JP6554426B2 - Exercise intensity estimation device, exercise intensity estimation method, program - Google Patents

Description

  The present invention relates to, for example, an exercise intensity estimation apparatus for estimating exercise intensity of a person, an exercise intensity estimation method, and a control program thereof.

  Workers engaged in construction work and manufacturing work their bodies vigorously on a daily basis.

  In order to realize such labor management as having these workers get rest according to the degree of exercise and instruct them not to consume excessive calories, exercise intensity during work (for each worker) ( It is required to accurately measure the intensity of exercise.

  Conventionally, in order to measure exercise intensity, an acceleration sensor is worn on a waist or the like, and acceleration information measured there has been used (for example, see Non-Patent Document 1).

  However, in the method of estimating exercise intensity based on such acceleration information, the activity amount is underestimated when performing an operation with a small change in acceleration (such as a plastering operation to move the upper body in a crouched state or an operation to support a heavy object). I think that. Further, it is difficult to determine the difference between the walking motion and the transporting motion (walking while supporting a heavy object) from the acceleration information, and it is considered that the estimation error becomes large.

  As described above, it is considered that the motion that cannot be accurately estimated by the estimation method from the acceleration information is frequently performed in construction work, etc., and the motion intensity estimation for the worker cannot cope with such motion. Desired.

  On the other hand, a method of estimating exercise intensity based on heart rate information has been considered (see, for example, Non-Patent Document 2).

K. Ohkawara, Y. Oshima, Y. Hikihara, K. Ishikawa-Takata, I. Tabata, and S. Tanaka, “Real-time estimation of daily physical activity intensity by a degree of accelerometry and a gravity-removal classification algorithm,” Br. J. Nutr., Vol. 105, no. 11, pp. 1681-1691, Jun. 2011. Internet <URL: http://journals.cambridge.org/abstract_S0007114510005441> M. Li, K.-C. Kwak, and Y.-T. Kim, “Intelligent predictor of energy opportunity with the use of patch-type sensor module.” Sensors (Basel)., Vol. 12, no. 11 , pp. 14382-96, Jan. 2012.

  However, many of the conventional estimation methods aim at estimation with a single motion such as running, and do not correspond to multiple motions. In construction work, workers alone perform various operations, and it is difficult to apply the conventional estimation method directly to worker exercise intensity estimation.

  The present invention has been made in view of such a problem, and an exercise intensity estimation device and exercise intensity capable of estimating exercise intensity with high accuracy in a situation where a plurality of operations including an operation with small movement occur. An object is to provide an estimation method and a control program thereof.

A first exercise intensity estimation device according to the present invention comprises data storage means for storing data in which exercise intensity and heart rate are associated with each of a plurality of types of motions of a subject, and a heartbeat of an estimation target person. The difference between the heart rate input means for inputting the number and the heart rate of the person to be estimated which is input by the heart rate input means among the heart rates corresponding to the plurality of types of actions stored in the data storage means a data extracting means for extracting by selecting the exercise intensity associated with the plurality of heart rate and the plurality of heart rate from the smaller, plurality extracted from the data storage means by said data extracting means of the heart rate as a parameter representing the relationship between them of when and the exercise intensity and heart rate was left and the right side of the equation in each set of the exercise intensity, the numerical difference between the left and right side of the equation is the most Become smaller A regression analysis section for determining the parameters and regression analysis, based on the parameters determined by the regression analysis means calculates the exercise intensity corresponding to the heart rate estimation subject input by the heart rate input means exercise intensity And a calculating means.

A second exercise intensity estimation apparatus according to the present invention comprises data storage means for storing data in which exercise intensity, heart rate and acceleration are associated with each of a plurality of types of motions of a subject, and an estimation target person Among the accelerations according to each of a plurality of types of operations stored in the data storage means, the acceleration input means for inputting the heart rate input means, the acceleration input means for inputting the acceleration of the person to be estimated Data extraction means for selecting and extracting a plurality of accelerations and a heart rate and exercise intensity associated with the plurality of accelerations from the smaller difference with the acceleration of the estimation target person input by the method; Table both relationships when with the exercise intensity and heart rate and the left and right side of the equation in each set of a plurality of heart rate and exercise intensity extracted from the data storage means by the extraction means Of be parameters, and the regression analysis means for determining by regression analysis parameters numerical difference between the left and right side of the equation is minimized, based on the parameters determined by the regression analysis means, an input by the heart rate input means And exercise intensity calculation means for calculating exercise intensity corresponding to the estimated subject's heart rate.

