CN104200076A - Athlete athletic injury risk early warning method - Google Patents

Abstract

The invention discloses an athlete athletic injury risk early warning method. On the bases of reference to models provided by foreign scholars and comprehensive analysis of track and field athlete athletic injury factors, a track and field athlete motion injury risk early warning injury factor dynamic chain model is provided, based on the model, corresponding factors are selected from an athlete risk early warning database, and a track and field athlete motion injury risk early warning injury factor dynamic chain quantitative model is established through the analytic hierarchy process. Index data are discretized according to an SOM neutral network discretization method, a decision table is simplified according to a method based on a discernibility matrix in a rough set, an RBF neutral network is established based on the simplified decision table, the RBF neutral network is trained, and finally a correct diagnosis result is acquired. By the adoption of the method, accurate early warning can be conducted on athletic injuries, the athlete athletic injury risk grade is effectively predicted, and treatment and prevention of the athletic injuries are facilitated.

Description

A kind of Athletes damage method for prewarning risk

Technical field

The present invention relates to a kind of Athletes damage method for prewarning risk based on Rough Sets Neural Networks.

Background technology

In track and field sports, the factor relevant to injury gained in sports is a lot, there are inner Risk Factors (age, sex, damage history etc.), outside Risk Factors (place, apparatus, weather etc.), bring out incentive condition (training burden, match density etc.), relation how to find out they and injury gained in sports risk in so many factor is a large order.And in these factors the existing index that can accurately measure as physiological and biochemical index, have again some qualitatively index as medical history, physical characteristic etc., and the index that each sportsman surveys in passing data is not quite similar again, therefore will carry out comprehensive assessment to injury gained in sports risk is a very difficult job.Radial basis function (RBF:Radial Basis Function) Learning Algorithm, because of its good non-linear mapping capability and unique network structure, the every field such as pattern-recognition, data prediction, system identification, image processing, speech understanding and Function Fitting have now been widely used in.But, utilize neural network to carry out injury gained in sports risk assessment, there is the problems such as selected characteristic parameter is too much, scale of neural network is too huge, learning training time overlength, badly influence practical effect and the real-time of this model.This is the deficiencies in the prior art part.

Summary of the invention

The technical problem to be solved in the present invention is to provide a kind of Athletes damage method for prewarning risk, and effectively predicted motion person's injury gained in sports risk class, is beneficial to treatment and the prevention of injury gained in sports.

The present invention adopts following technological means to realize goal of the invention:

An Athletes damage method for prewarning risk, is characterized in that, comprises the steps:

(1) analysis of the injury gained in sports Risk-warning factor;

(2) the sports injury muscle injury Risk-warning factor dynamic chain model of causing injury is proposed;

(3) set up sportsman's Risk-warning database;

(4) use analytical hierarchy process, set up the sports injury muscle injury Risk-warning factor dynamic chain quantitative model of causing injury;

(5) according to sports events, choose sportsman and the Risk-warning factor, utilize the corresponding data in exercise risk database, set up exercise risk early warning Sample Storehouse and the test sample book data of this project;

(6) discreteness of the Risk-warning factor corresponding data in judgement exercise risk early warning Sample Storehouse, if not discrete, call SOM neural network discrete method data is carried out to discretize processing;

(7) by the synthetic decision table of the achievement data after utilizing SOM network discrete and the initial data with regard to discrete valuation;

(8) according to the yojan process of RS theory, decision table is carried out to attribute reduction and Rule Extraction, obtain one group of yojan result;

(9) decision table based on simplifying builds RBF neural network;

(10) according to the decision table after simplifying, choose the corresponding data in exercise risk early warning Sample Storehouse, Training RBF Neural Network;

(11) judge whether yojan result restrains, and whether meets accuracy requirement, if do not meet, modified RBF network structure be connected weights, go to step (10);

(12) utilize test sample book data, whether the RBF network that training of judgement is good obtains correct diagnostic result, if do not have, selects other yojan result from step (8), goes to step (9);

(13) preserve the RBF network training, utilize the new sportsman's risk assessment index detecting of RBF neural network assessment training.

As the further restriction to the technical program, the SOM neural network discrete method of described step (6) comprises the steps:

(6.1) calculate the importance degree of each connection attribute in decision table and sort by order from big to small;

(6.2) each continuous injury gained in sports factor attribute a that causes injury _i(i=1,2 ..., | a C|) corresponding SOM network, below with an attribute discrete, turn to example explanation, the initial value of given SOM network output layer neuron number is 3, input layer number is 1, by the connection weights wj of network, gives [Min (a _i), Max (a _i)] random value in interval, determining the initial value α (0) of learning rate α (t), α (t) should slowly reduce with the increase of t, total the study number of times T determining;

(6.3) by attribute a _iin numerical value x be input to som network;

(6.4) calculate input numerical value x and the distance being connected between weights wj;

D _j=(x-w _j) ², j=1,2,3 (formula 1)

(6.5) choose and make the minimum neuron k of distance as triumph node;

(6.6) the connection weights of triumph set of node are upgraded

W _j(t+1)=w (t)-α (t) (x-w _j) (formula 2)

(6.7) return to step (3), choose another data input network, until all mode in sample is offered to network;

(6.8) renewal learning rate:

$\mathrm{\α}\left(t\right)=\mathrm{\α}\left(0\right)[1-\frac{t}{T}];$ (formula 3)

(6.9) make t=t+1, return to step (6.2), be repeated to t=T;

(6.10) incompatibility of decision table after calculating discretize, calculates end if meet the requirement of incompatibility threshold values, otherwise turns next step;

(6.11) select the attribute of importance sorting first, its SOM network output layer neuron number is adjusted into 4, the SOM network output layer neuron number that all the other attributes are corresponding is constant, again discretize decision table, and go to step (6.10), until the maximum cluster numbers of each SOM network is adjusted into 4, the maximum cluster numbers of each SOM network is defined as 4, if its SOM network clustering number of the attribute of importance degree maximum has reached requirement, select the attribute that importance degree takes second place to adjust, by that analogy;

(6.12) complete final discretize operation.

As the further restriction to the technical program, described step (8) adopts the clearly demarcated matrix method based in rough set to carry out Decision Table Reduction.

As the further restriction to the technical program, the described clearly demarcated matrix method based in rough set comprises the steps:

(8.1) read in exercise risk early warning Sample Storehouse data, as decision table;

(8.2) calculate the clearly demarcated Matrix C of this decision table _d;

(8.3) in clearly demarcated matrix, all values are non-zero, 1, element c _ij, set up corresponding logical expression L _ij, wherein

(formula 4)

(8.4) by all expression formula L that extracts _ijdo conjunction computing, obtain a conjunctive normal form L,

(formula 5)

(8.5) form that above-mentioned conjunctive normal form is converted to disjunctive normal form is

(8.6) output attribute yojan result, the corresponding attribute reduction result of each conjunct in disjunctive normal form, the conditional attribute set of the set of properties comprising in each conjunct after into about letter.

