CN112509697B

CN112509697B - Multi-module intelligent health management information system

Info

Publication number: CN112509697B
Application number: CN202011407218.0A
Authority: CN
Inventors: 杨洋; 张萌; 盛姝; 解绮雯; 娄威; 李祯然; 王维; 黄奇
Original assignee: Nanjing University
Current assignee: Nanjing University
Priority date: 2020-12-04
Filing date: 2020-12-04
Publication date: 2023-08-22
Anticipated expiration: 2040-12-04
Also published as: CN112509697A

Abstract

The invention provides a multi-module intelligent health management information system which comprises a first crowd classification module, a second muscle-increasing weight-increasing scoring module and a third self-adaptive parameter optimization module, wherein the first crowd classification module is used for classifying people; the first module acquires data such as user age, gender, weight, body fat rate, exercise proficiency, diseases and the like provided by the collected user to classify the health parameters of the crowd, and obtains the health parameter range of the classified crowd; the second module obtains the crowd classification parameters provided by the first module, the user diet provided by the user and the exercise record parameters to perform health management target-muscle-increasing and weight-increasing target score calculation to obtain the user score; the third module obtains the score record of each classified user provided by the second module, obtains the muscle and weight increasing variable quantity data provided by the user, carries out self-adaptive parameter optimization, obtains the score parameter after optimization, and feeds the score parameter back to the second module so as to optimize the score of the second module, namely, optimize the score parameter as the health management parameter and the target of the user; the invention has simple structure, strong function, convenient operation and extremely high commercial value.

Description

Multi-module intelligent health management information system

Technical Field

The invention relates to an information comprehensive, analysis and evaluation system for helping human physiological health, in particular to a multi-module intelligent health management information system.

Background

CN2017111065179 discloses an analysis and evaluation system for physiological information data, which comprises a data input module, a data classification module, a single health index calculation module and a health index comprehensive weighting module, wherein the single health index calculation module comprises a rising normal distribution data calculation module, a falling normal distribution data calculation module and a platform type distribution data calculation module, a historical data analysis module is further arranged in the single health index calculation module, a normalization index adjustment module is arranged in the historical data analysis module, the historical data analysis module is used for carrying out comparative analysis on new physiological information data, corresponding historical physiological information data and a reference target value to obtain a historical variation trend result of the physiological information data, and the normalization index adjustment module is used for adjusting the normalization index of the new physiological information data according to the historical variation trend result; the single health index calculation module is also provided with a health index adjustment module, and the health index adjustment module comprehensively weights the new health index and the corresponding historical health index thereof to obtain an adjustment value of the health index; however, due to the wide variety of the sex, the weight, the age, the life, the exercise habit and the like of individuals, a unified standard is adopted to judge and guide the loss of healthy life of people; and the user score is obtained through the comprehensive score calculation in aspects of scientific classification, final perfection, movement and health management indexes, muscle increase, weight increment and the like. This can provide a more effective health management effect.

Disclosure of Invention

The invention aims to provide a multi-module intelligent health management information system, which is used for helping health management and advice of people, including advice in aspects of diet and exercise, counting and calculating and evaluating the health state of the individuals through a plurality of intelligent means by a scientific and matched crowd classification module, a muscle and weight increasing scoring module and a module capable of adaptively giving out health and exercise parameter optimization indexes, and has a relatively recognized health evaluation basis at present.

The technical scheme of the invention is that the multi-module intelligent health management information system is characterized by comprising a first crowd classification module, a second muscle-increasing weight-increasing scoring module and a third self-adaptive parameter optimization module; the first module acquires data such as user age, gender, weight, body fat rate, exercise proficiency, diseases and the like provided by the collected user to classify the health parameters of the crowd, and obtains the health parameter range of the classified crowd; the second module obtains the crowd classification parameters provided by the first module, the user diet provided by the user and the exercise record parameters to perform health management target-muscle-increasing and weight-increasing target score calculation to obtain the user score; the third module obtains the score record of each classified user provided by the second module, obtains the muscle and weight increasing variable quantity data provided by the user, carries out self-adaptive parameter optimization, obtains the score parameter after optimization, and feeds the score parameter back to the second module so as to optimize the score of the second module, namely, optimize the score parameter as the health management parameter and the target of the user;

The module one-person group classification module comprises two sub-modules, namely a sub-module 1.1 data collection sub-module and a sub-module 1.2 user data-crowd classification mapping sub-module;

the sub-module 1.1 is used for acquiring personal data of a user provided by the user and collecting the user data; the sub-module obtains the following user data provided by the user: the method comprises the steps of performing user data collection on race, gender, body fat rate, disease, exercise habit/fatigue, pump feeling, pain feeling, height H, weight W, age Y, weight delta W required to be increased in one week, exercise time t, anaerobic exercise group number m, anaerobic exercise group number n, anaerobic exercise unit intensity weight omega, 4 nutrient intake, exercise type and exercise intensity data, and obtaining user data required by crowd classification mapping;

the sub-module 1.2 is used for obtaining user data required by the crowd classification mapping provided by the sub-module 1.1, carrying out user data-crowd classification mapping calculation, and obtaining crowd classification parameters;

in the specific example, one user is Asian, healthy person, regular sporter, male and BFR is less than or equal to 25%, and the classification parameter is g ₁ ＝0.95； g ₂ ＝1，ν ₁ ＝1/3，ν ₂ ＝1/3，ν ₃ ＝1/3；DCE ₂ ＝275kcal，α＝66.47、β＝13.75、γ＝5.0033、Ω＝6.775；WWRF ＝485kcal/kg。

The sub-module 1.2 randomly divides crowd data into a test set and a training set; in a training set, carrying out feature selection on crowd features, adopting an information gain standard, and recursively constructing a decision tree; setting a loss function in consideration of the complexity of the decision tree, and pruning the decision tree; finally classifying the crowd according to the constructed decision tree model to obtain a training model, and testing the classifying model by using a testing set; constructing a decision tree; obtaining a decision tree finally until the information gain of all the features is very small or no features can be selected; the characteristics are selected as 5 types of characteristics of race, disease, exercise habit, male and female body fat rate, and the crowd is finally divided into 32 types;

the second muscle-increasing and weight-increasing scoring module comprises five sub-modules, namely a user data collecting sub-module 2.1, a motion stimulation scoring computing sub-module 2.2, a caloric scoring computing sub-module 2.3, a diet proportion scoring computing sub-module 2.4 and a muscle-increasing and weight-increasing comprehensive scoring computing module 2.5; the sub-module 2.1 is used for acquiring diet and exercise records provided by a user and collecting user data to obtain diet and exercise data of the user; the sub-module 2.2 motion stimulation score calculation sub-module obtains crowd classification parameters provided by the module group classification module, obtains user motion data provided by the sub-module 2.1 user data collection sub-module, calculates motion stimulation scores, and obtains user motion state calculation data and motion stimulation scores; the sub-module 2.3 caloric score calculating sub-module obtains the user diet and exercise data provided by the sub-module 2.1 user data collecting sub-module, obtains the user exercise state data provided by the sub-module 2.2 exercise stimulus score calculating sub-module, obtains the crowd classification parameters provided by the crowd classification module of the module, carries out caloric score calculation, and obtains caloric score; the sub-module 2.4 diet proportion score calculation sub-module obtains the diet data of the user provided by the sub-module 2.1 user data collection sub-module, obtains the crowd classification parameters provided by the crowd classification module of the module, calculates diet proportion scores, and obtains diet proportion scores; the sub-module 2.5 muscle and weight increasing comprehensive score calculation module obtains the motor stimulation score provided by the sub-module 2.2 motor stimulation score calculation sub-module, obtains the caloric score provided by the sub-module 2.3 caloric score calculation sub-module, obtains the diet proportion score provided by the sub-module 2.4 diet proportion score calculation sub-module, obtains the optimized comprehensive score superparameter provided by the module three self-adaptive parameter optimization module, and carries out muscle and weight increasing comprehensive score calculation to obtain the user score;