  According to the present invention, it is possible to estimate exercise intensity with high accuracy in a situation where a plurality of motions including a motion with small motion occur.

BRIEF DESCRIPTION OF THE DRAWINGS The block diagram which shows the system configuration | structure of the exercise intensity estimation apparatus 10 which concerns on embodiment of this invention. The figure which shows the content of the data memorize | stored in exercise intensity: heart rate: acceleration DB12c of the said exercise intensity estimation apparatus 10. FIG. The simplification figure which shows the basic procedure of the method of estimating exercise intensity by the said exercise intensity estimation apparatus 10. FIG. The figure which shows the procedure of the exercise intensity estimation method (1) by the said exercise intensity estimation apparatus 10. FIG. The figure which shows the procedure of the exercise intensity estimation method (2) by the said exercise intensity estimation apparatus 10. FIG. 3 is a flowchart showing exercise intensity estimation processing (1) (2) according to the exercise intensity estimation program 12b of the exercise intensity estimation device 10. Each estimated value obtained by estimating the exercise intensity of the person to be estimated in the process of actually performing the work according to the estimation processes (1) and (2) of the exercise intensity estimation apparatus 10 and the actual measurement value measured using an oxygen intake meter The figure which compares and shows (true value). The figure which shows the average of the absolute error of each estimated value which estimated the exercise | movement intensity | strength of the estimation object person in the process actually performed according to the said estimation process (1) (2), and a measured value (true value). The figure which shows the average of the correlation coefficient of each estimated value which estimated the exercise | movement intensity | strength of the estimation object person in the process which actually performed according to the said estimation process (1) (2), and an actual value (true value).

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a block diagram showing a system configuration of an exercise intensity estimation apparatus 10 according to an embodiment of the present invention.

  The exercise intensity estimation apparatus 10 is implemented using an information terminal device such as a personal computer or a tablet terminal, and includes a CPU 11 which is a computer.

  The CPU 11 executes the device control program 12a and the exercise intensity estimation program 12b stored in the storage unit 12 using the RAM 16 as a working memory, and controls the operation of each part of the circuit. The exercise intensity estimation program 12b is read from the external recording medium 14 such as a ROM card into the storage unit 12 via the recording medium reading unit 13, or from a Web server on the Internet (in this case, a program server). It may be read into the storage unit 12 through the communication unit 15.

  The device control program 12a and the exercise intensity estimation program 12b stored in the storage unit 12 are input signals corresponding to user operations from the operation unit 19 such as a keyboard and a touch panel, or the heart rate measurement unit 17 and triaxial acceleration. It is activated in response to each measurement signal from the measurement unit 18 or a communication signal with each Web server (not shown) on the Internet connected via the communication unit 15.

  The measurement functions of the heart rate measurement unit 17 and the triaxial acceleration measurement unit 18 are provided by, for example, an electrocardiographic sensor, and are realized by attaching the electrocardiographic sensor to an exercise intensity estimation target person. The estimation target person's heart rate HR ′ and triaxial acceleration (AccX ′, AccY ′, AccZ ′) measured by the electrocardiographic sensor may be input via the communication unit 15 by radio signals.

  Thus, in addition to the storage unit 12, the recording medium reading unit 13, the communication unit 15, the RAM 16, the heart rate measurement unit 17, the triaxial acceleration measurement unit 18, and the operation unit 19, the CPU 11 is connected to the CPU 11. A display unit 20 with a touch panel is connected.

  The device control program 12a stored in the storage unit 12 includes a system program that controls the overall operation of the exercise intensity estimation device 10, each Web server on the Internet via the communication unit 15, and an external device (not shown). A communication program for data communication is stored.

