Big data intelligent wearable device

Abstract

The invention discloses a big data intelligent wearable device. Belongs to the field of human body wearing equipment, and comprises a data acquisition unit, a central processing unit, an alarm unit and the like, and can monitor the physiological status of a wearer in real time, and give an alarm when an abnormality is detected to guarantee the personal safety of the wearer; the sensor collects physiological data and transmits the physiological data to the data acquisition unit through a wire or in a wireless mode, the data acquisition unit sends the data to the central processing unit, and the central processing unit analyzes the data and gives an alarm through the alarm unit when detecting abnormity. Meanwhile, the central processing unit can also send the data to the display unit; according to the invention, the physical condition of a wearer can be effectively detected in real time through the data of each sensor; and according to the detected data, the condition of the wearer can be known more timely and accurately, and rescue time can be won for the wearer when necessary.

Description

Technical Field

[0001] The present invention belongs to the field of human wearable devices, and relates to a big data intelligent wearable device; specifically, it relates to detecting the physiological conditions of the wearer to ensure the safety of the wearer. Background Art

[0002] With the continuous growth of the demand for living standards, intelligent wearable devices have emerged as the times require; the health monitoring functions of existing intelligent wearable devices mostly focus on common indicators such as heart rate and blood pressure, with a narrow monitoring range, and lack fall detection and location tracking functions; once a user is in danger, accurate rescue information cannot be provided in a timely manner; the accuracy and reliability of the health monitoring data of some devices are insufficient, which may mislead users' health decisions; with the popularization of intelligent wearable devices, data security and privacy protection have become important issues; in the prior art, there is a risk of leakage in the storage and transmission of user data, and some devices do not fully consider data encryption and privacy protection measures; although some enterprises have improved the battery life through technological innovation, there are still deficiencies in power consumption control of existing devices, especially in the process of data transmission and processing, the battery life is limited, affecting the user experience; the interaction methods of existing intelligent wearable devices are relatively single, relying on touch, buttons or simple voice commands, lacking an efficient and natural interaction experience. Summary of the Invention

[0003] In view of the above problems, the object of the present invention is to propose a big data intelligent wearable device, which can be used to analyze the physiological conditions of the human body, etc., provide useful information for the daily life of the wearer, detect various indicators of the human body physiology, prevent the wearer from being in danger, and provide a safer and more comfortable life experience for the wearer.

[0004] The technical solution of the present invention is: a big data intelligent wearable device described in the present invention includes a data acquisition unit, a central processing unit and an alarm unit that are connected to each other.

[0005] The data acquisition unit includes a number of integrated sensors for collecting physiological data and environmental data.

[0006] The central processing unit is used to process the data collected by the sensors and transmit the data to the alarm unit.

[0007] The alarm unit is used to give an alarm when abnormal data is detected.

[0008] Further, the integrated sensors include an intelligent safety helmet sensor, an intelligent glove sensor, an intelligent vest sensor and an environmental data sensor.

[0009] The intelligent safety helmet sensor monitors the impact on the head and environmental conditions; it includes a heart rate sensor, a temperature sensor and an accelerometer.

[0010] The intelligent glove sensor monitors hand movements and grip strength, including a pulse sensor and a grip strength sensor;

[0011] The intelligent vest sensor monitors chest physiological parameters and body postures; it includes a heart rate sensor, a body temperature sensor, and a respiration sensor;

[0012] The environmental data sensor includes a temperature sensor, a humidity sensor, and a light sensor.

[0013] Data collection by the data sensors can collect the body data and environmental data of the human body at that time, and classify them more accurately; comparison with the baseline, classify them according to different types, diagnose and alarm according to different situations; analyze individual physiological baselines to provide personalized health monitoring.

[0014] Furthermore, the data collected by the integrated sensor includes user basic information, behavior data information, and surrounding environmental data information;

[0015] Among them, the user basic information includes data information such as blood pressure and body mass index;

[0016] The behavior data information includes detection information such as heart rate and blood pressure;

[0017] For the data collected by the integrated sensor, using the heart rate data (HR), temperature data (T), acceleration data (Acc), and pulse data (P) as inputs, a CC-RLS objective function is constructed;

[0018] By applying the CC-RLS algorithm, high-precision estimation of the detected results is achieved;

[0019] Among them, the construction of the CC-RLS objective function is specifically:

[0020]

[0021] b(n) = b(n - 1) + k(n)e(n)

[0022]