In the specific example, one user initial score parameter is ρ ₁ ＝0.4，ρ ₂ ＝0.3，ρ ₃ After the correction by the periodic data importing module III, the score parameter is optimized to be ρ =0.3 ₁ ＝0.35，ρ ₂ ＝0.35，ρ ₃ ＝0.3。

The third score super-parameter self-adaptive optimization module comprises two sub-modules, namely a user data collection sub-module 3.1 and a score super-parameter self-adaptive optimization calculation sub-module 3.2; the sub-module 3.1 is used for acquiring the muscle-increasing weight-increasing variable quantity data provided by a user, acquiring the user score provided by the second muscle-increasing weight-increasing score module and collecting the user data; the sub-module 3.2 score super-parameter self-adaptive optimization calculation sub-module obtains the user data required by score super-parameter optimization provided by the sub-module 3.1 user data collection sub-module, and performs score super-parameter self-adaptive optimization calculation to obtain the optimized comprehensive score super-parameter; and the score hyper-parameter self-adaptive optimization submodule adopts a feature selection method to evaluate the relation between each input variable and the target variable, so as to obtain the optimal weight.

The sub-module 2.5 muscle and weight increasing comprehensive score calculation module obtains the motor stimulation score provided by the sub-module 2.2 motor stimulation score calculation sub-module, obtains the caloric score provided by the sub-module 2.3 caloric score calculation sub-module, obtains the diet proportion score provided by the sub-module 2.4 diet proportion score calculation sub-module, obtains the optimized comprehensive score superparameter provided by the module three self-adaptive parameter optimization module, and performs muscle and weight increasing comprehensive score calculation to obtain the user score.

The beneficial effects are that: according to scientific evidence study, the weight index, exercise and macro nutrient reference ratio related by the invention accords with the human health and energy metabolism rule, and has the purposes of muscle increase and health. On the premise that macro nutrients are determined by a macro nutrient large-scale framework, new ideas of the modern health are to be further absorbed, such as: (1) a protein source; (2) suitable use of unsaturated fatty acids; (3) The main food is matched with substances with different blood sugar loads, so that the blood sugar health is improved. And obtaining the diet proportion score by obtaining the crowd classification parameters provided by the crowd classification module.

The invention has simple structure, strong function, convenient operation and extremely high commercial value.

Drawings

Fig. 1: a multi-module intelligent health management information system;

fig. 2: a crowd classification module data flow diagram;

fig. 3: a muscle-increasing weight-increasing scoring module data flow diagram;

fig. 4: and a score super-parameter self-adaptive optimization module data flow diagram.

Detailed Description

Referring to the drawings, FIG. 1 is a multi-module intelligent health management information system:

module-crowd classification module

For a module-person group classification module, the sub-module data flow diagram is shown in FIG. 2. The crowd classification module comprises two sub-modules, namely a data collection sub-module and a user data-crowd classification mapping sub-module. The sub-module 1.1 is used for acquiring personal data of a user provided by the user and collecting the user data; the sub-module 1.2 is used for obtaining the user data required by the crowd classification mapping provided by the sub-module 1.1, carrying out the user data-crowd classification mapping calculation, and obtaining crowd classification parameters.

Referring to fig. 2: crowd classification module data flow graph

Module 1.1 user data collecting sub-Module

For a 1.1 user data collection sub-module, the sub-module obtains the following user data provided by the user: the method comprises the steps of collecting user data, namely, race, sex, body fat rate, whether a patient suffers from diseases, exercise habit/fatigue, pump feeling, pain feeling, height H, weight W, age Y, weight delta W required to be increased in one week, exercise time t, anaerobic exercise group number m, anaerobic exercise group number n, anaerobic exercise unit intensity weight omega, 4 nutrient intake, exercise type and exercise intensity data, and obtaining user data required by crowd classification mapping.

Module 1.2 user data-crowd Classification mapping submodule

And for the 1.2 user data-crowd classification mapping sub-module, the sub-module acquires user data required by the crowd classification mapping provided by the user data collecting sub-module of the module 1.1, and performs user data-crowd classification mapping to obtain crowd classification parameters.

The specific practice is that the module 1.2 randomly divides crowd data into a test set and a training set. In the training set, the crowd characteristics are subjected to characteristic selection, information gain standards are adopted, and a decision tree is constructed recursively. And then, setting a loss function by considering the complexity of the decision tree, and pruning the decision tree. And finally classifying the crowd according to the constructed decision tree model to obtain a training model, and testing the classification model by using a test set. The information gain standard employs l (x) _i )＝-log ₂ p(x _i ) Wherein the symbol x _i Representing a sort of case; p (x) _i ) Representing a probability of selecting the category; l (x) _i ) Represents x _i Is used for the information amount of the (a). To calculate entropy we need to calculate the expected values of the information contained in all possible values of all classesWherein p (x) _i ) The probability of selecting the class is represented, and n represents a total of n class cases. Empirical entropy ofWherein D represents a training data set, C _k Represents a certain classification situation of the training data set, K represents a total of K classification situations, |C _k I is belonging to class C _k Is a number of samples of (a).

The conditional entropy H (y|x) represents the uncertainty of the random variable Y under the condition that the random variable X is known,wherein, when the random variable X is X _i In the case of such classification, the probability p of the classification _i ＝P(X＝x _i ) N represents a total of n classification cases.

The information gain g (D, a) =h (D) -H (d|a) of the feature a to the training data set D, wherein H (D) is the empirical entropy of the set D, and H (d|a) represents the difference between the empirical conditional entropy of D given the feature a. Then constructing a decision tree, starting from a root node, calculating information gains of all possible features of the node, selecting the feature with the largest information gain as the feature of the node, establishing child nodes according to different values of the feature, and recursively calling the child nodes to construct the decision tree; and finally, obtaining a decision tree until the information gain of all the features is small or no features can be selected. Then pruning the decision tree by minimizing the integral loss function or cost function of the decision tree, wherein the loss function is that Wherein T represents the current leaf node, and T represents all leaf nodes of the subtree; h _t (T) represents entropy of the T-th leaf; n (N) _t Representing the number of training samples contained in the t leaf node; alpha represents a penalty coefficient.Wherein the N is _t Representing the number of training samples contained in the t leaf node; n (N) _tk The number of training samples contained in the kth class of the t leaf node is represented, and K represents a total of K classification cases. In this case, the characteristics are selected as 5 types of characteristics (race, disease, exercise habit, man and woman, body fat rate), and the population is finally divided into 32 types. The crowd classification module variable table and the crowd classification parameter table are shown in table 1 and table 2.

