

[0096] Furthermore, the first recommended speed is:

[0097] RSped=MeanSped*(1+k1*CFC+k2*PHC);

[0098] Among them, RSped represents the first recommended speed; MeanSped represents the basic recommended speed, which is used to represent the average speed of the user during the riding process; k1 represents the first speed adjustment coefficient; CFC represents the first riding fatigue coefficient; k2 represents the second speed adjustment coefficient; PHC represents the second user physiological abnormality coefficient.

[0099] This embodiment obtains a first recommended speed based on the user's first fatigue coefficient and second physiological abnormality coefficient, thereby achieving personalized speed recommendations. This can effectively reduce imbalances in riding status caused by riding too fast or too slow, and help users find the optimal speed range during exercise, especially during long-term or high-intensity riding, which can help ensure the user's safety and experience during riding.

[0100] This embodiment establishes a cycling reminder message mapping set to obtain a user's historical cycling data, real-time cycling data, and physiological data. A user fatigue recognition model is constructed to calculate a first cycling fatigue coefficient based on the historical and real-time cycling data. If the coefficient exceeds a preset fatigue coefficient threshold, a first fatigue message reminder is sent to the mobile phone. A user physiological state recognition model is also constructed to analyze the physiological data and calculate a second physiological abnormality coefficient. If the coefficient exceeds a preset physiological abnormality coefficient threshold, a second physiological state abnormality message reminder is sent to the mobile phone. If both coefficients are within the threshold range, a first recommended speed is calculated based on these two coefficients. If the difference between the cycling speed and the first recommended speed exceeds a preset speed threshold, a third speed mismatch message reminder is sent to the mobile phone. This invention can monitor the cyclist's status in real time and ensure cycling safety.

[0101] Example 2

[0102] To improve the safety of cycling, user B applied a message reminder system based on a smartwatch;

[0103] Reference Figure 2 , for a structural diagram of a message reminder system based on a smart watch, including:

[0104] a message reminder mechanism determination module, configured to set a cycling reminder message mapping set, the set comprising message reminder key-value pairs, where the key represents the cycling event number that triggers the message reminder, and the value represents the message that needs to be transmitted to a first medium when the cycling event occurs, the first medium being a mobile phone connected to the smartwatch;

[0105] The riding events include a first fatigue event, a second physiological state abnormality event, and a third speed mismatch event; the keys in the riding reminder message mapping set include K1, K2, and K3, which correspond to the numbers of the first fatigue event, the second physiological state abnormality event, and the third speed mismatch event, respectively;

[0106] A data acquisition module, configured to acquire the user's historical riding data, first riding data, and second physiological data;

[0107] Furthermore, the historical riding data includes time series data of the user in various stages of past riding, including historical weather data, historical riding mileage, historical riding time, historical riding speed, historical riding slope, historical riding acceleration, historical riding road slope and historical riding road traffic volume; the first riding data includes time series data of the user during real-time riding, including weather data, riding speed, riding acceleration, riding time, riding distance, riding road slope and riding road traffic volume; the second physiological data includes heart rate, respiratory rate and body temperature.

[0108] The historical riding data is obtained from the user's past riding data, the historical weather data includes temperature, humidity and wind speed, and the weather data is obtained through the weather application software in the mobile phone connected to the smart watch. The historical riding mileage, riding mileage, historical riding time, riding time, historical riding speed, riding speed, historical riding acceleration and riding acceleration are obtained through the keep software used by the user during each riding process. The historical riding road slope, riding road slope, riding road traffic volume and historical riding road traffic volume are obtained through the map software in the mobile phone connected to the smart watch. The heart rate, respiratory rate and body temperature are obtained through the sensors in the smart watch.