  As the exercise intensity estimation program 12b, the estimation target person's heart rate HR 'measured by the heart rate measurement unit 17 or the estimation target measured by the heart rate HR' and the triaxial acceleration measurement unit 18 is used. Based on the three-axis acceleration (AccX ′, AccY ′, AccZ ′) of the person, with reference to the exercise intensity: heart rate: acceleration DB (database) 12c described later, for estimating the exercise intensity of the estimation target person The program is memorized.

  FIG. 2 is a diagram showing the contents of data stored in the exercise intensity: heart rate: acceleration DB 12 c of the exercise intensity estimation device 10.

  This exercise intensity: heart rate: acceleration DB12c, subject subject exercise intensity PA, heart rate HR and 3-axis acceleration (AccX, AccY, AccZ, AccZ Are stored in association with each of the plurality of types of motions (exercises).

  The heart rate data memory 16a storing the heart rate HR 'of the person to be estimated measured and input by the heart rate measuring unit 17 and the 3-axis acceleration measuring unit 18 measure and input to the RAM 16 The acceleration data memory 16b and the like in which the three-axis accelerations (AccX ', AccY', AccZ ') of the person to be estimated are stored are secured.

  In the exercise intensity estimation apparatus 10 configured as described above, the CPU 11 controls the operation of each part of the circuit in accordance with the instructions described in the apparatus control program 12a and the exercise intensity estimation program 12b. By operating in cooperation, the exercise intensity estimation function described below is realized.

(1) Outline of Estimation Method FIG. 3 is a simplified diagram showing a basic procedure of a method for estimating exercise intensity by the exercise intensity estimation device 10.

  The method of estimating the exercise intensity of the estimation target person according to the present embodiment is a frame of supervised learning in which exercise intensity PA (Physical Activity, unit METs) is a target variable and heart rate HR (Heart Rate, unit bpm) is an explanatory variable. According to the work.

  The exercise intensity: heart rate: acceleration DB 12 c prepared in advance, exercise intensity PA and heart rate HR when subject performs a plurality of types of actions, triaxial acceleration values AccX, AccY, AccZ (x, y, z Three types of directions, units G) are stored.

From the exercise intensity PA and the heart rate HR stored in the acceleration DB 12c, parameters β ₀ and β ₁ representing the relationship between the two are determined by regression analysis based on the following equation 1.

Here, with respect to the heart rate HR ′ of the estimation target person newly measured, the estimated value PAe of the exercise intensity is calculated by the following equation 2 using the parameters β ₀ and β ₁ obtained based on the equation 1. expressed.

  As a result, it is possible to obtain an output PAe of exercise intensity from the heart rate HR ′ of the person to be estimated that has been measured and input by the heart rate measurement unit 17.

(1-1) Key Points of Estimation Method In the estimation method according to the present embodiment, the parameter β ₀ , which represents the relationship between the exercise intensity PA and the heart rate HR stored in advance in the exercise intensity: heart rate: acceleration DB 12c. Calculate β ₁ The exercise intensity: heart rate: acceleration DB 12 c stores exercise intensity PA and heart rate HR of various exercise states performed by the subject. The estimation target person is selected by selecting the data set of the state close to the current state of the estimation target person from among the data sets (sets) of the exercise intensity PA and the heart rate HR of various exercise states, and performing the regression analysis The exercise intensity PAe is analyzed with high accuracy.

(1-2) Selection Method of Learning Data There are two types of data selection methods.

  Estimation method (1) Select data with reference to heart rate HR.

  Estimation method (2) Select data with reference to 3-axis acceleration (AccX, AccY, AccZ).

(In the case of estimation method (1))
FIG. 4 is a diagram showing a procedure of the exercise intensity estimation method (1) by the exercise intensity estimation device 10.

  The input is only the estimation target person's heart rate HR '.

  The heart rate HR in the DB 12c having a small difference from the input heart rate HR 'is considered to be similar in data property, and N HRs having a small difference from the HR' are selected in order from the DB 12c (N is an arbitrary constant). The PA corresponding to the HR selected at the same time is also called from the DB 12c.