[0023] In the formula, cov(Y(t0,a(t)) represents the covariance between the signals Y(t) and a(t), D(Y(t)) represents the difference of the signal Y(t), D(a(t)) represents the variance of the signal a(t), C 1 C 2 C 3Denote the values of heart rate, temperature, pulse, and acceleration. n represents the superposition coefficient, b(n) represents the filter coefficient of w in the nth iteration, k(n) represents the function gain, e(n) represents the output error of the filter k(n) in the nth iteration, f(n) represents the input vector, λ represents the forgetting factor, and R -1 (n - 1) represents the inverse of the covariance rectangle;

[0024] The construction method of this objective function combines the characteristics of heart rate, temperature, acceleration, and pulse data, aiming to extract useful physiological features by eliminating noise and artifacts to improve the prediction accuracy of the model.

[0025] Furthermore, the central processing unit receives sensor data from the data acquisition unit, which is collected from the behavioral data information and environmental data information as unlabeled sample data, that is: the health model management system constructs an unsupervised learning algorithm, which can help record and analyze data after recording, generate healthy living or exercise suggestions, and actively give early warnings and alarms when abnormalities are detected;

[0026] The human action recognition technology of wearable devices based on deep learning realizes functions such as gesture recognition and motion state recognition, providing users with a more intelligent experience; learn the learning mode and structure from the behavioral data information as unlabeled sample data, and use the Feedforward Neural Network (FNN) algorithm, which is divided into five parts (1. input layer, 2. hidden layer, 3. activation function, 4. output layer, 5. output layer activation function); the specific is as follows:

[0027] X = [x 1 , x 2 ,..., x n (2)

[0028]

[0029] a hj = σ(z hj ) (4)

[0030]

[0031] a ok = σ(z ok ) (6)

[0032] In the formula, X represents the sensor input values (heart rate, temperature, pulse, acceleration), w jk represents the connection between the jth hidden neuron and the kth output neuron, w ik represents the weight connecting the ith input neuron and the jth hidden neuron, b kDenote the bias term of the k-th output neuron for the j-th neuron of the h-th hidden function. The output formula is as above, and for each neuron \(a\) in the output layer \(a\). ok The output of each neuron \(a\) in hj is also the result after processing its input through the activation function. \(j\).

[0033] By using the Sigmoid activation function to optimize the activation function part of the feedforward neural network, the nonlinearity is increased, the accuracy of data analysis and anomaly detection is improved, and the performance of the model is affected. The purpose is to increase the nonlinearity, as shown in the following formula:

[0034]

[0035] \(z = w\) 1 \(x\) 1 +\(w\) 2 \(x\) 2 +…+\(w\) n \(x\) n +\(b\) (8)

[0037]

[0038] In the formula, \(e\) represents the base of the natural logarithm, approximately equal to 2.71828, \(z\) represents the linear combination of input features, \(f(z)\) represents that the sample belongs to the positive class, \(w\) n represents the weight, \(x\) n represents the input feature, \(b\) represents the bias, and \(P\) represents the probability.

[0039] Furthermore, using the Softmax optimization algorithm allows the model to output a probability distribution instead of a single class prediction, improving the prediction accuracy of the model. By applying the Softmax optimization algorithm to the output layer of the feedforward neural network algorithm in the health model management system, the prediction output method of the model is optimized. The probability distribution output provides information to judge the physiological state or environmental condition of the wearer, providing support for generating healthy living or exercise suggestions and anomaly warnings. The specific formula is as follows:

[0040]

[0041] In the formula, \(k\) represents the number of classes, \(s\) i represents the probability that the i-th element in the input vector belongs to the corresponding class, and all \(i = 1, 2,..., K\). represents the exponential function, and the denominator \(z\) i is the sum of all , and \(s\) i ensures that the sum of all is 1.

[0042] Furthermore, the central processing unit has a data transmission function, and in the smart wearable device, it transmits the user's basic information to the data analysis system. The transmission method is based on the Shannon channel capacity formula and is transmitted to the AI intelligent analysis system. The specific process is as follows:

[0043]

[0044] In the formula, C represents the channel capacity, that is, the maximum data transmission rate (unit: bits per second), W represents the channel bandwidth (unit: Hertz), P represents the average signal power, and N 0 represents the power density of the noise, S(f) represents the power spectrum of the signal, N(f) represents the noise power spectrum, f represents the frequency, and B represents the channel bandwidth.