As the further restriction to the technical program, described step (9) comprises the steps:

(9.1) choose basis function

Select Gaussian function as the activation function of hidden layer unit, suppose that input vector is n dimension, have m output node, the number of hidden layer node is k, and the output of hidden node is expressed as:

$R_j\left(x\right)=e^{-\frac{{\left|\right|x-c_j\left|\right|}^2}{{2\mathrm{\σ}}_j^2}}$ (formula 6)

J={1 wherein, 2 ..., k}, Rj (x) be j hidden node to output, the n dimension input vector that x is network, c _jbe the kernel function center vector of j hidden layer node, it has identical dimension, σ with input vector x _jbe the width of j hidden layer node Gaussian function, || x-c _j|| represent the distance measure between input vector and central point.Network output is defined as:

$y_i=\underset{j=1}{\overset{k}{\mathrm{\Σ}}}{w}_{\mathrm{ij}}{R}_j\left(x\right)$ (formula 7)

I=1 wherein, 2 ..., m; Yi is the output of i output layer node, w _ijbe that j hidden layer node is to the connection weights of i output layer node.In this project, only need 1 output, i.e. i=1,

(9.2) hidden layer design

The hidden layer of design RBF network, mainly to determine hidden layer number of unit and their excitation function, adopt Gaussian function, Gaussian function is determined by cluster centre and cluster width, the classification of training mode is known, it is injury gained in sports Risk-warning grade, therefore can adopt a kind of simple mode to determine, be totally three of the corresponding Gaussian functions of each class, be k=3, the width parameter of Gaussian function can be got the average of three class samples and its central point distance, and the average that three class samples can be got respectively in the center of Gaussian function is as the central point of unit separately;

(9.3) radial basis function center, width and to the renewal of the weights between output layer

Radial basis function center and other parameter all experience learning process, conventionally adopt error correction learning process, and application gradient descent method is specific as follows:

Suppose to have N sample input, to all input samples, definition error function:

$\mathrm{\ξ}=\frac{1}{2}\underset{q=1}{\overset{N}{\mathrm{\Σ}}}{e}_q^2$ (formula 8)

E wherein _qfor error, be defined as follows:

$e_q=d_q-y\left(x_q\right)=d_q-\underset{j=1}{\overset{3}{\mathrm{\Σ}}}{w}_{1j}{R}_j\left(x_q\right)=d_q-\underset{j=1}{\overset{3}{\mathrm{\Σ}}}{w}_{1j}{e}^{-\frac{{\left|\right|x_q-x_j\left|\right|}^2}{{2\mathrm{\σ}}_j^2}}$ (formula 9)

D wherein _qfor sample x _qthe value of required type, as the d corresponding to sample of high risk class in this project _q=3.

As the further restriction to the technical program, described step (10) comprises the steps:

Step 1: determine hidden layer node number according to known class number, k=3 in this project, asks for the mean value of all input vectors of this classification as such center initial value c _j(1), the width initial value σ using the mean value of all input vectors of every class and centre distance as such _j(1), a sample of each classification input, asks for initial w _1j(1), set allowable error ε, ε is an a small amount of setting learn efficiency η that is greater than zero ₁=η ₂=η ₃=1/t, t is iterative loop variable, sets loop variable t and is initially 1, maximal value is MaxT;

Step 2: input t training sample, obtain the actual output of network y (t);

Step 3: calculate the error ξ (t) between actual output and desired output, if error ξ (t) < is ε, this sample does not need to adjust network parameter, jumps to step 6, otherwise enters next step;

Step 4: press formula $\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;w}_{1j}\left(n\right)}=-\underset{q=1}{\overset{n}{\mathrm{\&Sigma;}}}{e}_q\left(n\right)R_j\left(x_q\right)=-\underset{q=1}{\overset{n}{\mathrm{\&Sigma;}}}{e}_q{e}^{-\frac{{\left|\right|x_0-c_1\left|\right|}^2}{{2\mathrm{\&sigma;}}_1^2}}---\left(29\right)$

formula $\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;c}_j\left(n\right)}=-\underset{q=1}{\overset{n}{\mathrm{\&Sigma;}}}{e}_q\left(n\right)R_j\left(x_q\right)=-\underset{q=1}{\overset{n}{\mathrm{\&Sigma;}}}{e}_q\frac{w_{1j}\left(n\right)}{{\mathrm{\&sigma;}}_j^2}{e}^{-\frac{{\left|\right|x_0-c_1\left(n\right)\left|\right|}^2}{{2\mathrm{\&sigma;}}_1^2}}(x_q-c_j\left(n\right))---\left(31\right)$

formula $\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;\mathrm{\&sigma;}}_j\left(n\right)}=-\underset{q=1}{\overset{n}{\mathrm{\&Sigma;}}}{e}_q\left(n\right)\frac{w_{1j}\left(n\right)}{{\mathrm{\&sigma;}}_j^3\left(n\right)}{\left|\right|x_q-c_j\left|\right|}^2{R}_j\left(x_q\right)---\left(33\right)$

Calculate and press formula 29

$w_{1j}(n+1)=w_{1j}\left(n\right)-{\mathrm{\&eta;}}_1\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;w}_{1j}}---\left(30\right)$

formula $c_j(n+1)=c_j\left(n\right)-{\mathrm{\&eta;}}_2\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;c}_j}---\left(32\right)$

formula ${\mathrm{\&sigma;}}_j(n+1)={\mathrm{\&sigma;}}_j\left(n\right)-{\mathrm{\&eta;}}_3\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;\mathrm{\&sigma;}}_j\left(n\right)}---\left(34\right)$ Calculate w _1j(t+1), c _j(t+1), σ _j(t+1);

Step 5: upgrade new network parameter w _1j(t+1), c _j(t+1), σ _j(t+1), t=t+1; If t>MaxT, shows and do not restrain, enter step 6, otherwise forward step 2 to;

Step 6: whole learning process finishes, preserves each parameter of current network.

Compared with prior art, advantage of the present invention and good effect are: SOM neural network discrete method of the present invention carries out discretize processing to achievement data; The clearly demarcated matrix method of employing based in rough set carries out Decision Table Reduction; Decision table based on simplifying builds RBF neural network; Training RBF Neural Network, finally obtains correct diagnostic result.This method can to injury gained in sports make the effective predicted motion person's injury gained in sports of early warning comparatively accurately risk class, be beneficial to treatment and the prevention of injury gained in sports.

Accompanying drawing explanation

Fig. 1 is the injury gained in sports cause of disease dynamic chain model that Meeuwisse (1994) sets up.

Fig. 2 is Sample Storehouse collecting flowchart figure of the present invention.

Fig. 3 is physiological and biochemical index information acquisition FB(flow block) of the present invention.

Fig. 4 is the potential factor information collecting flowchart block diagram of causing injury of discrete motion of the present invention.

Fig. 5 is system master data fill flow path figure of the present invention

Fig. 6 is system day regular data fill flow path figure of the present invention

Fig. 7 is sports injury muscle injury Risk-warning cause of disease dynamic chain model of the present invention.

Fig. 8 is the athlete of the present invention factor rating figure that causes injury.

Fig. 9 is the process flow diagram of the preferred embodiment of the present invention.

Figure 10 is neural network topology structure figure of the present invention.

Figure 11 is the track and field team of the present invention injury gained in sports risk factors data acquisition table of causing injury.

Figure 12 is sports injury muscle injury Risk Factors risk assessment table of the present invention.

Embodiment

Below in conjunction with accompanying drawing and preferred embodiment, the present invention is further described in detail.

One, the analysis of injury gained in sports Risk-warning correlation factor

A. the inside factor of causing injury

(1) age

Table-1 athlete's Age Indices damages the risk assessment of causing injury

(2) sex

Table-2 track and field saltaroria sports events sex indexs are damaged the risk assessment of causing injury

(3) menstrual cycle

Table-3 athlete's menstrual cycle indexs are damaged the risk assessment of causing injury

(4) physical training amount (sport rank)

The risk assessment of causing injury of table-4 athlete's sport ranks (physical training amount)

(5) damage history

Table-5 athletes damage history index and damage the risk assessment of causing injury

(6) injury recovery situation

Table-6 athlete's injury recovery status indexes damage the risk assessment of causing injury

(7) body shape (BMI index)

Table-7 athlete BMI indexs are damaged the risk assessment of causing injury

(8) range of motion (ROM)

Table-8 athlete's range of motion (ROM) indexs are damaged the risk assessment of causing injury

(9) stability of joint (arthrochalasis degree)

Table-9 athlete's stabilities of joint (arthrochalasis degree) index is damaged the risk assessment of causing injury

(10) muscle strength and harmony

The harmonious index of table-10 athlete's muscle strengths is damaged the risk assessment of causing injury

(11) balance ability

Table-11 athlete's balance capacity index injurys gained in sports risk assessment of causing injury

(12) health anatomical structure is abnormal

Table-12 athlete's health anatomical structure abnormal index injurys gained in sports risk assessment of causing injury

B. the outside factor of causing injury

(1) training, contest situation

Table-13 training athletes, the injury gained in sports of the contest status index risk assessment of causing injury

(2) technical merit

Table-14 athlete's technical merit index injurys gained in sports risk assessment of causing injury