Table 1: crowd classification module variable table

Table 2: crowd classification parameter table

Module two muscle-increasing weight-increasing scoring module

For the module two, the muscle-increasing weight-increasing score module, the sub-module data flow diagram is shown in figure 3. The muscle and weight increasing scoring module comprises five sub-modules, namely a user data collecting sub-module 2.1, a motion stimulation scoring computing sub-module 2.2, a caloric scoring computing sub-module 2.3, a diet proportion scoring computing sub-module 2.4 and a muscle and weight increasing comprehensive scoring computing module 2.5. The sub-module 2.1 is used for acquiring diet and exercise records provided by a user and collecting user data to obtain diet and exercise data of the user; the sub-module 2.2 motion stimulation score calculation sub-module obtains crowd classification parameters provided by the module group classification module, obtains user motion data provided by the sub-module 2.1 user data collection sub-module, calculates motion stimulation scores, and obtains user motion state calculation data and motion stimulation scores; the sub-module 2.3 caloric score calculating sub-module obtains the user diet and exercise data provided by the sub-module 2.1 user data collecting sub-module, obtains the user exercise state data provided by the sub-module 2.2 exercise stimulus score calculating sub-module, obtains the crowd classification parameters provided by the crowd classification module of the module, carries out caloric score calculation, and obtains caloric score; the sub-module 2.4 diet proportion score calculation sub-module obtains the diet data of the user provided by the sub-module 2.1 user data collection sub-module, obtains the crowd classification parameters provided by the crowd classification module of the module, calculates diet proportion scores, and obtains diet proportion scores; the sub-module 2.5 muscle and weight increasing comprehensive score calculation module obtains the motor stimulation score provided by the sub-module 2.2 motor stimulation score calculation sub-module, obtains the caloric score provided by the sub-module 2.3 caloric score calculation sub-module, obtains the diet proportion score provided by the sub-module 2.4 diet proportion score calculation sub-module, obtains the optimized comprehensive score superparameter provided by the module three self-adaptive parameter optimization module, and performs muscle and weight increasing comprehensive score calculation to obtain the user score.

Referring to fig. 3: and a muscle and weight increasing scoring module data flow diagram, wherein for a sub-module 2.1, a user data collecting sub-module acquires diet and exercise records provided by a user, and user data collection is carried out to obtain diet and exercise data of the user.

Module 2.2 Sport stimulus score calculation submodule

And for the sub-module 2.2, the motion stimulation score calculation sub-module acquires the crowd classification parameters provided by the module group classification module, acquires the user motion data provided by the sub-module 2.1 user data collection sub-module, calculates the motion stimulation score, and obtains the user motion state calculation data and the motion stimulation score.

The control method comprises the steps of providing an absolute score S of daily exercise stimulus in the period _a The method is obtained by the following formula:

wherein the x= (a) ₁ ，a ₂ ，a ₃ ……a _F ) For the pre-motion description vector, it is defined as each actual component of F indexes for the process a, a _f Score for pre-exercise index f, (a) _f ) _max The score is the highest score of the index f before exercise; the y= (b) ₁ ，b ₂ ，b ₃ ……b _F ) For describing vectors after motion, defineFor each actual component of the f indexes of the b process, b _f Score for post-exercise index f, (b) _f ) _max The score is the highest score of the index f after exercise; θ is the in-motion adjustment coefficient, θmax is the maximum value of the in-motion adjustment coefficient. F is 1, 2 and 3 … … F (F is the maximum value of F), and only three indexes are considered at the present stage.

2. The daily motor stimulus of the athlete is related to a number of indicators, and the current phase analysis considers: the synergistic relationship of the three reference indexes is enhanced, the foundation of fatigue is not highlighted, and the two effective time points which are beneficial to evaluation, namely before training and after training, are obtained, and the row vectors which are measured under the three indexes are respectively subjected to dot product operation to obtain a number product. Based on realistic scenario diversity and individual variability. The expected value range should be expanded, so that the user self-evaluation dimension is improved, after the training is finished, the user is guided to recall and evaluate the whole training experience, and the theory is developed by combining the related high-index documents, presetting the adjustment coefficient.

(1) Research shows that at the beginning of one training, the fatigue of the organism is usually found in regular sports crowd, or the natural training frequency of the crowd and the target guiding type in the system training, and the stimulation depth of the successive training program to the trainee. Thus, when the default exerciser is in a non-fully recovered state, the perceived reversal of light fatigue by the exerciser can be equated to a benign state in which the target muscle is more prone to pumping fresh blood and helping to activate the muscle—the nervous system. Indeed, moderate to intense fatigue may give rise to a gain in training, but it should be noted that moderate to intense fatigue may significantly increase the probability of exercise risk (e.g. heart and lung system breakdown, muscle and fascia injury, rhabdomyolysis, etc.) occurring in training, whereas in reality, periodic, cyclic differentiation training concepts have profound effects in the above population, fatigue load training is unusual, and thus in the present system, it should still be discussed from a generalized muscle hypertrophy point of view, many demonstration studies indicate that under moderate deep fatigue, pressure of muscle work increases, there is a significant correlation between pressure level and pressure hormone such as cortisol in blood, at which time muscles often regulate part of biochemical metabolic processes, especially nitrogen metabolism (e.g. reverse nitrogen retention), based on self-protection mechanisms formed under long-term natural selection of humans, namely: causing the decomposition of more than usual muscles and bringing negative effects to muscle proliferation. Further enriching data, and providing similar quantitative regulation for the perceived fatigue of the body after exercise.

(2) For pump feel, a large amount of empirical data indicates that "pump feel rooted in muscle memory" is one of: such as the exercise-acquired training characteristics which are difficult to be displayed when a certain strengthening degree is not achieved for a motion. For higher-order sporters, the strong sense of pump is remarkable in the modeling of muscle morphology, and is an important reference index for measuring real benefit of training; for common sportsmen, the feeling of pump experience is obtained primarily through exogenous experimental design and adaptation of the sportsmen to the environment system of the executors, but due to the general short plates (poor muscle endurance, insufficient joint stability and low development degree of deep muscle groups) in the training stage, excessive feeling of pump or muscle injury with delay perception is resisted, and the storage of muscles is resisted.

(3) The feeling of soreness and distension is quite different from individual to individual, and can be generated by following the stimulation of muscle characteristics in a sufficient amount, generally has delay, can be accumulated in a short period, can be relieved by training and proper stretching, reduces lactic acid accumulation and promotes recovery. An important belief of training is: after the load is applied and the training is carried out, the nutrient substances are reasonably ingested, and the excessive recovery can be stimulated. While within the proper range, the pain sensation is positively correlated with the positive balance of protein.