[0045] Furthermore, the optimization improves the transmission speed through the following two methods; namely:

[0046] Optimizing signal processing and resource allocation: By optimizing coding, signal processing, and resource allocation to increase the channel capacity and reduce feedback information, the energy efficiency of the communication system can be effectively improved. The formula is as follows:

[0047]

[0048] In the formula, C ergodic represents the long-term average capacity of the channel, that is, the average capacity over all possible channel states, B represents the bandwidth of each sub-channel, and log 2 (1 + γ) represents the logarithmic function with base 2;

[0049] Using physical dimensions for capacity expansion: including time (modulation symbol rate and pulse shape), quadrant (using the real and imaginary parts of light waves), polarization (two orthogonal polarization states), frequency (WDM channel number and spectral band), and space (parallel transmission over more than one spatial path); as follows:

[0050]

[0051] In the formula, C ergodic represents the long-term average capacity of the channel, that is, the average capacity over all possible channel states, B represents the bandwidth of each sub-channel, and P j represents the power allocated to the j-th sub-channel, and N 0 is the power density of the noise, ||H j || 2 is the square of the channel gain of the j-th sub-channel.

[0052] Furthermore, the alarm unit is specifically: It performs data judgment and classification by comparing with the personal baseline of the input system. The specific process is as follows:

[0053]

[0054] wherein, max(x i ), min(x i ) represent the maximum or minimum value in the dataset, represents the sample mean, S represents the sample standard deviation; the judgment is made according to the human baseline.

[0055] The beneficial effects of the present invention are as follows: through the CC-RLS objective function, physiological data and environmental data of the present invention; then through the FNN algorithm, the data activation is converted into data to be processed; transmitted to the AI intelligent analysis system by using the Shannon channel capacity formula; finally, the AI intelligent analysis system makes a comparative analysis according to the personal baseline; the application of big data intelligent wearable devices is becoming more and more extensive; it monitors the physiological state of workers, such as heart rate, body temperature and fatigue degree, etc. through integrated sensors; these devices can collect data in real time and analyze it through big data technology to ensure the safety and health of workers. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 is a schematic structural diagram of the intelligent wearable device of the present invention;

[0057] Figure 2 is a flowchart of the operation of the intelligent wearable device of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0058] The following further elaborates on the specific technical solutions of the present invention in conjunction with specific examples.

[0059] As shown in the figure, a big data intelligent wearable device according to the present invention, the data acquisition unit includes a plurality of integrated sensors, which can collect human physiological data and environmental data; a central processing unit, which processes the data collected by the sensors and transmits the data to the alarm unit, and the alarm unit alarms when detecting abnormal data;

[0060] The integrated sensors include an intelligent safety helmet, intelligent gloves, an intelligent vest and environmental data sensors; the intelligent safety helmet sensor is used to detect the human head temperature, acceleration, and detect the human heart rate;

[0061] The intelligent glove sensor is used to detect the hand grip strength and pulse beating frequency;

[0062] The intelligent vest is used to detect the human heart rate, breathing rate and body temperature;

[0063] The environmental data sensors include a temperature sensor, a humidity sensor and a light sensor.

[0064] Data collection by sensors can collect the body data and environmental data of the human body at that time for more accurate classification; baseline comparison, classification according to different types, diagnosis and alarm according to different situations; analysis of individual physiological baselines to provide personalized health monitoring.

[0065] Specifically, there are multiple sensors in the big data intelligent wearable devices, which respectively detect the physiological conditions of the human body; in order to detect the physiological conditions of the wearer more efficiently, an AI intelligent discrimination system is introduced to quickly evaluate the abnormal level of the physiological conditions of the human body, and corresponding treatment measures will be taken according to different abnormal levels;

[0066] Among them, in the intelligent safety helmet, the temperature sensor can receive the ambient temperature and the head temperature, and the acceleration sensor can calculate the acceleration of the human head, so as to take protective measures for the head in advance. The heart rate sensor can record the arterial beating frequency of the human head, so as to calculate the physiological conditions of the human body; in the intelligent glove, the grip force sensor can detect the grip force of the hand and calculate the degree of damage to the hand by the external environment. The pulse sensor can detect the pulse beating frequency of the human hand, and the combination of the two can better output data; in the intelligent vest, the heart rate sensor can detect the heart beating frequency, the breathing sensor can detect the breathing frequency, and the body temperature sensor can detect the body surface temperature. The combination of the three can reduce the error generated by calculation, so as to improve the efficiency of AI judgment; the various sensors in the intelligent wearable devices cooperate with each other, which can reduce the mistakes in judgment and evaluate the abnormal level more efficiently.