The equation 1 including N sets of parameters β ₀ and β ₁ respectively corresponding to N sets of HR and PA is calculated, and the N sets of parameters β ₀ and β _{1 correspond} to the left side and the right side in the equation 1, respectively. A set of parameters β ₀ and β ₁ that minimizes the numerical difference is obtained by the least square method or the like. That is, the parameters β ₀ and β ₁ are calculated from the combination of N HR and PA by the regression analysis according to the equation 1. These values may be calculated in advance using the values of the DB 12c and recorded in the DB 12c. Then, on the basis of the set of parameters β ₀ and β _{1 obtained} as described above and the input heart rate HR ′ of the estimation subject, the exercise amount PAe of the estimation subject is calculated and output according to the equation 2.

(In the case of estimation method (2))
FIG. 5 is a diagram showing a procedure of the exercise intensity estimation method (2) by the exercise intensity estimation device 10.

  The inputs are the heart rate HR ′ of the person to be estimated and the three-axis acceleration (AccX ′, AccY ′, AccZ ′).

  The norm Acc 'of the input three-axis acceleration is calculated based on the following equation 3.

  The norm Acc of each three-axis acceleration stored in the DB 12 c is calculated based on the following equation 4.

  Among the norm Acc of each three-axis acceleration in the DB 12c calculated based on the equation 4, the three-axis acceleration (AccX, AccY, AccZ) with which the difference between the input three-axis acceleration and the norm Acc 'is small is The heart rate HR associated with the 3-axis acceleration (AccX, AccY, AccZ) of the DB 12c which has a small difference (Acc'-Acc) with respect to the difference (Acc'-Acc) is considered to be similar to the data. The N heart rates HR are selected in ascending order of differences in (where N is an arbitrary constant). At the same time, the PA corresponding to the selected HR is also called from the DB 12c.

The equation 1 including N sets of parameters β ₀ and β ₁ respectively corresponding to N sets of HR and PA is calculated, and the N sets of parameters β ₀ and β _{1 correspond} to the left side and the right side in the equation 1, respectively. A set of parameters β ₀ and β ₁ that minimizes the numerical difference is obtained by the least square method or the like. That is, the parameters β ₀ and β ₁ are calculated from the combination of N HR and PA by the regression analysis according to the equation 1. These values may be calculated in advance using the values of the DB 12c and recorded in the DB 12c. Then, on the basis of the set of parameters β ₀ and β _{1 obtained} as described above and the input heart rate HR ′ of the estimation subject, the exercise amount PAe of the estimation subject is calculated and output according to the equation 2.

  Next, the operation of the exercise intensity estimation device 10 having the above configuration will be described.

  FIG. 6 is a flowchart showing the exercise intensity estimation process (1) (2) according to the exercise intensity estimation program 12b of the exercise intensity estimation device 10.

(Exercise intensity estimation process (1))
The exercise intensity estimation process (1) is a process corresponding to the estimation method (1), and is executed by the processing procedure of steps e1 → e3 to e6.

  The user (estimation target person) wears an electrocardiographic sensor (step e0), and inputs the heart rate HR 'measured by the electrocardiographic sensor (heart rate measuring unit 17) (step e1).

  From the exercise intensity: heart rate: acceleration DB 12c, the heart rate HR having a small difference from the inputted heart rate HR 'of the estimation subject is selected and extracted in order from the smallest difference. At the same time, the exercise intensity PA corresponding to the selected N heart rates HR is also extracted from the DB 12c (step e3).

Based on the formula 1, a set of N heart rate HR and exercise intensity HR extracted from the exercise intensity: heart rate: acceleration DB 12c is input and regression analysis is performed on parameters β ₀ and β ₁ representing the relationship between the two. To calculate (step e4).

Using the parameters β ₀ and β ₁ calculated in step e4 and the current heart rate HR ′ of the person to be estimated input in step e1, the exercise intensity estimated value PAe is calculated according to the equation 2 Is calculated (step e5).

  The estimated value PAe of the exercise intensity of the estimation target person calculated in step e5 is output to the display unit 20 or other output device for notification (step e6).