Law chart:

index one:

table 3: index-score standard table for regular sporter

Index I (regular trainer)	Perception assessment	Score (before sports)	Score (after sports)
				Feeling of fatigue	Without any means for	3	2
	Mild and mild	4	3
					Moderate degree	3	4
	Strong intensity	2	3

Table 4: index-score standard table for common sporter

Index (I)One (general health person)	Perception assessment	Score (before sports)	Score (after sports)
				Feeling of fatigue	Without any means for	4	1
	Mild and mild	3	2
					Moderate degree	2	3
	Strong intensity	1	4

Index II:

table 5: index two-score standard table for regular sporter

Index II (regular sporter)	Perception assessment	Score (before sports)	Score (after exercise)
				Pump feel	Without any means for	2	2
	Slight	3	3
					Medium and medium	4	4
	Strong intensity	3	3

Table 6: standard table for two scores of index of common sporter

And (3) index III:

table 7: index three-score standard table for regular sporter

Table 8: index three-score standard table for common sporter

Index III (common sporter)	Perception assessment	Score (before sports)	Score (after exercise)
				Pain sensation of sour	Without any means for	3	1
	Slight	4	2
					Medium and medium	3	3
	Strong intensity	2	4

Full training experience assessment:

table 9: adjustment coefficient standard table

Perception assessment	Adjustment coefficient
		Intermittent type	0.8
Normal state	1
		Fluency	1.1
Immersion	1.2

Table 10: variable table of motor stimulation score calculation submodule

3. And taking uniformity and range limitation of score distribution into consideration, carrying out normal distribution conversion on the absolute score so that the placement degree is 95%, and taking the value of the central axis as the maximum value of the absolute score. I.e. at In the distribution of the particles,and μ= (S _a ) _max . S after transformation _a For the final motor stimulus score.

sub-Module 2.3 caloric score calculation sub-Module, daily caloric Absolute score S over period _b The method is obtained by the following formula:

wherein the S is _M The actual daily heat surplus in the period is calculated; WRF is the target weighting energy; s is S _yn Daily dietary energy intake for a period; s is S _ym Energy is consumed daily for each day in a period; s is S _s Energy is consumed for daily movements during the cycle.

Daily dietary energy intake during the period in the steps is obtained through the following formula:

wherein the FIE _i Intake of energy for nutrient i in the food; k (k) _i Taking the energy coefficient of the nutrient i of unit mass as the reference human digestibility; ADI (ADI) _i Intake of nutrient i; i is 1, 2 and 3 … … I (I is the maximum value of I), and only four nutrients are considered at the present stage, namely: FIE (FIE) ₁ Representing carbohydrate intake energy, FIE in food ₂ Representing the energy intake of protein in food, FIE ₃ Representing fat intake energy, FIE in a fat diet ₄ Represents dietary fiber intake energy in the food; ADI (ADI) ₁ ADI represents carbohydrate intake energy in food ₂ ADI for protein intake ₃ ADI for fat intake ₄ The intake of dietary fiber; k (k) ₁ In the case of reference to human digestibility, the unit of intakeThe energy coefficient of the mass carbohydrate is generally taken as 4kcal/g; k (k) ₂ Taking the energy coefficient of protein of unit mass under the condition of referencing the human digestibility, and generally taking the value of 4kcal/g; k (k) ₃ Taking the energy coefficient of fat of unit mass under the condition of referencing the human digestibility, and generally taking the value of 9.6kcal/g; k (k) ₄ In order to refer to the actual digestibility of the human body, the energy coefficient of dietary fiber per unit mass is generally taken as 2kcal/g.

Those skilled in the art understand that each macro-nutrient ingested produces less net available calories (as compared to the identified calories of the macro-nutrient before meal) due to the food heating effect. In addition, some macronutrients may require more energy consumption (i.e., greater food heating effects) during digestion, absorption, transportation, and storage. Specifically, the thermal effects of fat, carbohydrate and protein are 0% to 3%, 5% to 10% and 20% to 30%, respectively.

Daily energy consumption S daily in a cycle of said steps _ym The method is obtained by the following formula:

wherein the DCE _j For daily activities J to consume energy, J takes 1 and 2 … … J (J is the maximum value of J), only two daily activities are considered at present, namely: DCE (DCE) ₁ Representing basal metabolic energy expenditure, DCE ₂ Indicating that daily activities after exercise consume energy.

Wherein the BMR is a basal metabolic rate; g _r For correction coefficients, R is 1, 2 … … R (R is the maximum value of R), and only five correction coefficients are considered at the present stage, namely: g ₁ Correction coefficient for BMR race, g ₂ Correction coefficients for BMR disease.

BMR adopts Harris-Benedict formula,

BMR＝α+(β×W)+(γ×H)-(Ω×Y)

wherein W is weight, H is height, and Y is age. The user sexes are different, and the parameters alpha, beta, gamma and omega take different values. Determining the sex of the user based on the user basic information, and when the user is male, α= 66.47, β=13.75, γ= 5.0033, Ω=6.775; when the user is female, α= 655.1, β= 9.563, γ=1.850, Ω= 4.676.

Further, the g ₁ Correction factors for BMR ethnicity, g when the user is Asian ethnicity ₁ Take a value of 0.95, but g when the user is not asian ₁ The value is 1. Further, the g ₂ When a user is normal to have a disease, the BMR disease correction coefficient is generally set to 1, and the rest is set according to the comprehensive evaluation of the physical condition of the user.

DCE ₂ Comprises two values, 275 of which the user is a regular sporter and 100 of which the user is a common sporter, and determining DCE suitable for the user basic information based on the user basic information ₂ Substituting the numerical value.

Daily movement during the period of the step consumes energy S _s The method is obtained by the following formula:

wherein the SCE _q The energy is consumed for the motion Q, Q takes 1, 2 and 3 … … Q (Q is the maximum value of Q), and three motion forms are only considered at the present stage, namely: SCE ₁ Energy is consumed for basic exercise, SCE ₂ Energy consumption for aerobic exercise, SCE ₃ Energy is consumed for anaerobic exercises.

The BMR is the basal metabolic rate; said g _r G is the correction coefficient ₁ Correction coefficient for BMR race, g ₂ Correction coefficients for BMR disease.

The METs are physical activity intensities of different aerobic exercises, the W is the weight of a user, the t is exercise time, the AETs are physical activity intensities of different anaerobic exercises, the m is the number of anaerobic exercises, the n is the number of times of each anaerobic exercise, and the omega is the unit intensity weight of the anaerobic exercises.

Those skilled in the art understand that different exercise patterns have different exercise efficiencies for different areas of the body, and are classified into aerobic exercise and resistance exercise. The METs are different physical activity intensities of different aerobic exercises, and according to different aerobic exercise modes, the METs have different values, in a preferred embodiment, when a user selects basketball exercises, the METs have a value of 6.5, when the user selects gymnastics, the METs have a value of 5, when the user selects weight-losing exercises, the METs have a value of 6, when the user selects swimming exercises, the METs have a value of 8, and certainly, the user has corresponding values when the user selects riding bicycles, running or other aerobic exercises, and the values are not stated one by one. The AETs are the basis of different measurement indexes with different body movement intensity differences in the range of the anti-resistance exercise project; in a preferred embodiment, the value of AETs is 1.5 when the user exercise intensity is low; the value of AETs is 2.25 when the exercise intensity of the user is moderate; when the exercise intensity of the user is higher, the value of AETs is 3.5; of course, with reference to the detailed training content entered by the user, the resistive training energy consumption can be further quantified and corrected.