[0067] Furthermore, the integrated sensors in the data acquisition unit are responsible for collecting various types of data, including user basic information, behavior data information, and surrounding environment data information; among them, the user basic information covers data such as blood pressure and body mass index; the behavior data information includes the detection information of heart rate and blood pressure.

[0068] Furthermore, taking the heart rate data, temperature data, acceleration data, and pulse data as inputs, a CC-RLS objective function is constructed for the data collected by the integrated sensors; the constructed CC-RLS objective function is specifically as follows:

[0069]

[0070] b(n) = b(n - 1) + k(n)e(n)

[0071]

[0072] In the formula, cov(Y(t0,a(t)) represents the covariance between the signal Y(t) and a(t), D(Y(t)) represents the difference of the signal Y(t), D(a(t)) represents the variance of the signal a(t), C 1 C2 C 3 represents the values of heart rate, temperature, pulse, and acceleration. n represents the superposition coefficient, b(n) represents the filter coefficient of w in the nth iteration, k(n) represents the function gain, e(n) represents the output error of the filter k(n) in the nth iteration, f(n) represents the input vector, λ represents the forgetting factor, and R -1 (n - 1) represents the inverse of the covariance rectangle.

[0073] The construction method of this objective function combines the characteristics of heart rate, temperature, acceleration, and pulse data, aiming to extract useful physiological features by eliminating noise and artifacts to improve the prediction accuracy of the model.

[0074] Furthermore, the original data is processed based on the FNN algorithm to obtain processed data; the processed data is determined based on the mathematical algorithm of Grubbs' Test and compared with the personal benchmark to determine whether it meets the indicators of the said objective.

[0075] The present invention combines sensors and AI intelligent detection devices to judge the physiological condition of the human body by comparing with the personal baseline;

[0076] Build an unsupervised learning algorithm for the health model management system. The algorithm can help record and analyze data, generate healthy living or exercise suggestions, and actively give early warnings and alarms when abnormalities are detected; based on the deep learning wearable device human action recognition technology, realize functions such as gesture recognition and motion state recognition, providing a more intelligent experience for users; learn the mode and structure from the behavioral data information as unlabeled sample data, and use the Feedforward Neural Network (FNN) algorithm, which is divided into five parts (input layer, hidden layer, activation function, output layer, output layer activation function):

[0077] X = [x 1 , x 2 ,..., x n

[0078]

[0079] a hj = σ(z hj )

[0080]

[0081] a ok = σ(z ok )

[0082] In the formula, X is the sensor input value (heart rate, temperature, pulse, acceleration), w jk ​is the connection between the j-th hidden neuron and the k-th output neuron, w ik is the weight connecting the i-th input neuron and the j-th hidden neuron, b k is the bias term of the k-th output neuron. For the j-th neuron of the h-th hidden function, the output formula is as above. For each neuron a ok in the output layer a hj the output is also the result of processing its input through the activation function.

[0083] Specifically, by using the Sigmoid activation function to optimize the activation function part of the feedforward neural network, the model performance is affected, the nonlinearity is increased, so as to better process the complex data relationship of sensor input and improve the accuracy of data analysis and anomaly detection:

[0084]

[0085] z = w 1 x 1 + w 2 x 2 + … + w n x n + b

[0086]

[0087] In the formula, e represents the base of the natural logarithm, approximately equal to 2.71828, z represents the linear combination of input features, f(z) represents that the sample belongs to the positive class, w n represents the weight, x n represents the input feature, b represents the bias, and P represents the probability.

[0088] Furthermore, using the Softmax optimization algorithm allows the model to output a probability distribution instead of a single class prediction, improving the prediction accuracy of the model: By applying the Softmax optimization algorithm to the output layer of the feedforward neural network algorithm in the health model management system, the prediction output method of the model is optimized; The probability distribution output provides information to judge the wearer's physiological state or environmental conditions, providing support for generating healthy living or exercise suggestions and anomaly warnings. The specific formula is as follows:

[0089]

[0090] In the formula, k represents the number of classes, s i represents the probability that the i-th element in the input vector belongs to the corresponding class, and all i = 1, 2,..., K, represents the exponential function, and the denominator z i is all

[0091] si The sum is ensured so that all sums are 1.

[0092] The optimal solution is calculated as follows:

[0093]

[0094] Where a is the output of the neuron, f is the activation function, wi is the weight, xi is the input, b is the bias term, and n is the number of inputs.