[0109] a first judging module, configured to identify the historical riding data and the first riding data by building a user fatigue recognition model, obtain a first riding fatigue coefficient, and judge whether the first riding fatigue coefficient exceeds a preset fatigue coefficient threshold;

[0110] The user fatigue recognition model schematic diagram is as follows Figure 3 As shown;

[0111] The user fatigue recognition model includes a first input layer, a first preprocessing layer, a first feature extraction layer, a fatigue coefficient recognition layer and a first output layer;

[0112] The first input layer is used to input the historical riding data and the first riding data into a user fatigue recognition model;

[0113] The first preprocessing layer is used to preprocess the historical riding data and the first riding data, including denoising and abnormal data deletion, to obtain historical riding identification data and first riding identification data;

[0114] The first feature extraction layer is used to extract features from the historical riding identification data and the first riding identification data through an LSTM network to obtain a first fatigue feature, where the first fatigue feature includes tolerable riding intensity, tolerable riding load, real-time riding intensity, and real-time riding load;

[0115] The fatigue coefficient recognition layer is used to analyze the first fatigue feature to obtain a first riding fatigue coefficient;

[0116] The first output layer is used to output the first cycling fatigue coefficient.

[0117] Furthermore, the first riding fatigue coefficient is:

[0118]

[0119] Among them, CFC represents the first cycling fatigue coefficient; θ represents the first fatigue influence coefficient; WCI represents the real-time cycling intensity; CWCI represents the tolerable cycling intensity; WCL represents the real-time cycling load; CWCL represents the tolerable cycling load; γ represents the second fatigue influence coefficient; exp represents the exponential function with base e.

[0120] a second judgment module, configured to identify the second physiological data by building a user physiological state recognition model, obtain a second user physiological abnormality coefficient, and determine whether the second user physiological abnormality coefficient exceeds a preset physiological abnormality coefficient threshold;

[0121] The schematic diagram of the user physiological state recognition model is as follows Figure 4 As shown;

[0122] The user physiological state recognition model includes a second input layer, a second preprocessing layer, a second feature extraction layer, a physiological abnormality coefficient recognition layer and a second output layer;

[0123] The second input layer is used to input the second physiological data into the user physiological state recognition model;

[0124] The second preprocessing layer is used to perform data cleaning and denoising on the second physiological data to obtain second physiological recognition data respectively;

[0125] The second feature extraction layer is used to extract features from the second physiological recognition data through a convolutional neural network integrated with an attention mechanism; obtain second physiological state features, wherein the physiological state features include heart rate fluctuation amplitude, heart rate peak value, heart rate trough value, average heart rate, respiratory rate change rate, average respiratory rate, average body temperature and body temperature change amplitude;

[0126] The physiological abnormality coefficient identification layer obtains the second user's physiological abnormality coefficient by analyzing the second physiological state characteristics;

[0127] The second output layer is used to output the physiological abnormality coefficient of the second user;

[0128] The second user's physiological abnormality coefficient is:

[0129]

[0130] Wherein, PHC represents the physiological abnormality coefficient of the second user; ψ i represents the attention mechanism weight of the i-th second physiological state feature; PSC i represents the i-th second physiological state characteristic value.

[0131] a third judgment module, configured to determine whether the first cycling fatigue coefficient and the second user physiological abnormality coefficient are both less than corresponding preset thresholds; if so, deriving a first recommended speed based on the first cycling fatigue coefficient, the basic recommended speed, and the second user physiological abnormality coefficient; and determining whether the first recommended speed exceeds a preset speed threshold;

[0132] The execution module triggers a first fatigue event if the first cycling fatigue coefficient exceeds a preset fatigue coefficient threshold, and sends a first fatigue message reminder to the user's mobile phone; if it is determined that the second user physiological abnormality coefficient exceeds a preset physiological abnormality coefficient threshold, the smart watch triggers a second physiological state abnormality event, and sends a second physiological state abnormality message reminder to the user's mobile phone; if the error between the first recommended speed and the cycling speed exceeds a preset speed threshold, the smart watch triggers a third speed mismatch event, and sends a third speed mismatch message reminder to the user's mobile phone.

[0133] Furthermore, the first recommended speed is:

[0134] RSped=MeanSped*(1+k1*CFC+k2*PHC);

[0135] Among them, RSped represents the first recommended speed; MeanSped represents the basic recommended speed, which is used to represent the average speed of the user during the riding process; k1 represents the first speed adjustment coefficient; CFC represents the first riding fatigue coefficient; k2 represents the second speed adjustment coefficient; PHC represents the second user physiological abnormality coefficient.