The WRF is obtained by the following formula:

WRF＝WWRF*ΔW

wherein WWRF is the unit weight gain factor corresponding to different BMIs, and DeltaW is the weight required to be increased in one week. Wwrf=769 Kcal/kg when the user is male and the Body Fat Rate (BFR) is greater than 25%; when the user is male, body fat rate is less than 25%, wwrf=485 Kcal/kg; when the user is female, the body fat rate is greater than 35%, wwrf=769 Kcal/kg; when the user is female, the body fat rate is less than 35%, wwrf=485 Kcal/kg.

It will be appreciated by those skilled in the art that adipocytes are composed of 80% fat, 2% protein, 18% water, muscle cells are composed of 27% protein, 73% water, and that for each increase in body weight in a fat population, 30% -40% muscle cells and 0% -70% adipocytes are increased, and for each increase in body weight in a lean population, 1kg body weight is increased, 60% -70% muscle cells and 30% -40% adipocytes are increased. The fatness of men is more than 25% and the fatness of women is more than 30%. The muscle increasing amount is preferably 113g-675g per week. The following list describes, as an example, the case of a value of a fatter population of men, and the other cases will not be repeated.

Table 11: value taking condition of fatter men

Table 12: calorimeter score calculation submodule variable meter

/>

And for the submodule 2.3, a caloric score calculating submodule acquires the user diet and exercise data provided by a submodule 2.1 user data collecting submodule, acquires the user exercise state data provided by a submodule 2.2 exercise stimulus score calculating submodule, acquires crowd classification parameters provided by a crowd classification module of the module, calculates caloric scores and obtains caloric scores.

Taking into account the uniformity and range limitation of the score distribution, the absolute score is subjected to normal distribution conversion so that the placement degree is 9And 5, taking the value of the central axis as the maximum value of the absolute score. I.e. atIn the distribution of the particles,and μ= (S _b ) _max . S after transformation _b Is the final caloric score.

Module 2.4 diet proportion score calculation sub-Module: the absolute score Sc of the daily diet proportion in the period in the step is obtained by the following formula:

wherein the U is _h Is the absolute score of nutrient h, v _h For the weight of nutrient H, H is 1, 2 and 3 … … H (H is the maximum value of H), and only three nutrients are considered at the present stage, namely: u (U) ₁ Represents absolute score of carbon water, U ₂ Representing absolute score of protein, U ₃ Representing an absolute score of fat; v (v) ₁ Representing the weight of the water, v ₂ Representing protein weight, v ₃ Representing fat weights.

The ratio P of the total daily dietary intake energy of the daily nutrient h in the cycle in said step _h The method is obtained by the following formula:

wherein the S is _yn Total energy intake for diet; the FIE _h For the energy intake of nutrient H in food, H is 1, 2 and 3 … … H (H is the maximum value of H), only three nutrients are considered at the present stage, namely: FIE (FIE) ₁ Representing carbohydrate intake energy, FIE ₂ For protein uptake energy, FIE ₃ Energy intake for fat; p (P) ₁ Represents the ratio of carbon to water, P ₂ Represents the protein ratio, P ₃ Indicating fat ratio.

3. When the user is healthy, each nutrient is assigned the same weight by default, and only 3 nutrients, namely v, are considered at the present stage ₁ ＝ν ₂ ＝ν ₃ =1/3. However, to take into account individual variability, certain nutrients may be present in higher proportions. For example, some trainers suffer from diet-related chronic diseases, including cardiovascular disease, type 2 diabetes. By increasing the weight of the nutrient, a user with a high score on the nutrient will obtain a higher score in the overall diet proportion score, i.e., v ₁ 、ν ₁ 、ν ₁ The value is related to the type of chronic diseases related to diet.

Table 13: reference table for nutrient intake standard

In particular, in general, the energy intake ratio of carbohydrate is within 55% to 60%, the energy intake ratio of protein is within 25% to 30%, and the energy intake ratio of fat is within 15% to 20%. Therefore, in general, (P) ₁ ) _min Namely 55% (P) ₁ ) _max Namely 60% (P) ₂ ) _min Namely 25% (P) ₂ ) _max Namely 30% (P) ₃ ) _min Namely 15% (P) ₃ ) _max I.e. 20%.

Absolute score U of nutrient h in daily diet during the period of the step _h The method is obtained by the following formula:

wherein the P is _h The total energy intake in the diet for the daily nutrient h in the cycleIn the ratio, H is 1, 2 and 3 … … H (H is the maximum value of H), and only three nutrients are considered at the present stage, namely: p (P) ₁ Represents the ratio of carbon to water, P ₂ Represents the protein ratio, P ₃ Representing fat ratio; (P) _h ) _min A minimum value of the ratio of total energy taken in the diet for nutrient h; (P) _h ) _max The maximum value of the ratio of total energy taken in the diet for nutrient h.

Those skilled in the art understand that the ratio of the three macronutrients in the total daily dietary intake energy of a population with definite muscle development needs should be 55% to 60% carbohydrate, 25% to 30% protein and 15% to 20% fat. When the whole-day energy condition of the organism is evaluated, the total consumption of diet intake is not lower than 15% and the surplus proportion is required for metabolic synthesis of new tissues. The proposal ensures that sufficient protein is taken in to promote muscle growth, sufficient carbohydrate is taken in to provide sufficient energy for high-strength anti-resistance training, and proper amount of fat is also provided to ensure that blood contains sufficient testosterone. Adenosine Triphosphate (ATP) is also required for protein synthesis, so that sufficient energy is available to ensure protein synthesis. Finally, scientific evidence researches show that the macro nutrient reference ratio meets the human energy metabolism rule and achieves the purposes of muscle enhancement and health. In fact, with reference to recent and developed national dietary guidelines and the like, the new views of digital and logical contemporary health are found to be further absorbed under the premise of being determined by a macro nutrient large scale framework: (1) The protein source selects high quality animal proteins such as fish as much as possible; (2) Unsaturated fatty acid is properly used for replacing saturated fat to promote cardiovascular health; (3) In the aspect of staple food, substances with different blood sugar loads can be matched and combined to optimize the intake of carbon water and improve the blood sugar health.

Table 14: diet proportion score calculating submodule variable table

The sub-module 2.4 diet proportion score calculation sub-module obtains the user diet data provided by the sub-module 2.1 user data collection sub-module, obtains the crowd classification parameters provided by the crowd classification module of the module, calculates diet proportion scores, and obtains diet proportion scores.

And taking uniformity and range limitation of score distribution into consideration, carrying out normal distribution conversion on the absolute score so that the placement degree is 95%, and taking the value of the central axis as the maximum value of the absolute score. I.e. atIn the distribution of the particles,and μ= (S _c ) _max . S after transformation _c Score for the final diet proportion.

S _w The calculation formula is obtained by the formula above for the comprehensive score of muscle and weight increment. Wherein, Representing the weight gain and muscle gain comprehensive score coefficient vector, +.> λ ₁ Representing the motor stimulus score coefficient, lambda ₂ Representing caloric score coefficient, lambda ₃ A score coefficient representing the proportion of diets; />A comprehensive score vector representing muscle and weight gain>S _a Indicating a motor stimulus score, S _b Indicating caloric score, S _c A score representing the proportion of diet; for->And continuously optimizing by a module III self-adaptive parameter optimizing module.