(Exercise intensity estimation process (2))
The exercise intensity estimation process (2) is a process corresponding to the estimation method (2), and is executed by the processing procedure of steps e2 to e6.

  A user (estimation target) wears an electrocardiographic sensor (step e0), and the heart rate HR ′ and triaxial acceleration (AccX) measured by the electrocardiographic sensor (heart rate measuring unit 17, triaxial acceleration measuring unit 18). ', AccY', AccZ ') are input (step e2).

  The norm Acc 'of the input three-axis acceleration is calculated based on the equation 3. Further, the norm Acc of each three-axis acceleration stored in the exercise intensity: heart rate: acceleration DB 12 c is calculated based on the equation 4 above. Then, among the norms Acc of each 3-axis acceleration in the DB 12c, the 3-axis acceleration (AccX, AccY, AccZ) reduces the difference (Acc'-Acc) with the norm Acc 'of the input 3-axis acceleration. N associated heart rates HR are selected and extracted in the order of small difference in the norm. At the same time, the exercise intensity PA corresponding to the selected N heart rates HR is also extracted from the DB 12c (step e3).

Based on the formula 1, a set of N heart rate HR and exercise intensity HR extracted from the exercise intensity: heart rate: acceleration DB 12c is input and regression analysis is performed on parameters β ₀ and β ₁ representing the relationship between the two. To calculate (step e4).

Using the parameters β ₀ and β ₁ calculated in step e4 and the current heart rate HR ′ of the person to be estimated input in step e2, the exercise intensity estimated value PAe is calculated according to the equation 2 Is calculated (step e5).

  The estimated value PAe of the exercise intensity of the estimation target person calculated in step e5 is output to the display unit 20 or other output device for notification (step e6).

  FIG. 7 shows the estimated values obtained by estimating the exercise intensity of the person to be estimated in the process of actually performing the work according to the estimation processes (1) and (2) of the exercise intensity estimation apparatus 10 and an oxygen intake meter. It is a figure which compares and shows the measured value (true value) measured.

  The exercise intensity here is represented by a value obtained by dividing the oxygen intake during exercise by the oxygen intake during rest, and the unit is (METs).

  In addition, the actual work consists of a series of five rests [rest1], ..., [rest5] with a predetermined preparation exercise [radio] and three work [task1] [task2] [task3]. Done by work.

  In each of the tasks [task1] [task2] [task3], 10 sets of walking (4 km / h), transportation (10 kg, 4 km / h), unloading (6 kg) (6 walking, 2 transportation, 2 unloading) combination It was done.

  The estimated value by the comparison method shown in FIG. 7 is an estimated value when the exercise intensity is calculated using all data sets without selecting the data in the exercise intensity: heart rate: acceleration DB 12c. .

  FIG. 8 shows an average of absolute errors between the estimated values obtained by estimating the exercise intensity of the person to be estimated and the measured values (true values) in the process of actually performing the work according to the estimation processes (1) and (2). FIG.

  FIG. 9 shows the average of the correlation coefficients between the estimated values obtained by estimating the exercise intensity of the person to be estimated and the measured values (true values) in the process of actually performing the work according to the estimation processes (1) and (2). FIG.

  According to this, as shown in FIG. 8, the average of the absolute error with the actually measured value (true value) by the comparison method is “0.412 (METs)”, whereas in the estimation method (1), “0.341 (METs)” and the estimation method (2) are “0.364 (METs)”, and the error is much smaller, and as shown in * a, the estimation method (1) provides a more accurate estimation. It can be seen that the value is obtained.

  As shown in FIG. 9, as for the correlation coefficient, there is no big difference in any method around “0.9”, and as shown by * b, the estimation method (1) is also an actual measurement value (true value). It can be seen that a sufficient correlation is obtained.