(3) safeguard measure

The injury gained in sports risk assessment of causing injury of table-15 athlete's safeguard measures (injury protection utensil)

(4) sport footwear

The damage risk assessment of causing injury of table-16 track and field running type games project sport footwears

(5) athletic ground

Table-17 athlete athletic ground index injurys gained in sports risk assessment of causing injury

(6) weather conditions (climate condition)

The cause injury assessment of risk of table-18 athlete's weather conditions (climate condition) injurys gained in sports

C. the Induced by Stimulation factor

(1) error of performance

Table-19 athlete's error of performance index injurys gained in sports risk assessment of causing injury

(2) training, competition schedule is unreasonable

Table-20 athlete's exercise load figureofmerit injurys gained in sports risk assessment of causing injury

Table-21 athlete's training and competition plan index injurys gained in sports risk assessment of causing injury

(3) physical function is in poor shape

Table-22 athlete's physical function status index injurys gained in sports risk assessment of causing injury

(4) psychological factor

Table-23 athlete's psychological factor index injurys gained in sports risk assessment of causing injury

(5) with other people or apparatus accident collision

In athletic training and match, due to the fierce contact-impact of sportsman between mutually or with the accidental cause such as apparatus collision, cause injured possibility comparatively rare, may occur in middle-distance race, marathon etc. and because sportsman squeezes mutually to touch in order to fight for vantage point, cause damage when the match start of a race or intermediate run or spurt, also may occur in hurdler and cause with the contact-impact of hurdle frame in training and competition.In a word, this class accidental cause is comparatively rare in track and field event, once but occur, just probably cause injured, therefore, with other people or apparatus accident collision be the risk factors that cause athlete to damage, concrete evaluation can reference table-24.

Table-24 athletes and other people or the injury gained in sports of the apparatus accident collision factor risk assessment of causing injury

Two, the foundation of injury gained in sports Risk-warning cause of disease dynamic chain model

A. track and field event spy sends out breakdown diagnosis

By inquiry lot of documents, collect the data of the common damage of relevant track and field projects, by projects category induction, be summarized as follows.

(1) short in class

1) dash class

The special breakdown diagnosis of sending out of table-25 sprint eventses

2) class of hurdling

Table-26 special breakdown diagnosis of sending out of project of hurdling

(2) middle-distance race intermediate item

The special breakdown diagnosis of sending out of table-27 middle-distance race intermediate item

(3) saltaroria project

The special breakdown diagnosis of sending out of table-28 high jump projects

The special breakdown diagnosis of sending out of table-29 long-jump projects

(4) throw intermediate item

1) javelin

The special breakdown diagnosis of sending out of table-30 javelin projects

2) discus

The special breakdown diagnosis of sending out of table-31 discus projects

3) hammer

The special breakdown diagnosis of sending out of table-32 hammer projects

4) shot

The special breakdown diagnosis of sending out of table-33 shot projects

B. potential sick and wounded generation sign is analyzed

By analysis, cause the reason of various different motion damages to find, athletes ' functions ability declines, and produce exercise induced fatigue and fatigue and fail to recover in time, be the key factor of bringing out damage.In Xu Yong Deng Dui China long jumper injury gained in sports epidemiological study, find, long jumper when training different somatosensory with the direct relation that has of injury gained in sports, while having 40.3% damage to occur, somatosensory is " fatigue ".That is to say if body occurs fatigue or recover not good general meaning the large increase of having a big risk of damage.Therefore, can or recover the sign that bad appearance occurs as latent lesion using exercise induced fatigue.In addition, early stage in some sick and wounded appearance, have some specific signs, if can find in time, carry out early warning, take corresponding reduction to cause injury measure that the factor occurs, just can avoid sick and wounded generation or further develop, thereby avoids or reduce sick and wounded to athletes ' body and the adverse effect of normally training.

(1) exercise induced fatigue and recover bad diagnostic factro

The conventional physiological fatigue diagnostic factro of β 1

The conventional physiological fatigue diagnostic factro A of table-34

The conventional physiological fatigue diagnostic factro B of table-35

The conventional mental fatigue diagnostic factro of β 2:

Motility mental fatigue is that sportsman is when reply endogenous pressure and exogenous pressure, Psychological Resources and physiology resource are constantly consumed and the mental function that occurred while not obtaining supplementing in time can not maintain the phenomenon that original psychological activity level mental function declines, and are in particular in the change of emotional dimension, cognitive dimension, power dimension, behavior dimension and physiology dimension.Motility mental fatigue has passive impact to athletic mental health state, physiological fatigue and damage etc., therefore motility mental fatigue is monitored and is just seemed especially important.At present, the monitoring of motility mental fatigue is mainly contained to following methods:

The conventional mental fatigue diagnostic factro of table-36

(2) potential sick and wounded omen sign (feature)

Table-37 potential sick and wounded omen signs (feature)

(3) sportsman's self-monitoring

Sportsman is by self-monitoring, can understand in time some body abnormalities and change, and doctor, trainer coordinate, adjust in time training and competition plan, contribute to prevent over training and wound accident, also can make sportsman form the custom of noting observing training philosophy and personal hygiene.The content that oneself detects comprises subjective sensation and objective material.Subjective sensation refers to general sensation, sleep, appetite, sport mood etc.Objective material refers to pulse, body weight, vital capacity, grip, back strength, perspiration situation etc.These materials are wanted simultaneously and training content, match or performance of the test combine and analyze.

It is energetic that sportsman feels normally, happiness, and training enthusiasm is high, and after exercise, the slightly tired and tired sense of muscle acid, recovers after rest very soon.As feel lassitude, and unable, sleepy, dizziness, passionnate, the aching and limp numbness of local articular muscle, pain even, chest is felt oppressed, breathes hard, abdominal distension, stomachache etc., are all abnormal occurrencies.Sleep is physiological requirements, can make people fully be had a rest and regain one's strength.Often insomnia, easily forms overtired.Appetite is all good under normal training, but in over training, overstretched or can reduce while there is some morbid state.After a hard training or the appetite decline phenomenon at once occurring after nervous match, be temporary transient, can recover very soon normal.Sport mood refers to that sportsman participates in the hope of training and competition.Train when good, in high spirits, excitability is higher, full of physical strength, and sports achievement is generally also better.When physical function is not good, can occur training indifference, even place, apparatus are weary of etc. to situation.Pulse generally records from radial artery, also can record by arteria carotis, should note pulse frequency, rhythmicity, the variation such as strong and weak of beating.When quiet, pulse may be slack-off with the raising of training degree, as unexpected appearance acceleration or excessively slow, will find reason.Sometimes can there is irregularity of pulse with breathing in pulse, be normal phenomenon; As occur obviously irregularly, should notify doctor at once.Body weight is general in group-up player keeps constant, can slightly alleviate, but may go up soon while just participating in training.As body weight continuous decrease, should note whether having certain deeline or seriously overtired.After a hard training, body weight may decline 1～4 kilogram, but after 1～2 day, recovers again normal.Grip, arm strength all increase when training is good, and continuous decrease is abnormal.How much perspire because of temperature, amount of drinking water, the level of training different and different with personal touch.Sportswoman also will record menstruation.

Physical function is easy to cause damaging the increase of risk while declining, this tells about above, therefore by athletic self-supervision, timely and effectively athletic health is fed back, and will contribute to reduce the generation of sports wound.At this, sportsman's self-monitoring (sportsman of mainly take feels as main) to be classified as to evaluation and damage the risk factors of causing injury, concrete evaluation can reference table-38.

Table-38 athlete's self-supervision indexs are damaged the risk assessment of causing injury

(4) trainer's monitoring

As trainer, in arranging the process of athletic training and competition, not only to understand the implementation status of training athlete and match, also should be able to judge athletic physical function situation.Can not only when appearring in sportsman, technical movements mistake correct in time, can also in the athletic words and deeds of observation, judge sportsman by the experience of self and whether occur fatigue and tired degree, so just can better grasp athletic body building situation and avoid sportsman to occur sick and wounded.In addition, outstanding trainer also should be able to understand or grasp the body function monitor indexs such as some conventional Physiology and biochemistries, could observe in conjunction with self like this and objective indicator is reasonably adjusted sport training plan and amount of exercise, can reach and improve locomitivity and achievement, can play the object of avoiding exercise injury and disease again.