For the calculated muscle and weight increasing comprehensive user score S _w User scores in different ranges represent different muscle and weight gain conditions: when 0 is less than or equal to S _w When the weight of the body is less than 60, the hormone level and the metabolic level of the body of the user are reduced at the stage, and the weight of the body is not obvious for increasing the muscle and the weight of the body for maintaining the internal balance state; when 60 is less than or equal to S _w When the weight of the user is less than 80, the hormone, the metabolism level and various metabolism adaptations of the user are slowly restored in the stage, the user reaches a new weight level, the user keeps for 1-2 weeks in the stage, the relationship with food is reestablished, the caloric intake and the caloric surplus are kept, and the weight gain of the muscle is increased obviously; when 80 is less than or equal to S _w And when the weight is less than 100, the condition of muscle and weight increment at the stage of the user is stable.

Table 15: meter for increasing muscle and weight

User score S _w	Condition of muscle and weight increase
		0～60	Reduced hormone level and metabolism level, and insignificant muscle and weight increase
60～80	Hormone, metabolism level and various metabolism adaptation are slowly restored, and the muscle and weight increment is obviously improved
		80～100	Stable condition of muscle and weight increase

Table 16: variable table of muscle and weight increasing comprehensive score calculation module

Variable name	Meaning of variable	Variable unit	Variable value
				S _w	Comprehensive score for muscle and weight increase	Dividing into	0100
λ ₁	Score coefficient of motor stimulus	Dividing into	01
				λ ₂	Coefficient of caloric score	Dividing into	01
λ ₃	Diet proportion score coefficient	Dividing into	01
				S _a	Sport stimulus score	Dividing into	0100
S _b	Caloric score	Dividing into	0100
				S _c	Diet proportion score	Dividing into	0100

Module three-score super-parameter self-adaptive optimization module

For a module three-score super-parameter self-adaptive optimization module, a sub-module data flow diagram is shown in fig. 4. The self-adaptive parameter optimization module comprises two sub-modules, namely a user data collection sub-module and a score super-parameter self-adaptive optimization calculation sub-module. The sub-module 3.1 is used for acquiring the muscle-increasing weight-increasing variable quantity data provided by a user, acquiring the user score provided by the second muscle-increasing weight-increasing score module and collecting the user data; the sub-module 3.2 score super-parameter self-adaptive optimization calculation sub-module obtains the user data required by score super-parameter optimization provided by the sub-module 3.1 user data collection sub-module, and performs score super-parameter self-adaptive optimization calculation to obtain the optimized comprehensive score super-parameter.

Referring to fig. 4: score superparameter self-adaptive optimization module data flow diagram

Sub-module 3.1 user data acquisition sub-module obtains muscle and weight increasing variation data ΔM provided by the user and obtains user score S provided by the module two muscle and weight increasing scoring module _w And collecting user data.

Table 17: user data collection submodule variable table

Variable name	Meaning of variable	Variable unit	Variable value
				ΔM	Weekly weight gain	kg	/
S _w	Comprehensive score for muscle and weight increase	Dividing into	0-100

The sub-module 3.2 score super-parameter self-adaptive optimization calculation sub-module obtains the user data required by score super-parameter optimization provided by the sub-module 3.1 user data collection sub-module, and performs score super-parameter self-adaptive optimization calculation to obtain the optimized comprehensive score super-parameter.

And the score hyper-parameter self-adaptive optimization submodule adopts a feature selection method to evaluate the relation between each input variable and the target variable, so as to obtain the optimal weight. Here, the pearson correlation coefficient is adopted to obtain the diet proportion score S _a Score S of exercise stimulus _b Score S of caloric content _c And the weight gain variation S of muscle and body _ΔM The relationship of the scores, namely:wherein, the muscle and weight increasing variable quantity score S _ΔM The calculation formula is that the submodule 3.1 user data collecting submodule obtains:

In view of uniformity and range limitation of score distribution, for S _ΔM Normal distribution transformation is carried out by the following steps of _ΔM Normal distribution transformation is carried out to ensure that the placement degree is 95 percent, and the value of the central axis is (S _ΔM ) _max . I.e. atIn the distribution of the particles, and μ= (S _ΔM ) _max . S after transformation _ΔM Score for the final diet proportion.

Is determined by (a)The mode is that each correlation coefficient accounts for the proportion of the total correlation coefficient sum, and the calculation formula is as follows:

as user data increases, the difference between the actual value and the measured value can be compared and used as an optimization parameterFor parameter optimization, the error calculation formula is:

wherein, is a super parameter vector->Lower S _ΔM And S is equal to _w Is a function of the error of (a). The training set and the test set are randomly divided and the parameters are updated continuously using a gradient descent algorithm. Super parameter vector->The update formula is:

wherein, is the gradient of the parameter, η is the learning rate. When->During minimization, the muscle and weight increasing comprehensive score superparameter vector ++>Is a global optimum.

Table 18: score superparameter self-adaptive optimization calculation submodule variable table

Claims

1. A multi-module intelligent health management information system is characterized by comprising a first crowd classification module, a second muscle-increasing weight-increasing scoring module and a third self-adaptive parameter optimization module; the first module acquires user age, gender, weight, body fat rate, exercise proficiency and disease data provided by the collected user to classify the health parameters of the crowd, and obtains the health parameter range of the classified crowd; the second module obtains the crowd classification parameters provided by the first module, the user diet provided by the user and the exercise record parameters to perform health management target-muscle-increasing and weight-increasing target score calculation to obtain the user score; the third module obtains the score record of each classified user provided by the second module, obtains the muscle and weight increasing variable quantity data provided by the user, carries out self-adaptive parameter optimization, obtains the score parameter after optimization, and feeds the score parameter back to the second module so as to optimize the score of the second module, namely, optimize the score parameter as the health management parameter and the target of the user;

the sub-module 1.1 is used for acquiring personal data of a user provided by the user and collecting the user data; the sub-module obtains the following user data provided by the user: the method comprises the steps of collecting user data according to race, gender, body fat rate, disease, exercise habit/fatigue, pump feeling, pain feeling, height H, weight W, age Y, weight delta W required to be increased in one week, exercise time t, anaerobic exercise group number m, anaerobic exercise group number n, anaerobic exercise unit intensity weight omega, 4 nutrient intake, exercise type and exercise intensity data, and obtaining user data required by crowd classification mapping;

The third score super-parameter self-adaptive optimization module comprises two sub-modules, namely a user data collection sub-module 3.1 and a score super-parameter self-adaptive optimization calculation sub-module 3.2; the sub-module 3.1 is used for acquiring the muscle-increasing weight-increasing variable quantity data provided by a user, acquiring the user score provided by the second muscle-increasing weight-increasing score module and collecting the user data; the sub-module 3.2 score super-parameter self-adaptive optimization calculation sub-module obtains the user data required by score super-parameter optimization provided by the sub-module 3.1 user data collection sub-module, and performs score super-parameter self-adaptive optimization calculation to obtain the optimized comprehensive score super-parameter; the score hyper-parameter self-adaptive optimization submodule adopts a feature selection method to evaluate the relation between each input variable and the target variable to obtain the optimal weight;