Therefore, according to the exercise intensity estimation device 10 having the above-described configuration, the exercise intensity PA, the heart rate HR, and the triaxial acceleration (AccX, AccY, AccZ) corresponding to each of a plurality of types of movements of the subject are associated with each other. : Heart rate: Acceleration DB 12c is stored in advance. In the estimation method (1), the heart rate HR ′ of the estimation target person is measured and input by the heart rate measuring unit 17, and the heart rate HR corresponding to the input heart rate HR ′ of the estimation target person is input from the DB 12c. The exercise intensity PA is selected and extracted. Then, parameters β ₀ and β ₁ indicating the relationship between the heart rate HR extracted from the DB 12c and the exercise intensity PA are calculated according to the equation 1, and the calculated heart rate HR ′ of the estimation target person is calculated. Based on the parameters β ₀ and β ₁ , the exercise strength PAe of the estimation target person is calculated by regression analysis according to the equation 2.

Further, in the estimation method (2), the heart rate HR ′ of the estimation target person is measured and inputted by the heart rate measuring unit 17, and the three-axis acceleration (AccX ′, AccY ′, AccZ ′) of the estimation target person is also input. Is measured and input by the three-axis acceleration measuring unit 18. Then, the heart rate HR and the exercise intensity associated with the three-axis acceleration (AccX, AccY, AccZ) corresponding to the three-axis acceleration (AccX ′, AccY ′, AccZ ′) input from the DB 12c. Select and extract PA. Then, parameters β ₀ and β ₁ indicating the relationship between the heart rate HR extracted from the DB 12c and the exercise intensity PA are calculated according to the equation 1, and the calculated heart rate HR ′ of the estimation target person is calculated. Based on the parameters β ₀ and β ₁ , the exercise strength PAe of the estimation target person is calculated by regression analysis according to the equation 2.

  This makes it possible to estimate exercise intensity PAe with high accuracy even in a situation where a plurality of operations including a small motion such as a construction worker occur.

  In the embodiment, the numerical value (unit METs) corresponding to the oxygen intake is used as the numerical value representing the exercise intensity, but is not limited to this, and another numerical value that changes according to the heart rate HR. May be used as an index.

  Each processing method by the exercise intensity estimation apparatus 10 described in each of the above embodiments, that is, exercise intensity estimation processing by the estimation method (1) shown in the flowcharts of FIGS. 4 and 6, and estimation shown in the flowcharts of FIGS. Each method such as exercise intensity estimation processing by method (2) is a program that can be executed by a computer, such as a memory card (ROM card, RAM card, etc.), magnetic disk (floppy (registered trademark) disk, hard disk, etc. ), Optical disc (CD-ROM, DVD, etc.), semiconductor memory, etc., can be stored and distributed in a medium of an external recording device. Then, the computer (control device: CPU) of the electronic device capable of inputting the estimation target person's heart rate HR and triaxial acceleration (AccX, AccY, AccZ) stores the program stored in the medium of the external recording device. And the operation is controlled by the read program, the exercise intensity estimation function described in each of the above embodiments can be realized, and the same processing by the method described above can be executed.

  The program data for realizing each method can be transmitted as a program code form on a network, and the program data is transmitted from a computer device connected to the network to the estimation subject's heart rate. It is also possible to implement the exercise intensity estimation function described above by taking HR and triaxial acceleration (AccX, AccY, AccZ) into an electronic device that can be input and storing them in a storage device.

  The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the invention when it is practiced. Furthermore, the above embodiments include inventions of various stages, and various inventions can be extracted by appropriate combinations of a plurality of disclosed configuration requirements. For example, even if some constituent elements are deleted from all the constituent elements shown in each embodiment or some constituent elements are combined in different forms, the problems described in the column of the problem to be solved by the invention If the effects described in the column “Effects of the Invention” can be obtained, a configuration in which these constituent requirements are deleted or combined can be extracted as an invention.