Trainer's monitoring can comprise the body action situation of observing in athlete's life and training, exchange and to commonly using the judgement etc. of physiological and biochemical index with athletic, therefrom draw and self to the appearance of sportsman's possibility is sick and wounded, make corresponding judgement, and show that sportsman damages the height of the risk of causing injury (reference table-39).

Table-39 track and field monitoring moving persons damage the risk assessment of causing injury

C. model is set up

On the multifactor dynamic chain of Meeuwisse injury gained in sports cause of disease basis, we attempt by injury gained in sports risk feedback fusion wherein, have set up the injury gained in sports Risk-warning cause of disease dynamic chain model for athlete, as Fig. 7.

In this dynamic chain, more to the right, the possibility that damage occurs is larger, and more to the left, the possibility that damage occurs is less.When the category-A factor (inside cause injury the factor) occurs, sportsman becomes " having damage tendency crowd "; Now, when the different B class factor (outside cause injury the factor) occurs, sportsman becomes " crowd is easily sent out in damage ", and the risk that certain class specific Damage occurs the sportsman of track and field disparity items will increase greatly.Now should carry out early warning for the first time, by reduction, damage the measure of risk, the state of sportsman in dynamic chain is moved to the left, return to " having damage tendency crowd " stage, its danger that damage occurs is reduced.If sportsman belongs to " crowd is easily sent out in damage ", the appearance of the C class factor (bringing out stimulating factor), further increases the risk that makes sportsman face damage.Now sportsman often may show the sick and wounded sign that will occur, feature, now should carry out early warning for the second time, removes and causes the sick and wounded all kinds of risk factors of causing injury, and the position of sportsman in dynamic chain moved to left, and avoids the generation of its damage; Otherwise sportsman will there will be sick and wounded.

Three, set up sportsman's Risk-warning database

A. injury gained in sports risk data gathers and pre-service

After the analysis of injury gained in sports Risk factors completes, can set about carrying out the foundation of injury gained in sports risk warning model, modeling method both can adopt dynamic chain model qualitatively, also can adopt quantitative model.In quantitative model, can choose traditional linear model, also can adopt non-linear modeling method based on this, as rough set or neural network etc.No matter adopt which kind of modeling method, first all will carry out the Data Collection work of sample database.The collection of injury gained in sports risk data mainly comprises the potential injury gained in sports obtaining by the injury gained in sports analysis factor of causing injury, and according to the value of the factor, can be divided into two classes: 1) the continuous data factor, is used for the conventional physiological and biochemical index of monitoring moving person's fatigue; 2) the discrete data factor, mainly comprises the information such as sports events, sport rank, stability of joint.The flow process of data acquisition, as shown in Figure 2.

Data acquisition and pretreated concrete steps are as follows:

Step 1: the measurement of physiological and biochemical index;

Step 2: the processing of physiological and biochemical index data;

Step 3: sports injury muscle injury cause injury establishment and the data acquisition of risk factors data acquisition table;

Step 4: the sports injury muscle injury risk factors data acquisition table data statistics of causing injury;

A.1. forecasting fatigue relevant physiological biochemical indicator gathers and pre-service

(1) measurement of forecasting fatigue relevant physiological biochemical indicator

1) routine blood test:

Routine blood test detects and comprises the contents such as RBC number, haemoglobin, hematocrit, leucocyte.

Play on every Mondays test, extracting vein blood morning.

Testing tool is that German BAYER Automatic Blood Cell Analyzer is tested.

2) serum testosterone, cortisol, serum testosterone (T)/serum cortisol (C) ratio:

Serum testosterone, cortisol are according to getting a venous blood test the every 2-4 of project characteristic week, and it is 1-2 week that pre-games generally increases test frequency.Serum testosterone (T)/serum cortisol (C) ratio generates index for it.

Testing tool is that German BAYER automatic clinical chemistry analyzer carries out.

3) creatine kinase and urea nitrogen:

Rise morning on every Mondays and emphasis training session after get peripheral blood.

Testing tool is German EPPENDORF semi-automatic biochemical analyzer.

(2) processing of physiological and biochemical index data

Collection completes after the physiological and biochemical index value that track and field team measures weekly, for the ease of final assessment Risk-warning grade, by physiological and biochemical index discretize, then comprehensively becomes the fatigue monitoring factor for dynamic chain model the standard according in table-40.

Table-40 physiological and biochemical index grade scales

In physiological and biochemical index discretize process, owing to there being larger difference between the physiological and biochemical index of Different Individual, can not go to weigh Different Individual by unified standard, in order to facilitate computerized algorithm, the physiological and biochemical index of Different Individual is compared, and therefrom excavate and injury gained in sports risk between incidence relation, therefore when submitting sample data to, must complete normalized to physiological and biochemical index.Concrete grammar be Physiological And Biochemical Values computation of mean values that sportsman was measured in the different periods of cycle of training as its individual base value, by single measurement data in contrast, using ratio as this achievement data input value.

A.2. cause injury Qualitative factor data acquisition and pre-service of potential injury gained in sports

(1) the sports injury muscle injury establishment of risk factors data acquisition table of causing injury

For setting up injury gained in sports Risk-warning cause of disease dynamic chain quantitative model, and the athlete sick and wounded early warning system of research and development based on data mining technology and mobile computing technology, consulting a large amount of domestic and foreign literatures, and with expert and the special trainer of Shandong Province track and field team of medicine in field of sports medicine, sportsman carries out interview, understand incidence and the common traumatogenic reason of relevant track and field event injury gained in sports, and on the basis that the factor of causing injury conducts in-depth analysis and system is sorted out, we have drafted " sports injury muscle injury cause injury risk factors data acquisition table ", collection for cause injury factor information and sick and wounded library information.This data acquisition table is within cycle of training, and every circumferentially sportsman provides once, and sportsman fills according to self actual conditions.

(2) the sports injury muscle injury data statistics of risk factors data acquisition table of causing injury

First in will " sports injury muscle injury cause injury risk factors data acquisition table ", the Factor Selection relevant to injury gained in sports risk out, and carries out assignment according to it to the size of injury gained in sports risk contribute by option corresponding to this factor.The 3 option factors of take are example, the option grade assignment 1 to injury gained in sports risk contribute minimum, the assignment 3 of maximum option.In some specific factor, for example, damage history, if certain sportsman does not have previously damage, this factor assignment 0.On this basis, the data acquisition table of with criteria table, sportsman being filled in is added up.

B. the sick and wounded early warning system data of athlete typing

The data general data that the sick and wounded early warning system of athlete is relevant to injury gained in sports early warning is distributed in four databases, comprise sportsman's basic information database, training monitoring management subsystem database, injury gained in sports Risk-warning subsystem database and individual sick and wounded database, according to the stability of data, the filling in of whole database data be divided into fill in each cycle of training, fill in weekly, fill in every day and irregularly upgrade four classes.Wherein filling according to sports injury muscle injury Warning System master data fill flow path block diagram of first two data carried out, filling according to sports injury muscle injury Warning System daily workout data fill flow path block diagram of latter two data carried out, and refers to Fig. 5, Fig. 6.In each database, contained data are as shown in table-41.

The sick and wounded early warning system sportsman's data of table-41 athlete input database

Three, use analytical hierarchy process, set up sports injury muscle injury Risk-warning and hinder the sub-dynamic chain quantitative model of reason

Although qualitative dynamic chain model can prevent the generation of injury gained in sports to a certain extent, the factor that affects early warning judgement is too much, and standard is comparatively fuzzy, practical operation difficulty.Therefore this project has been carried out the structure of dynamic chain quantitative model on this basis.In documents and materials previously, have no the contrast of importance between different Risk Factors, the foundation of this specified rate dynamic chain model has brought difficulty.The U.S. plans strategies for scholar A.L.saaty in analytical hierarchy process (the Analytic Hierarchy Process of 20 century 70s propositions, be called for short AHP method), be a kind of stratification, the structured decision method that many index systems of scheme are analyzed, it is policy-making thought process modelling, the quantification to complication system by decision maker.Should be in this way, decision maker is by being decomposed into challenge some levels and some factors, carries out simply relatively and calculate between each factor, just can draw the weight of different factors, for the selection of preferred plan provides foundation.The method can effectively solve in quantitative dynamic chain model process of establishing, the problem of different motion damage Risk Factors weight.