[0095] Furthermore, the central processing unit has a data transmission function, and in the smart wearable device, it transmits the user's basic information to the data analysis system. The transmission method is based on the Shannon channel capacity formula and transmitted to the AI intelligent analysis system. The specific process is as follows:

[0096]

[0097] Where C represents the channel capacity, that is, the maximum data transmission rate (unit: bits per second), W represents the channel bandwidth (unit: hertz), P represents the average signal power, N 0 represents the power density of the noise, S(f) represents the power spectrum of the signal, N(f) represents the noise power spectrum, f represents the frequency, and B represents the bandwidth of the channel.

[0098] Furthermore, the optimization improves the transmission speed in the following two ways; that is:

[0099] Optimize signal processing and resource allocation: By optimizing coding, signal processing, and resource allocation, the channel capacity can be increased and the feedback information can be reduced, which can effectively improve the energy efficiency of the communication system. The formula is as follows:

[0100]

[0101] Where C ergodic represents the long-term average capacity of the channel, that is, the average capacity over all possible channel states, B represents the bandwidth of each sub-channel, log 2 (1 + γ) represents the logarithmic function with base 2;

[0102] Use physical dimensions for capacity expansion: including time (modulation symbol rate and pulse shape), quadrant (using the real and imaginary parts of the optical wave), polarization (two orthogonal polarization states), frequency (WDM channel number and spectral band), and space (parallel transmission over more than one spatial path); as follows:

[0103]

[0104] Where C ergodic represents the long-term average capacity of the channel, that is, the average capacity over all possible channel states, B represents the bandwidth of each sub-channel, Pj represents the power allocated to the j-th subchannel, N 0 is the power density of the noise, ||H j || 2 is the square of the channel gain of the j-th subchannel.

[0105] Furthermore, the alarm unit specifically: performs data judgment and classification by comparing with the personal baseline of the input system; then, establishes a network model using the optimized FNN algorithm, and the network model includes an input layer, a judgment layer, and an output layer; wherein, the input layer receives physiological data and environmental data, the judgment layer divides the range according to the personal baseline, divides the abnormal level within each range, and calculates the human body state at each moment; the output layer executes the corresponding abnormal level; determines whether it is higher than the personal baseline, using the following formula:

[0106]

[0107] In the formula, max(x i ), min(x i ) represents the maximum or minimum value in the dataset, represents the sample mean, S represents the sample standard deviation; and makes a judgment according to the human body baseline.

[0108] The present invention uses a variety of sensors to collect data, including heart rate, temperature, pressure, and acceleration data; the present invention can detect various physiological indicators of the wearer and ensure the life safety of the wearer in real time; the application fields of intelligent wearable devices are extensive, such as tracking sports, health monitoring, and chronic management treatment; intelligent wearable devices can accurately classify the events that occur; take timely actions through radio technology; can achieve shortening the rescue time; the use of intelligent wearable devices provides a safer daily life experience for the wearer through technologies such as human physiological detection, intelligent classification, and emergency alarm. In summary, the intelligent wearable device uses a variety of technologies such as sensors, artificial intelligence, and wireless communication technology to ensure the safety of the wearer during daily sports and life; not only improves the level of life safety, but also reduces the risk of death due to inability to obtain timely medical treatment in daily life, and guarantees the safety of the wearer.

Claims

1. A big data smart wearable device, characterized in that: It includes interconnected data acquisition unit, central processing unit and alarm unit. The data acquisition unit includes a plurality of integrated sensors for collecting physiological data and environmental data; The central processing unit is used to process the data collected by the sensor and transmit the data to the alarm unit; The alarm unit is used to alarm when data abnormality is detected.

2. A big data smart wearable device according to claim 1, characterized in that: The integrated sensors include smart helmet sensors, smart glove sensors, smart vest sensors and environmental data sensors; The smart helmet sensor includes a heart rate sensor, a temperature sensor and an accelerometer; The smart glove sensor includes a pulse sensor and a grip force sensor; The smart vest sensor includes a heart rate sensor, a body temperature sensor and a breathing sensor; The environmental data sensors include a temperature sensor, a humidity sensor and a light sensor.

3. A big data smart wearable device according to claim 1, characterized in that: The data collected by the integrated sensor includes basic user information, behavioral data information and surrounding environment data information; Among them, user basic information includes blood pressure and body mass index data; Behavioral data information includes heart rate and blood pressure detection information.