[0136] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A message reminder method based on a smart watch, characterized in that: include: S1. Set a riding reminder message mapping set, the set includes each message reminder key-value pair, the key is used to represent the riding event number that triggers the message reminder, and the value is used to represent the message that needs to be transmitted to the first medium when the riding event occurs. The first medium is a mobile phone connected to the smart watch; S2. Obtain the user's historical riding data, the first riding data, and the second physiological data; S3. Build a user fatigue recognition model, identify the historical riding data and the first riding data, obtain a first riding fatigue coefficient, determine whether the first riding fatigue coefficient exceeds a preset fatigue coefficient threshold, and if so, the smartwatch triggers a first fatigue event and sends a first fatigue message reminder to the user's mobile phone; S4. Build a user physiological state recognition model, identify the second physiological data, obtain the second user physiological abnormality coefficient, determine whether the second user physiological abnormality coefficient exceeds the preset physiological abnormality coefficient threshold, and if so, the smart watch triggers a second physiological state abnormality event and sends a second physiological state abnormality message to the user's mobile phone to remind; S5. If both the first cycling fatigue coefficient and the second user physiological abnormality coefficient are less than corresponding preset thresholds, a first recommended speed is obtained based on the first cycling fatigue coefficient, the basic recommended speed, and the second user physiological abnormality coefficient. If the error between the first recommended speed and the cycling speed exceeds a preset speed threshold, the smartwatch triggers a third speed mismatch event and sends a third speed mismatch message reminder to the user's mobile phone.

2. The message reminder method based on a smart watch according to claim 1, characterized in that: The historical riding data includes time series data of the user in various stages of past riding, including historical weather data, historical riding mileage, historical riding time, historical riding speed, historical riding slope, historical riding acceleration, historical riding road slope and historical riding road traffic volume; the first riding data includes time series data of the user during real-time riding, including weather data, riding speed, riding acceleration, riding time, riding distance, riding road slope and riding road traffic volume; the second physiological data includes heart rate, respiratory rate and body temperature.

3. The message reminder method based on a smart watch according to claim 1, characterized in that: The user fatigue recognition model includes a first input layer, a first preprocessing layer, a first feature extraction layer, a fatigue coefficient recognition layer and a first output layer; The first input layer is used to input the historical riding data and the first riding data into a user fatigue recognition model; The first preprocessing layer is used to preprocess the historical riding data and the first riding data, including denoising and abnormal data deletion, to obtain historical riding identification data and first riding identification data; The first feature extraction layer is used to extract features from the historical riding identification data and the first riding identification data through an LSTM network to obtain a first fatigue feature, where the first fatigue feature includes tolerable riding intensity, tolerable riding load, real-time riding intensity, and real-time riding load; The fatigue coefficient recognition layer is used to analyze the first fatigue feature to obtain a first riding fatigue coefficient; The first output layer is used to output the first cycling fatigue coefficient.

4. The message reminder method based on a smart watch according to claim 3, characterized in that: The first riding fatigue coefficient is: Among them, CFC represents the first cycling fatigue coefficient; θ represents the first fatigue influence coefficient; WCI represents the real-time cycling intensity; CWCI represents the tolerable cycling intensity; WCL represents the real-time cycling load; CWCL represents the tolerable cycling load; γ represents the second fatigue influence coefficient; exp represents the exponential function with base e.

5. The message reminder method based on a smart watch according to claim 1, characterized in that: The user physiological state recognition model includes a second input layer, a second preprocessing layer, a second feature extraction layer, a physiological abnormality coefficient recognition layer and a second output layer; The second input layer is used to input the second physiological data into the user physiological state recognition model; The second preprocessing layer is used to perform data cleaning and denoising on the second physiological data; obtaining second physiological recognition data respectively; The second feature extraction layer is used to extract features from the second physiological recognition data through a convolutional neural network fused with an attention mechanism; Obtaining a second physiological state characteristic, wherein the physiological state characteristic includes heart rate fluctuation amplitude, heart rate peak value, heart rate trough value, average heart rate, respiratory rate change rate, average respiratory rate, average body temperature, and body temperature change amplitude; The physiological abnormality coefficient identification layer obtains the second user's physiological abnormality coefficient by analyzing the second physiological state characteristics; The second output layer is used to output the physiological abnormality coefficient of the second user.