Because the injury gained in sports risks and assumptions of causing injury is a lot, for the ease of analyzing, this project is divided into three levels by the injury gained in sports factor weight analysis problem of causing injury.The superiors are destination layer, are injury gained in sports Risk-warning; Middle layer is indicator layer, refers to by interview and survey to sportsman, trainer and sports medical science expert three classifications that injury gained in sports Risk Factors are divided; Orlop is that object layer is concrete injury gained in sports Risk Factors.In order to set up indicator layer, injury gained in sports Risk Factors to be classified, this project has been made " athlete's injury gained in sports factor risk factors assessment table ".By being distributed to 15 senior above track and fields and team doctor, fill in result statistics following (take middle-distance runner as example).

The α class factor (the most easily cause damage factor) of causing injury: comprise damage history, injury recovery situation, technical movements are wrong, stability of joint is poor, training and competition load is misorient, psychology is not in good state etc. in training and competition.

The β class factor (more easily cause damage factor) of causing injury: comprise muscle strength and harmony, balance ability, joint motion amplitude (pliability), have or not safeguard measure, health anatomical structure have or not extremely, and other people or apparatus have collisionless etc.

The γ class factor (cause injury impact less factor) of causing injury: comprise age, sex, menstrual cycle, sport rank, athletic ground, weather conditions, sport footwear, training or competition status etc.

Cause injury after the factor having determined three classes, we adopt analytical hierarchy process to carry out weight analysis to it, with draw α, β, this three class of γ cause injury the factor on cause injury impact weight, and adopt each the sub-factor that three classes are caused injury in the factor that uses the same method and carry out weight analysis, draw respectively the weight size of each sub-factor.Circular is as follows.

1) set up hierarchy Model.The foundation of hierarchical structure can be problem methodization, stratification, the factor that causes damage is built to the three classes hierarchical model of causing injury, and the individual factor of causing injury is encoded.

2) Judgement Matricies A, compares between two for each element at the same level in hierarchy Model, and its criterion is as table 42.

$A=\left[\begin{array}{cccc}{a}_{11}& a_{12}& ...& a_{1n}\\ a_{21}& a_{22}& ...& a_{2n}\\ ...& ...& ...& ...\\ a_{n1}& a_{n2}& ...& a_{\mathrm{nn}}\end{array}\right]$ (formula 10)

Table 42 judgment matrix scale and implication thereof

3) calculate each judgment matrix weight, and do consistency check.

The first step, by each column element normalization of judgment matrix, obtains the matrix after normalization

${\stackrel{\&OverBar;}{a}}_{\mathrm{ij}}=\frac{a_{\mathrm{ij}}}{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{a}_{\mathrm{ij}}}(i,j=\mathrm{1,2},...,n)$

(formula 11)

Second step, is added the judgment matrix after normalization by row, obtain row vector

${\stackrel{\&OverBar;}{W}}_i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{\stackrel{\&OverBar;}{a}}_{\mathrm{ij}}(i=\mathrm{1,2},...,n)$ (formula 12)

The 3rd step, by row vector normalization, obtains characteristic vector W, is the weight of the factor of respectively causing injury

$W_i=\frac{{\stackrel{\&OverBar;}{W}}_i}{\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{\stackrel{\&OverBar;}{W}}_j}(i=\mathrm{1,2},...,n)$ (formula 13)

The 4th step, the maximum characteristic root of calculating judgment matrix

${\mathrm{\&lambda;}}_{\max }=\frac{1}{n}\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}\frac{\left({\mathrm{AW}}_i\right)}{W_i}$ (formula 14)

The 5th step, carry out consistency check:

$\mathrm{CI}=\frac{{\mathrm{\&lambda;}}_{\max }-n}{n-1}$ (formula 15)

Looking into exponent number is the judgment matrix mean random coincident indicator RI of n, in Table 43.

Table 43RI value table

And calculate CR=CI/RI, when CR≤0.1, consistance that it is generally acknowledged judgment matrix is acceptable; Otherwise, defective.

Concrete assessment result reference table 44.

The cause injury total weight assessment of the factor of table 44 α, β, γ tri-classes

As shown in Table 44, cause injury in three class factors of risk affecting athlete's injury gained in sports, the α class factor is the most remarkable, and next is respectively β class and the γ class factor of causing injury, and α, β, the shared weight proportion of the γ tri-class factor are respectively 72.4%, 19.3%, 8.3%.This shows, once α class is caused injury, the various factors of the factor occurs, to causing sportsman to damage having the greatest impact of generation, this is also that sportsman and trainer need the major risk factors of evading and reducing in peacetime training and match.

The cause injury right assessment of factor various factors inside of table-45 α classes

From table-45, in the weight analysis assessment result of inner each Risk Factors of the factor that α class is caused injury, damage history (34.1%) is the most obvious to sportsman's risk of causing injury, and next is respectively error of performance (19.8%), training and competition misorient (15.3%), injury recovery situation (12.5%), psychological factor (5.3%) and stability of joint (4.1%) etc.This is basically identical with document and investigation before, has its ratio of again damaging of sportsman of damage history apparently higher than the sportsman who was not subject to wound; Technical movements mistake, action distortion easily causes the stressed improper generation that causes damage; Training and competition are misorient refers to that training and competition arrangement does not meet athletes ' body function rule more, causes exercise intensity and amount of exercise excessive, causes that athletes ' body is overtired, thereby causes sick and wounded generation; Injury recovery situation is most important for sportsman, sick and wounded more not or participate in training and competition while not recovering completely and be easy to cause again generation injured and that other positions are injured; Athletic psychological factor also can have influence on the performance of training and competition, be overexcited, nervous, anxiety or psychological regulation ability etc. may cause athletes ' body muscle rigidity, inharmonious etc., thereby increase sick and wounded danger; Stability of joint mainly refers to have or not loose-jointed situation, and stability of joint declines and also makes easily to cause one of sick and wounded reason.Therefore, sportsman, in peacetime training and match, should fully understand oneself health, has not damaged and injury recovery situation, whether technical movements are wrong, whether training and competition amount is excessive etc., and should strengthen psychological regulation training, to reduce and to reduce the generation of damage.

The cause injury right assessment of factor various factors inside of table 46 β class

As shown in Table 46; in the weight analysis assessment of inner each Risk Factors of the factor that β class is caused injury; with other people or apparatus bump (33.5%) be the factor that the most easily causes damage, next is respectively health anatomical structure abnormal (26.0%), safeguard measure (13.9%), muscle strength and harmony (12.5%), range of motion (6.2%), balance ability (4.6%) and technical merit (3.2%).Bump comparatively rare in track and field event with other people or apparatus, generally see the sportsmen such as middle-distance race off or other when fighting for position, squeeze mutually and touch or in the accident collision of middle athletes ' body and hurdle frame of hurdling, once occur, be easy to cause injured, therefore, sportsman should focus one's attention in training and competition, avoids the generation of this type of accident; Whether health anatomical structure extremely mainly refers to that whether sportsman's lower limb length is consistent, have in knee, to turn up and whether the shape of inside of the foot has abnormal situation, if sportsman has Leg length to differ or other abnormal conditions, all can increase the risk of damage; Safeguard measure is most important for pre-antisitic defect, especially has the position of old wound, answers more protection, to reduce the generation of damage; Muscle strength is a basic fitness, most important for athletic achievement performance, sportsman is easy to cause the generation of damage when development force as the balanced development of muscle group before and after not noting and left and right limb muscle strength, especially more obvious at lower limb, therefore, sportsman, in development muscle strength, must take into account the balanced development of strength; The main range of motion of athletes ' body (pliability), balance ability and technical merit also can have influence on sick and wounded generation, and therefore, sportsman wants to reduce damage, just need to improve the Comprehensive Quality of health and improve technical merit.