10 ... Exercise strength estimation device 11 ... CPU
12 ... storage unit 12a ... device control program 12b ... exercise intensity estimation program 12c ... exercise intensity: heart rate: acceleration DB (database)
13 ... recording medium reading unit 14 ... external recording medium 15 ... communication unit 16 ... RAM
16a ... Heart rate data memory 16b ... Acceleration data memory 17 ... Heart rate measuring unit 18 ... 3-axis acceleration measuring unit

Claims (5)

Data storage means for storing data in which exercise intensity and heart rate corresponding to each of a plurality of types of movements of a subject are associated;
Heart rate input means for inputting the heart rate of the person to be estimated;
Among the heart rates corresponding to each of the plurality of types of actions stored in the data storage means, the heart rates of the person to be estimated that are input by the heart rate input means are smaller Data extraction means for selecting and extracting exercise intensity associated with the plurality of heart rates;
The relationship between the exercise intensity and the heart rate in each of a plurality of heart rate and exercise intensity sets extracted from the data storage means by the data extraction means when the left side and the right side of the equation are used is represented. Regression analysis means for performing regression analysis to find the parameter that minimizes the numerical difference between the left side and the right side of the equation among the parameters;
Exercise intensity calculation means for calculating an exercise intensity corresponding to the heart rate of the person to be estimated input by the heart rate input means based on the parameter obtained by the regression analysis means ;
An exercise intensity estimation device comprising: Data storage means for storing data in which exercise intensity, heart rate, and acceleration corresponding to each of a plurality of types of movements of a subject are associated;
Heart rate input means for inputting the heart rate of the person to be estimated;
Acceleration input means for inputting an acceleration of the person to be estimated;
Among the accelerations corresponding to each of the plurality of types of motion stored in the data storage means, the plurality of accelerations are the same as the plurality of accelerations from the smaller difference with the acceleration of the person to be estimated input by the acceleration input means . Data extraction means for selecting and extracting the heart rate and the exercise intensity associated with the acceleration;
The relationship between the exercise intensity and the heart rate in each of a plurality of heart rate and exercise intensity sets extracted from the data storage means by the data extraction means when the left side and the right side of the equation are used is represented. Regression analysis means for performing regression analysis to find the parameter that minimizes the numerical difference between the left side and the right side of the equation among the parameters;
Exercise intensity calculation means for calculating an exercise intensity corresponding to the heart rate of the person to be estimated input by the heart rate input means based on the parameter obtained by the regression analysis means ;
An exercise intensity estimation device comprising: A heart rate input step for inputting the heart rate of the person to be estimated;
Among the heart rates stored in the data storage means storing data in which the exercise intensity and the heart rate are associated with each of a plurality of types of motions of the subject, the heart rate input step A data extraction step of selecting and extracting the plurality of heart rates and the exercise intensity associated with the plurality of heart rates from the side with the smallest difference from the heart rate of the estimation subject;
Both relationships when with the exercise intensity and heart rate and the left and right side of expression in each of the set of data a plurality of heart rate extracted from the storage means and the exercise intensity by the data extraction step A regression analysis step of performing regression analysis to obtain a parameter in which the numerical difference between the left side and the right side of the equation is the smallest among the parameters to be represented;
An exercise intensity calculation step of calculating an exercise intensity corresponding to the heart rate of the person to be estimated input in the heart rate input step based on the parameter obtained in the regression analysis step ;
An exercise intensity estimation method comprising: A heart rate input step for inputting the heart rate of the person to be estimated;
An acceleration input step for inputting the acceleration of the person to be estimated;
Among the accelerations stored in the data storage means storing data correlating exercise intensity, heart rate and acceleration according to each of a plurality of types of motions of the subject, the acceleration input step A data extraction step of selecting and extracting a plurality of accelerations and a heart rate and exercise intensity associated with the plurality of accelerations from the smaller difference with the acceleration of the estimation subject;
Both relationships when with the exercise intensity and heart rate and the left and right side of expression in each of the set of data a plurality of heart rate extracted from the storage means and the exercise intensity by the data extraction step A regression analysis step of performing regression analysis to obtain a parameter in which the numerical difference between the left side and the right side of the equation is the smallest among the parameters to be represented;
An exercise intensity calculation step of calculating an exercise intensity corresponding to the heart rate of the person to be estimated input in the heart rate input step based on the parameter obtained in the regression analysis step ;
An exercise intensity estimation method comprising:   A program for causing a computer to function as the exercise intensity estimation device according to claim 1 or 2.