2. The multi-module intelligent health management information system of claim 1, wherein the sub-module 1.2 randomly divides the crowd data into a test set and a training set; in the training set, selecting the characteristics of crowd characteristics, adopting an information gain standard, and recursively constructing a decision tree;

the information gain standard employs l (x) _i )＝-log ₂ p(x _i ) Wherein the symbol x _i Representing a sort of case; p (x) _i ) Representing a probability of selecting the category; l (x) _i ) Represents x _i Is an information amount of (a);

to calculate entropy we need to calculate the expected values of the information contained in all possible values of all classesWherein p (x) _i ) Representing the probability of selecting the class, n representing a total of n class cases;

empirical entropy ofWherein D represents a training data set, C _k Represents a certain classification situation of the training data set, K represents a total of K classification situations, |C _k I is belonging to class C _k Is the number of samples;

the conditional entropy H (y|x) represents the uncertainty of the random variable Y under the condition that the random variable X is known,wherein, when the random variable X is X _i In the case of such classification, the probability p of the classification _i ＝P(X＝x _i ) N represents a total of n classification cases;

the information gain g (D, A) =H (D) -H (D|A) of the feature A on the training data set D, wherein H (D) is the empirical entropy of the set D, and H (D|A) represents the difference of the empirical conditional entropy of D under the given condition of the feature A; and then constructing a decision tree, calculating the information gain of all possible features from a root node, selecting the feature with the largest information gain as the feature of the node, establishing child nodes by different values of the feature, and recursively calling the child nodes.

3. The multi-module intelligent health management information system of claim 1, wherein,

the sub-module 2.5 muscle and weight increasing comprehensive score calculation module obtains the motor stimulation score provided by the sub-module 2.2 motor stimulation score calculation sub-module, obtains the caloric score provided by the sub-module 2.3 caloric score calculation sub-module, obtains the diet proportion score provided by the sub-module 2.4 diet proportion score calculation sub-module, obtains the optimized comprehensive score super-parameters provided by the three-adaptive parameter optimization module, performs muscle and weight increasing comprehensive score calculation, and obtains the user score:

S _w for the comprehensive score of muscle and weight increment, the calculation formula is obtained as above; wherein, representing the weight gain and muscle gain comprehensive score coefficient vector, +.> λ ₁ Representing the motor stimulus score coefficient, lambda ₂ Representing caloric score coefficient, lambda ₃ A score coefficient representing the proportion of diets; />Comprehensive score vector for representing muscle and weight increment，/>S _a Indicating a motor stimulus score, S _b Indicating caloric score, S _c A score representing the proportion of diet; for->And continuously optimizing by a module III self-adaptive parameter optimizing module.

4. The multi-module intelligent health management information system of claim 1, wherein,

For the sub-module 2.2 motion stimulation score calculation sub-module, the crowd classification parameters provided by the module group classification module are obtained, the user motion data provided by the sub-module 2.1 user data collection sub-module is obtained, the motion stimulation score calculation is carried out, and the user motion state calculation data and the motion stimulation score are obtained;

wherein the x= (a) ₁ ，a ₂ ，a ₃ ……a _F ) For the pre-motion description vector, it is defined as each actual component of F indexes for the process a, a _f Score for pre-exercise index f, (a) _f ) _max The score is the highest score of the index f before exercise; the y= (b) ₁ ，b ₂ ，b ₃ ……b _F ) For the motion post-description vector, it is defined that the b-process faces each actual component of the f indices, b _f Score for post-exercise index f, (b) _f ) _max The score is the highest score of the index f after exercise; θ is an adjustment coefficient in motion, θmax is the maximum value of the adjustment coefficient in motion, F is 1, 2 and 3 … … F (F is the maximum value of F), and only three reference indexes are considered at the present stage; daily exercise stimulation of the athleteA number of indices are related, and the current phase analysis considers: the synergistic relation of the three reference indexes is enhanced, the foundation of fatigue is not highlighted, and the two effective time points which are beneficial to evaluation, namely before training and after training, are obtained, and the row vectors which are measured under the three indexes are respectively subjected to dot product operation to obtain a number product;

The three reference indexes refer to organism fatigue; the sense of pumping rooted in muscle memory ", excessive sense of pumping or muscle damage that will accompany delayed perception, against muscle storage; feeling of acid swelling.

5. The system of claim 1, wherein in the module 2.3 caloric score calculation sub-module, the absolute daily caloric score S in the cycle in the step _b The method is obtained by the following formula:

wherein the S is _M The actual daily heat surplus in the period is calculated; WRF is the target weighting energy; s is S _yn Daily dietary energy intake for a period; s is S _ym Energy is consumed daily for each day in a period; s is S _s Energy is consumed for daily movements during the cycle;

wherein the FIE _i Intake of energy for nutrient i in the food; k (k) _i Taking the energy coefficient of the nutrient i of unit mass as the reference human digestibility; ADI (ADI) _i Intake of nutrient i; i is 1, 2, 3 … … I (I is the maximum value of I), and only four nutrients, namely FIE, are considered at the present stage ₁ Representing carbohydrate intake energy, FIE in food ₂ Representing the energy intake of protein in food, FIE ₃ Representing fat intake energy in fat foods Quantity, FIE ₄ Represents dietary fiber intake energy in the food; ADI (ADI) ₁ ADI represents carbohydrate intake energy in food ₂ ADI for protein intake ₃ ADI for fat intake ₄ The intake of dietary fiber; k (k) ₁ In order to refer to the human digestibility, the energy coefficient of the carbohydrate in unit mass is taken, and the value is generally 4kcal/g; k (k) ₂ Taking the energy coefficient of protein of unit mass under the condition of referencing the human digestibility, and generally taking the value of 4kcal/g; k (k) ₃ Taking the energy coefficient of fat of unit mass under the condition of referencing the human digestibility, and generally taking the value of 9.6kcal/g; k (k) ₄ In order to refer to the actual digestibility of the human body, the energy coefficient of dietary fiber per unit mass is generally taken as 2kcal/g.