The cause injury right assessment of factor various factors inside of table 47 γ class

As shown in Table 47, in the weight analysis assessment of inner each Risk Factors of the factor that γ class is caused injury, training or match (24.0%) are the most remarkable to the motion risk of causing injury, and next is respectively athletic ground (19.7%), menstrual cycle (14.1%), physical training amount (9.7%), body shape (BMI) (8.2%) and sport footwear (8.2%), weather conditions (7.5%), age (5.0%) and sex (3.6%) etc.For sportsman, the training and competition of participating in large intensity can increase injured possibility undoubtedly, its exercise intensity is larger, competition Shaoxing opera is strong, injured chance is just more, for sportsman, this type of factor not too can be controlled undoubtedly, can only try one's best and reasonably arrange training and match to reach the object that reduces damage; The variation on athletic ground also can affect concerning sportsman, especially sportsman trains or competes at hard place (as road etc.), can increase injured possibility undoubtedly, and in wet and slippery place or out-of-flatness place equally also can increase injured may, therefore, sportsman should reduce the harmful effect being brought by place as far as possible; For sportswoman, menstrual period participates in training or match may bring bad impact, especially the situation of irregular or amenorrhoea of menstrual cycle, can increase the probability of player injuries, therefore, sportswoman should understand the physical condition of self in time, adopts the harmful effect of adequate measures to avoid the menstrual period to cause; Athletic physical training amount mainly refers to the height of sportsman's grade, and research shows, sport rank level is higher, and the rate of injury is also larger, this with high performance athlete's exercise load more greatly, more more relevant; And the also correlativity certain with having of injury gained in sports respectively such as athletes ' body form (BMI), sport footwear, weather conditions, age, sex, sportsman should be appreciated that the harmful effect that various Risk Factors may be brought in peacetime training and match, evade in time and reduce the various risks of causing injury, sick and wounded to reach minimizing, the object improving results.

The table 48 inside factor information statistical form of causing injury

The table 49 outside factor information statistical form of causing injury

Table 50 incitant Information Statistics table

According to the weight of the above different factors, different motion shown in table 50-52 is damaged the factor score of causing injury, can calculate each athletic PTS, based on sportsman, damage the normal distribution hypothesis of risk, by the data definition within the scope of next standard deviation in all sportsman's score averages, it is medium risk grade, on be high-risk grade, under be low risk level.Injury gained in sports Risk-warning result based on dynamic chain quantitative model is specifically in Table 51, and in table, 1,2,3 represent respectively three Risk-warning grades from low to high.

Early warning result table after the injury gained in sports Risk-warning result of table 51 based on dynamic chain quantitative model and correction

In sum, the injury gained in sports risk factors factor that to be sportsmen all likely face in each stage of training and competition and link of causing injury, but the probability of happening of the various risk factors of causing injury and the impact causing are not quite similar, therefore, by all kinds of factors being carried out to weight analysis assessment based on analytical hierarchy process, contribute to us to understand the cause injury size of risk of the various factor pair injurys gained in sports of causing injury, for the high factor of the risk of causing injury, as the α class factor of causing injury, once occur, should adjust in time, to reduce injured possibility; And the β class factor of causing injury equally also can not be ignored, sportsman should fully understand self health in peacetime training and match, improves comprehensive fitness and technical merit, and reducing damage morbidity may; γ class cause injury the factor for cause sportsman damage impact less, sportsman should strengthen self-monitoring in peacetime training, evades and reduces all kinds of risks of causing injury.The size of causing injury and impacting by calculating and assess all kinds of factor pair sportsmen that cause injury, has set up damage Risk-warning cause of disease dynamic chain quantitative model.

Five, according to sports events, choose sportsman and the Risk-warning factor, utilize the corresponding data in exercise risk database, set up exercise risk early warning Sample Storehouse and the test sample book data of this project;

In track and field sports, may cause the factor of injury gained in sports a lot, different sports eventss is different again, therefore will set up one to the general injury gained in sports risk warning model of all sports eventss infeasible.Set up like this Athletes damage risk assessment database, for different sports eventss, when doing risk assessment, it will be optimal selection that selected part sportsman and part index number are set up risk warning model targetedly.It is shown in table 28 that database should comprise variable, when wherein sportsman's essential information is set up with different motion project model, and the screening of sportsman's sample.After sportsman has screened, according to the characteristic of disparity items, according to the conclusion of injury gained in sports early warning related-factors analysis, the factor, outside of causing injury from inside caused injury and chosen and set up the required variable of injury gained in sports early warning Sample Storehouse the factor and the Induced by Stimulation factor, form Risk-warning Sample Storehouse with the injury gained in sports risk class that utilizes actual conditions+dynamic chain quantitative model+expert evaluation composite algorism to obtain, the factor in Risk-warning Sample Storehouse is stored in different databases in the sick and wounded early warning system of athlete, while setting up Risk-warning Sample Storehouse, can call respectively.Take long-distance runner as example, and the concrete factor of Risk-warning Sample Storehouse is as follows:

1) sportsman's essential information: name, project, sportsman's grade, BMI;

2) injury gained in sports Risk-warning information: whether state at heart, the training plan in training load amount, error of performance, joint safeguard measure, Subjective fatigue situation, training and competition changes, the frequency, main training site are changed in muscle strength contrast before and after training site weather condition, stability of joint, joint flexibility, thigh, the contrast of thigh circumference degree, balance ability, the shape of inside of the foot, technical merit, sport footwear, damage history, resume training situation, menstrual cycle after damaging;

3) training monitor message: RBC number, haemoglobin, hematocrit, SCK, blood urine element, serum testosterone, serum cortisol, serum testosterone (T)/serum cortisol (C), leukocyte count.

Six, judge the discreteness of the achievement data that Risk-warning index is corresponding, if not discrete, call SOM neural network discrete method achievement data is carried out to discretize processing

Factor that can continuous coverage numerical value, as training condition (physiological and biochemical index), is carried out discretize processing.The requirement of discretize is: (1) terseness, and the breakpoint number that discretize adopts is more few better; (2) consistance, does not change the compatibility of decision table after discretize; (3) accuracy, the precision of prediction of discretize result is more high better.

Discrete processes adopts SOM algorithm, SOM is a kind of without supervision self-organizing competing learning type feedforward neural network, can by a large amount of training sample data of Ad hoc mode utilization, adjust the weights of network, classification results can reflect the actual distribution situation of data comparatively objectively.When application SOM neural network is carried out discrete processes to connection attribute, it is very crucial correctly choosing clusters number, if because clusters number is very few, may obtain inconsistent decision system, while causing practical application, according to Rule of judgment, cannot make a policy; If clusters number is too much.There will be discrete case, increased the complicacy of calculating.In actual analysis, first using the foundation that the incompatibility of decision table after discretize and the incompatibility of each attribute pass judgment on as discrete performance.So just consider the incompatibility of decision table and the distribution characteristics of data itself after Data Discretization, both reduced as much as possible the number of breakpoint, guaranteed again the compatibility of decision table.

Definition 1: the importance degree of connection attribute a ∈ C (conditional attribute collection) is defined as follows:

1) calculate U/D, according to decision attribute, domain is divided into n class.Wherein U is that in Sample Storehouse, sportsman is all; D is decision attribute, i.e. final injury gained in sports Risk-warning grade.

2) calculate the mean value AVG (ai) of connection attribute a in each class (exercise risk damages the factor of causing injury), i=1,2 ..., n, and find out maximal value Max (AVG (ai)) and the minimum M in (AVG (ai)) of mean value.

3) calculate the standard deviation std (a) of all object connection attribute a.

The importance degree SGF (a) of attribute a is:

$\mathrm{SGF}\left(a\right)=\frac{\mathrm{Max}\left(\mathrm{AVG}\left(a_i\right)\right)-\mathrm{Min}\left(\mathrm{AVG}\left(a_i\right)\right)}{\mathrm{std}\left(a\right)}$ (formula 16)

Definition 2: a given infosystem S=(U, C ∪ D), if incompatible object has m after discretize, concrete calculating seen formula 21, the incompatibility α that defines decision table is:

$\mathrm{\&alpha;}=\frac{m}{\left|U\right|}$ (formula 17)

Wherein | U| is the sum of object in decision table.In practical application, only need to set a less incompatibility threshold values.In the applied research of current most of rough set, the neuronic number of SOM network output layer is subjective definite, easily occurs deviation.For this reason, according to above two definition, propose a kind of SOM network discretization method that is applicable to rough set, step is as follows:

1) calculate the importance degree of each connection attribute in decision table and sort by order from big to small.