6. The message reminder method based on a smart watch according to claim 5, characterized in that: The second user's physiological abnormality coefficient is: Wherein, PHC represents the physiological abnormality coefficient of the second user; ψ i represents the attention mechanism weight of the i-th second physiological state feature; PSC i represents the i-th second physiological state characteristic value.

7. The message reminder method based on a smart watch according to claim 1, characterized in that: The first recommended speed is: RSped=MeanSped*(1+k1*CFC+k2*PHC); Among them, RSped represents the first recommended speed; MeanSped represents the basic recommended speed, which is used to represent the average speed of the user during the riding process; k1 represents the first speed adjustment coefficient; CFC represents the first riding fatigue coefficient; k2 represents the second speed adjustment coefficient; PHC represents the second user physiological abnormality coefficient.

8. A message reminder system based on a smart watch, characterized in that: include: a message reminder mechanism determination module, configured to set a cycling reminder message mapping set, the set comprising message reminder key-value pairs, where the key represents the cycling event number that triggers the message reminder, and the value represents the message that needs to be transmitted to a first medium when the cycling event occurs, the first medium being a mobile phone connected to the smartwatch; A data acquisition module, configured to acquire the user's historical riding data, first riding data, and second physiological data; a first judging module, configured to identify the historical riding data and the first riding data by building a user fatigue recognition model, obtain a first riding fatigue coefficient, and judge whether the first riding fatigue coefficient exceeds a preset fatigue coefficient threshold; a second judgment module, configured to identify the second physiological data by building a user physiological state recognition model, obtain a second user physiological abnormality coefficient, and determine whether the second user physiological abnormality coefficient exceeds a preset physiological abnormality coefficient threshold; a third judging module, configured to judge whether the first cycling fatigue coefficient and the second user physiological abnormality coefficient are both less than corresponding preset thresholds; If yes, obtaining a first recommended speed according to the first cycling fatigue coefficient, the basic recommended speed, and the second user physiological abnormality coefficient; and determining whether the first recommended speed exceeds a preset speed threshold; An execution module, if the first cycling fatigue coefficient exceeds a preset fatigue coefficient threshold, the smartwatch triggers a first fatigue event and sends a first fatigue message reminder to the user's mobile phone; If it is determined that the second user's physiological abnormality coefficient exceeds the preset physiological abnormality coefficient threshold, the smartwatch triggers a second physiological state abnormality event and sends a second physiological state abnormality message reminder to the user's mobile phone; If the error between the first recommended speed and the riding speed exceeds a preset speed threshold, the smartwatch triggers a third speed mismatch event and sends a third speed mismatch message reminder to the user's mobile phone.

9. The message reminder system based on a smart watch according to claim 8, characterized in that: The first riding fatigue coefficient is: Wherein, CFC represents the first cycling fatigue coefficient; θ represents the first fatigue influence coefficient; WCI represents the real-time cycling intensity; CWCL represents the tolerable cycling intensity; WCL represents the real-time cycling load; CWCL represents the tolerable cycling load; γ represents the second fatigue influence coefficient; exp represents the exponential function with base e; The second user's physiological abnormality coefficient is: Wherein, PHC represents the physiological abnormality coefficient of the second user; ψ i represents the attention mechanism weight of the i-th second physiological state feature; PSC i represents the i-th second physiological state characteristic value.

10. The message reminder system based on a smart watch according to claim 8, characterized in that: The first recommended speed is: RSped=MeanSped*(1+k1*CFC+k2*PHC); Among them, RSped represents the first recommended speed; MeanSped represents the basic recommended speed, which is used to represent the average speed of the user during the riding process; k1 represents the first speed adjustment coefficient; CFC represents the first riding fatigue coefficient; k2 represents the second speed adjustment coefficient; PHC represents the second user physiological abnormality coefficient.

Images