6. The system according to claim 1, wherein daily energy consumption S is daily over a period of time in said step _ym The method is obtained by the following formula:

wherein the DCE _j For daily activities J to consume energy, J takes 1 and 2 … … J (J is the maximum value of J), only two daily activities are considered at present, namely: DCE (DCE) ₁ Representing basal metabolic energy expenditure, DCE ₂ Representing the energy consumed by the daily activities after exercise;

wherein the BMR is a basal metabolic rate; g _r For correction coefficients, R is 1, 2 … … R (R is the maximum value of R), and only five correction coefficients are considered at the present stage, namely: g ₁ Correction coefficient for BMR race, g ₂ Correction coefficients for BMR disease;

BMR adopts Harris-Benedict formula,

BMR＝α+(β×W)+(γ×H)-(Ω×Y)

wherein W is weight, H is height, Y is age; the gender of the users is different, and the parameters alpha, beta, gamma and omega take different values; determining the sex of the user based on the user basic information, and when the user is male, α= 66.47, β=13.75, γ= 5.0033, Ω=6.775; when the user is female, α= 655.1, β= 9.563, γ=1.850, Ω= 4.676;

said g ₁ Correction factors for BMR ethnicity, g when the user is Asian ethnicity ₁ Take a value of 0.95, but g when the user is not asian ₁ The value is 1; further, the g ₂ When a user is normal and suffering from the disease, the BMR disease correction coefficient is generally valued by 1, and the rest is set according to the comprehensive evaluation of the physical condition of the user;

7. The system of claim 1, wherein daily movements in the cycle of said steps consume energy S _s The method is obtained by the following formula:

wherein the SCE _q The energy is consumed for the motion Q, Q takes 1, 2 and 3 … … Q (Q is the maximum value of Q), and three motion forms are only considered at the present stage, namely: SCE ₁ Energy is consumed for basic exercise, SCE ₂ Energy consumption for aerobic exercise, SCE ₃ Energy is consumed for anaerobic exercises;

the BMR is the basal metabolic rate; said g _r G is the correction coefficient ₁ Correction coefficient for BMR race, g ₂ Correction coefficients for BMR disease;

the METs are the physical activity intensities of different aerobic exercises, the W is the weight of a user, the t is the exercise time, the AETs are the physical activity intensities of different anaerobic exercises, the m is the number of anaerobic exercises, the n is the number of times of each anaerobic exercise, and the omega is the weight of the anaerobic exercise unit intensity;

the METs are the physical activity intensity of different aerobic exercises, and the values of the METs are different according to different aerobic exercise modes.

8. The system of claim 1, wherein the WRF is obtained by the formula:

WRF＝WWRF*△W

wherein WWRF is a unit weight gain factor corresponding to different BMI, and DeltaW is the weight required to be increased in one week; wwrf=769 Kcal/kg when the user is male and the Body Fat Rate (BFR) is greater than 25%; when the user is male, body fat rate is less than 25%, wwrf=485 Kcal/kg; when the user is female, the body fat rate is greater than 35%, wwrf=769 Kcal/kg; when the user is female, the body fat rate is less than 35%, wwrf=485 Kcal/kg.

9. The system of claim 1 wherein for sub-module 2.3 caloric score calculation sub-module obtains user diet, exercise data provided by sub-module 2.1 user data collection sub-module, obtains user exercise state data provided by sub-module 2.2 exercise stimulus score calculation sub-module, obtains crowd classification parameters provided by module one crowd classification module, performs caloric score calculation, and obtains caloric score; taking uniformity and range limitation of score distribution into consideration, carrying out normal distribution conversion on absolute scores so that the placement degree is 95%, and taking the value of the central axis as the maximum value of the absolute scores; i.e. atIn distribution, the->And μ= (S _b ) _max The method comprises the steps of carrying out a first treatment on the surface of the S after transformation _b A final caloric score;

module 2.4 diet proportion score calculation sub-Module, the absolute score of the diet proportion S per day during the period in the step _c The method is obtained by the following formula:

wherein the U is _h For nutrient h absolute score, v _h For the weight of nutrient H, H is 1, 2 and 3 … … H (H is the maximum value of H), and only three nutrients, namely U, are considered at the present stage ₁ Represents absolute score of carbon water, U ₂ Representing absolute score of protein, U ₃ Representing an absolute score of fat; v (v) ₁ Represents the weight of the carbohydrate, v ₂ Representing protein weight, v ₃ Representing fat weights;

wherein the S is _yn Total energy intake for diet; the FIE _h For the energy intake of nutrient H in food, H is 1, 2 and 3 … … H (H is the maximum value of H), and only three nutrients are considered at the present stage, namely FIE ₁ Representing carbohydrate intake energy, FIE ₂ For protein uptake energy, FIE ₃ Energy intake for fat; p (P) ₁ Represents the ratio of carbon to water, P ₂ Represents the protein ratio, P ₃ Representing fat ratio;

when the user is a healthy person, each nutrient is assigned the same weight by default, and only 3 nutrients, i.e. v, are considered at the present stage ₁ ＝v ₂ ＝v ₃ =1/3; by increasing the weight of the nutrient, a user with a high score on the nutrient will obtain a higher score in the overall diet proportion score, i.e., v ₁ 、v ₁ 、v ₁ The value is related to the chronic disease type related to diet; the handler suffers from diet-related chronic diseases, including cardiovascular disease, type 2 diabetes; absolute score U of nutrient h in daily diet during the period of the step _h The method is obtained by the following formula:

wherein the P is _h For the ratio of daily nutrient H in total energy of diet intake in period, H is 1, 2, 3 … … H (H is the maximum value of H), only three nutrients are considered at present, namely P ₁ Represents the ratio of carbon to water, P ₂ Represents the protein ratio, P ₃ Representing fat ratio; (P) _h ) _min A minimum value of the ratio of total energy taken in the diet for nutrient h; (P) _h ) _max The maximum value of the ratio of total energy taken in the diet for nutrient h.

10. The system of claim 1 wherein the sub-module 3.1 user data collection sub-module obtains the muscle and weight gain variation data Δm provided by the user and obtains the user score S provided by the module two muscle and weight gain scoring module _w Collecting user data;

the sub-module 3.2 score super-parameter self-adaptive optimization calculation sub-module obtains the user data required by score super-parameter optimization provided by the sub-module 3.1 user data collection sub-module, and performs score super-parameter self-adaptive optimization calculation to obtain the optimized comprehensive score super-parameter;

score superparameter self-adaptive optimizerThe module adopts a feature selection method to evaluate the relation between each input variable and the target variable to obtain the optimal weight; here, the pearson correlation coefficient is adopted to obtain the diet proportion score S _a Score S of exercise stimulus _b Score S of caloric content _c And the weight gain variation S of muscle and body _ΔM The initial relationship of the scores, namely:wherein, the muscle and weight increasing variable quantity score S _ΔM The calculation formula is that the submodule 3.1 user data collecting submodule obtains:

in view of uniformity and range limitation of score distribution, for S _ΔM Normal distribution transformation is carried out by the following steps of _ΔM Normal distribution transformation is carried out to ensure that the placement degree is 95 percent, and the value of the central axis is (S _ΔM ) _max The method comprises the steps of carrying out a first treatment on the surface of the I.e. atIn the distribution of the particles, and μ= (S _ΔM ) _max The method comprises the steps of carrying out a first treatment on the surface of the S after transformation _ΔM Scoring the final diet proportion;

the initial determination mode of the method is that the specific gravity of each correlation coefficient to the total correlation coefficient sum is calculated as follows:

as user data continues to increase, the communication is continuedOver-comparing the error between the true value and the measured value and taking it as an optimization parameterFor parameter optimization, the error calculation formula is:

wherein, is a super parameter vector->Lower S _ΔM And S is equal to _w Error of (2); randomly dividing a training set and a testing set, and continuously updating parameters by using a gradient descent algorithm; super parameter vector->Updating the formula to

Wherein, is the gradient of the parameter, η is the learning rate; when->During minimization, the muscle and weight increasing comprehensive score superparameter vector ++>Is a global optimum.