2) each continuous injury gained in sports factor attribute a that causes injury _i(i=1,2 ..., | a C|) corresponding SOM network, below with an attribute discrete, turn to example explanation, the initial value of given SOM network output layer neuron number is 3, input layer number is 1, by the connection weights wj of network, gives [Min (a _i), Max (a _i)] random value in interval, determining the initial value α (0) of learning rate α (t), α (t) should slowly reduce with the increase of t, total the study number of times T determining;

3) by attribute a _iin numerical value x be input to som network;

4) calculate input numerical value x and the distance being connected between weights wj;

D _j=(x-w _j) ², j=1,2,3 (formula 18)

5) choose and make the minimum neuron k of distance as triumph node;

6) to triumph node and connection weights upgrade

W _j(t+1)=w (t)-α (t) (x-w _j) (formula 19)

7) return to step (3), choose another data input network, until all mode in sample is offered to network;

8) renewal learning rate:

$\mathrm{\&alpha;}\left(t\right)=\mathrm{\&alpha;}\left(0\right)[1-\frac{t}{T}];$ (formula 20)

9) make t=t+1, return to step 2), be repeated to t=T.

10) incompatibility of decision table after calculating discretize, calculates end if meet the requirement of incompatibility threshold values, otherwise turns next step.

11) select the attribute of importance sorting first, its SOM network output layer neuron number is adjusted into 4, and the SOM network output layer neuron number that all the other attributes are corresponding is constant, again discretize decision table, and go to step (10), until the maximum cluster numbers of each SOM network is adjusted into 4.The maximum cluster numbers of each SOM network is defined as 4, if its SOM network clustering number of the attribute of importance degree maximum has reached requirement, selects the attribute that importance degree takes second place to adjust, by that analogy.

12) complete final discretize operation.

Therefore because rough set theory is not high to the granularity requirements of knowledge, maximum cluster numbers is defined as 4 and just can substantially meets the requirement to incompatibility, if chance special circumstances suitably increase the neuronic maximum constraints number of SOM network output layer.

Seven, by the synthetic decision table of the achievement data after utilizing SOM network discrete and the initial data with regard to discrete valuation

Eight, according to the yojan process of RS theory, decision table is carried out to attribute reduction and Rule Extraction

By after the synthetic decision table of the data after utilizing SOM network discrete and the initial data with regard to discrete valuation, then according to the yojan process of RS theory, decision table is carried out to attribute reduction and Rule Extraction.The basic thought of RS is to use relation of equivalence that the element in set is classified, and generates certain division of set, corresponding with relation of equivalence.According to the theory of relation of equivalence, the element in same grouping equivalence class can not be differentiated, and to the processing of information, can in the granularity of equivalence class, carry out, and can reach the object that information is simplified thus.It is a kind of uncertain information disposal route based on set theory.Can effectively analyze and process out of true, incomplete various sizings, quantitatively or the incomplete information of mixed type, therefrom find tacit knowledge, disclose potential rule.

Attribute reduction has diverse ways, and this project is carried out yojan by discernible matrix.A.Skowron proposed the method for the clearly demarcated matrix representation knowledge of a kind of use in 1991, this method for expressing is conducive to us to be explained and computational data core and yojan.

Definition 3: establishing S=(U, A, V, f) is a decision table.Wherein U is domain, is all of sportsman's Sample Storehouse in this application; A is community set, A=C ∪ D, and C is conditional attribute collection, i.e. all potential injurys gained in sports set of factor attribute of causing injury, D is decision kind set, only comprises injury gained in sports Risk-warning grade attribute in this example; represent the set of all properties value, be the set of each attribute value of all sportsmen in Sample Storehouse;

F:U * A → V is information function, and it determines the value rule of each attribute of sportsman in Sample Storehouse.Definition a (x) is the value of sample x on attribute a, the clearly demarcated Matrix C of this decision table _dcan be expressed as:

(formula 21)

Matrix be one along the matrix of principal diagonal symmetry, when analyzing, only consider triangle or lower triangular portions on it.When the decision attribute value of two samples is identical, their corresponding differential matrix element values are 0; When the decision attribute of two samples is different and can be distinguished by the value difference of some conditional attribute, the value of their corresponding differential matrix elements is the different conditional attribute set of value on these two sample attributes, can distinguish the conditional attribute set of these two samples; When two samples clash, when all conditional attribute values value identical and decision attribute is different, the element value of their corresponding differential matrixs is 1.In differential matrix element, whether comprise 1 can be used as determination decisions table system whether comprise inconsistent information according to the foundation that is inconsistent decision table.

Attribute reduction based on differential matrix can be divided into following six steps.

1) read in injury gained in sports Risk-warning Sample Storehouse data, as decision table;

2) calculate the clearly demarcated Matrix C of this decision table _d;

3) in clearly demarcated matrix, all values are non-zero, 1, element c _ij, set up corresponding logical expression L _ij, wherein

(formula 22)

4) by all expression formula L that extracts _ijdo conjunction computing, obtain a conjunctive normal form L,

(formula 23)

5) form that above-mentioned conjunctive normal form is converted to disjunctive normal form is

6) output attribute yojan result.The result of the corresponding attribute reduction of each conjunct in disjunctive normal form, the conditional attribute set of the set of properties comprising in each conjunct after into about letter.Available these different yojan results, Training RBF Neural Network, chooses the final yojan result of the best conduct of classifying quality respectively.

Specific algorithm is as follows:

Nine, the decision table based on simplifying builds RBF neural network

RBF neural network structure and feedforward neural network are similar, and it is a kind of three layers of feedforward network.Input layer is comprised of signal source node.The second layer is hidden layer, the needs of visual the described problem of nodes and determining.The 3rd layer is output, and it makes response to the effect of input pattern.Conversion from the input space to hidden layer space is nonlinear, and conversion from hidden layer space to output layer space is linear.The transforming function transformation function of hidden layer unit is radial basis function, and it is a kind of non-negative nonlinear function to the decay of central point radial symmetry of local distribution.

RBF neural network topology structure is as shown in table 8.The conversion of RBF network from the input space to implicit space is nonlinear, and the conversion from hidden layer space to output region is linear.This is a kind of topological structure of feedforward network.

By RBF neural network, approach nonlinear system, given, after learning sample data, its algorithm mainly solves two problems:

1) Neural Network Structure Design problem, comprises determining of network hidden layer node number, and the cluster centre C of RBF determines;

2) weights correction problem, adjusts hidden layer space to the connection weight matrix w in output layer space.

This project is mainly used RBF neural network, according to input message distinguish athletic injury gained in sports Risk-warning grade be high, in or low, namely sportsman is divided three classes according to injury gained in sports risk size.For the purpose of simple, by 1,2,3 of RBF neural network output valve, represent this three classifications.Specific as follows:

Table-52RBF output and the risk class table of comparisons

Be that the object of this project is for being used sorter of RBF neural network configuration.Mainly comprise that basis function is chosen, the change of hidden layer design and radial basis function center, weights and width.This project neural network hidden layer center, width and definite to the weights employing gradient descent method between output layer.

(1) basis function chooses

Select Gaussian function as the activation function of hidden layer unit, suppose that input vector is n dimension, have m output node, the number of hidden node is k, and the output of hidden node is expressed as:

$R_j\left(x\right)=e^{-\frac{{\left|\right|x-c_j\left|\right|}^2}{{2\mathrm{\&sigma;}}_j^2}}$ (formula 24)

J={1 wherein, 2 ..., k}, Rj (x) is j hidden layer node to output, the n dimension input vector that x is network, c _jbe the kernel function center vector of j hidden layer node, it has identical dimension, σ with input vector x _jbe the width of j hidden layer node Gaussian function, || x-c _j|| represent the distance measure between input vector and central point.Network output is defined as:

$y_i=\underset{j=1}{\overset{k}{\mathrm{\&Sigma;}}}{w}_{\mathrm{ij}}{R}_j\left(x\right)$ (formula 25)

I=1 wherein, 2 ..., m; Yi is the output of i output layer node, w _ijbe that j hidden layer node is to the connection weights of i output layer node.In this project, only need 1 output, i.e. i=1.