4. A big data smart wearable device according to claim 3, characterized in that: For the data collected by the integrated sensor, the heart rate data, temperature data, acceleration data and pulse data are used as input to construct a CC-RLS objective function; the constructed CC-RLS objective function is specifically as follows: b(n)=b(n-1)+k(n)e(n) In the formula, cov(Y(t0,a(t)) represents the covariance between signals Y(t) and a(t), D(Y(t)) represents the difference of signal Y(t), D(a(t)) represents the variance of signal a(t), C1C2C3 represents the values ​​of heart rate, temperature, pulse, and acceleration, n represents the superposition coefficient, b(n) represents the filter coefficient of the nth iteration w, k(n) represents the function gain, e(n) represents the filter output error of the nth iteration, f(n) represents the input vector, λ represents the forgetting factor, and R -1 (n-1) represents the inverse of the covariance rectangle.

5. The big data smart wearable device according to claim 1, characterized in that: The central processing unit receives sensor data from the data collection unit, collects behavioral data information and environmental data information as unlabeled sample data, and divides them into five parts using a feedforward neural network algorithm; specifically, as follows: X=[x1,x2,...,x n ] (2) a hj =σ(z hj ) (4) a ok =σ(z ok ) (6) Where X represents the sensor input value, i.e. heart rate, temperature, pulse, acceleration, w jk represents the connection between the jth hidden neuron and the kth output neuron, w ik represents the weight connecting the i-th input neuron and the j-th hidden neuron, b k The bias term of the k-th output neuron is expressed as: For the j-th neuron of the h-th hidden function, the output formula is as above, and the output layer a ok Each neuron a hj The output of is also the result of its input being processed by the activation function; j optimizes the activation function part of the feedforward neural network by using the Sigmoid activation function, increases nonlinearity, and improves the accuracy of data analysis and anomaly detection, as shown in the following formula: z=w1x1+w2x2+…+w n x n +b (8) In the formula, e represents the base of the natural logarithm, z represents the linear combination of the input features, f(z) represents that the sample belongs to the positive class, and w n represents the weight, x n represents input features, b represents bias, and P represents probability.

6. A big data intelligent wearable device according to claim 5, characterized in that: The feedforward neural network algorithm includes five parts: input layer, hidden layer, activation function, output layer and output layer activation function.

7. The big data smart wearable device according to claim 5, characterized in that: By applying the Softmax optimization algorithm to the output layer of the feedforward neural network algorithm of the health model management system, the prediction output mode of the model is optimized; the probability distribution output provides information to judge the physiological state or environmental conditions of the wearer, and provide support for generating healthy life or exercise suggestions and abnormal warnings, which is as follows: In the formula, k represents the number of categories, s i represents the probability that the i-th element in the input vector belongs to the corresponding category, all i=1,2,...,K, represents the exponential function, the denominator z i Yes all The sum of s i Make sure all sums are 1.

8. The big data smart wearable device according to claim 1, characterized in that: The central processing unit has a data transmission function, and transmits the basic information of the user to the data analysis system in the smart wearable device. The transmission method is based on the Shannon channel capacity formula and is transmitted to the AI ​​intelligent analysis system. The specific process is as follows: Where C represents the channel capacity, i.e., the maximum data transmission rate, W represents the channel bandwidth, P represents the average signal power, N0 represents the power density of the noise, S(f) represents the power spectrum of the signal, N(f) represents the noise power spectrum, f represents the frequency, and B represents the bandwidth of the channel.

9. The big data intelligent wearable device according to claim 8, characterized in that: Optimization increases transfer speed in two ways; namely: Optimize signal processing and resource allocation: Improve channel capacity and the energy efficiency of the communication system by optimizing coding, signal processing, and resource allocation. The formula is as follows: In the formula, C ergodic represents the long-term average capacity of the channel, that is, the average capacity over all possible channel states, B represents the bandwidth of each subchannel, and log2(1+γ) represents the logarithmic function with base 2; Leverage physical dimensions for capacity expansion: including time, quadrant, polarization, frequency, and space; In the formula, C ergodic represents the long-term average capacity of the channel, that is, the average capacity over all possible channel states, B represents the bandwidth of each subchannel, and P j represents the power allocated to the jth subchannel, N0 is the power density of the noise, ||H j || 2 is the square of the channel gain of the jth subchannel.

10. The big data smart wearable device according to claim 1, characterized in that: The alarm unit uses a personal baseline comparison based on the input system to perform data judgment and classification, and the specific process is as follows: In the formula, max(x i ), min(x i ) represents the maximum or minimum value in the data set. represents the sample mean, and S represents the sample standard deviation.

Images