(2) hidden layer design

The hidden layer of design RBF network is mainly to determine hidden layer number of unit and their excitation function.This project adopts Gaussian function, and Gaussian function is determined by cluster centre and cluster width.In this project, the classification of training mode is known, and injury gained in sports Risk-warning grade, therefore can adopt a kind of simple mode to determine, the corresponding Gaussian function of each class is totally three, i.e. k=3.The width parameter of Gaussian function can be got the average of three class samples and its central point distance, and the average that three class samples can be got respectively in the center of Gaussian function is as the central point of unit separately.Although it is somewhat coarse that the network of this method construct seems relatively, by the strategy of effective learning algorithm and employing error correction, can effectively improve the performance of network.In actual classification application, the network of constructing by such mode is succinctly efficient, can not affect classification accuracy simultaneously.

(3) radial basis function center, width and to the renewal of the weights between output layer

Radial basis function center and other parameter all experience learning process.Conventionally adopt error correction learning process, application gradient descent method, specific as follows:

Suppose to have N sample input, to all input samples, definition error function:

$\mathrm{\&xi;}=\frac{1}{2}\underset{q=1}{\overset{N}{\mathrm{\&Sigma;}}}{e}_q^2$ (formula 26)

E wherein _qfor error, be defined as follows:

$e_q=d_q-y\left(x_q\right)=d_q-\underset{j=1}{\overset{3}{\mathrm{\&Sigma;}}}{w}_{1j}{R}_j\left(x_q\right)=d_q-\underset{j=1}{\overset{3}{\mathrm{\&Sigma;}}}{w}_{1j}{e}^{-\frac{{\left|\right|x_q-x_j\left|\right|}^2}{{2\mathrm{\&sigma;}}_j^2}}$ (formula 27)

D wherein _qfor sample x _qthe value of required type.As the d corresponding to sample of high risk class in this project _q=3.

The iterative process of each free parameter is as follows:

1) weights of output unit

$\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;w}_{1j}\left(n\right)}=-\underset{q=1}{\overset{n}{\mathrm{\&Sigma;}}}{e}_q\left(n\right)R_j\left(x_q\right)=-\underset{q=1}{\overset{n}{\mathrm{\&Sigma;}}}{e}_q{e}^{-\frac{{\left|\right|x_q-c_j\left|\right|}^2}{{2\mathrm{\&sigma;}}_j^2}}$ (formula 28)

$w_{1j}(n+1)=w_{1j}\left(n\right)-{\mathrm{\&eta;}}_1\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;w}_{1j}}$ (formula 29)

Wherein n represents current variable-value, and n+1 represents the revised value of iteration.

2) hidden unit center

$\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;c}_j\left(n\right)}=-\underset{q=1}{\overset{n}{\mathrm{\&Sigma;}}}{e}_q\left(n\right)R_j\left(x_q\right)=-\underset{q=1}{\overset{n}{\mathrm{\&Sigma;}}}{e}_q\frac{w_{1j}\left(n\right)}{{\mathrm{\&sigma;}}_j^2}{e}^{-\frac{{\left|\right|x_q-c_j\left(n\right)\left|\right|}^2}{{2\mathrm{\&sigma;}}_j^2}}(x_q-c_j\left(n\right))$ (formula 30)

$c_j(n+1)=c_j\left(n\right)-{\mathrm{\&eta;}}_2\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;c}_j}$ (formula 31)

3) function width

$\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;\mathrm{\&sigma;}}_j\left(n\right)}=-\underset{q=1}{\overset{n}{\mathrm{\&Sigma;}}}{e}_q\left(n\right)\frac{w_{1j}\left(n\right)}{{\mathrm{\&sigma;}}_j^3\left(n\right)}{\left|\right|x_q-c_j\left|\right|}^2{R}_j\left(x_q\right)$ (formula 32)

${\mathrm{\&sigma;}}_j(n+1)={\mathrm{\&sigma;}}_j\left(n\right)-{\mathrm{\&eta;}}_3\frac{\&PartialD;\mathrm{\&xi;}\left(n\right)}{{\&PartialD;\mathrm{\&sigma;}}_j\left(n\right)}$ (formula 33)

η wherein ₁, η ₂, η ₃for learning efficiency can also can be variable for constant.

Ten, according to the decision table after simplifying, choose the corresponding data in exercise risk early warning Sample Storehouse, Training RBF Neural Network, learning method detailed process is as follows:

Step 1: determine hidden node number according to known class number, k=3 in this project, asks for the mean value of all input vectors of this classification as such center initial value c _j(1), the width initial value σ using the mean value of all input vectors of every class and centre distance as such _j(1), a sample of each classification input, asks for initial w _1j(1), set allowable error ε, ε is an a small amount of that is greater than zero.Set learn efficiency η ₁=η ₂=η ₃=1/t, t is iterative loop variable.Set loop variable t and be initially 1, maximal value is MaxT.

Step 2: input t training sample, obtain the actual output of network y (t).

Step 3: calculate the error ξ (t) between actual output and desired output, if error ξ (t) < is ε, this sample does not need to adjust network parameter, jumps to step 6, otherwise enters next step.

Step 4: press formula 28,30,32 calculate and press formula 29,31,33 calculate w _1j(t+1), c _j(t+1), σ _j(t+1).

Step 5: upgrade new network parameter w _1j(t+1), c _j(t+1), σ _j(t+1), t=t+1; If t>MaxT, shows and do not restrain, enter step 6, otherwise forward step 2 to.

Step 6: whole learning process finishes, preserves each parameter of current network;

Step 7: utilize the network parameter of preserving to test new athletic damage alarming grade.

11, judge whether yojan result restrains, and whether meets accuracy requirement, if do not meet, modified RBF network structure be connected weights, go to step ten;

12, utilize test sample book data, whether the RBF network that training of judgement is good obtains correct diagnostic result, if do not have, selects other yojan result from step 8, goes to step nine;

13, preserve the RBF network training, utilize the new sportsman's risk assessment index detecting of RBF neural network assessment training.

Step 12 specifically comprises following process:

(1) set up test sample book data

When setting up exercise risk early warning Sample Storehouse Sample Storehouse, leave and take a part of data as test sample book, so just guaranteed the consistance of injury gained in sports Risk-warning grade discrimination, can effectively to the accuracy of model prediction, verify.Because athlete crowd more than domestic one-level is less, the present embodiment gathers 148 player information altogether, and 131 people's information are used for setting up training sample database, for Training RBF Neural Network; Choose 17 athletes as test sample book, the accuracy of the sick and wounded early warning system warning algorithm of athlete is tested.

(2) data importing

The cause injury typing of the factor of the sick and wounded Risk-warning of Athletes is mainly divided into four parts: individual essential information directly imports when building storehouse, and information can be revised by the personal information page in movable information management; Physiological and biochemical index adds, revises in monitor message management; Discrete damage is caused injury mostly, and the sick and wounded risk information page of the individual of factor information in sick and wounded information management adds, modification; The part often changing in the discrete factor, the sick and wounded early warning page in injury gained in sports early warning adds also directly to be submitted to and obtains early warning result.

(3) Risk-warning diagnostic result

This project utilizes respectively the sick and wounded intelligent early-warning system of actual conditions+quantitative model, actual conditions+quantitative model+expert evaluation and athlete to predict to the injury gained in sports advanced warning grade of test sample book, and result is as shown in table 53.In the sick and wounded intelligent early-warning system of athlete, adopt Rough Set-Neural Network algorithm to predict damage risk, in Sample Storehouse construction in combining information collection actual motion sick and wounded a situation arises, the composite algorism of coach's expertise and the quantification dynamic chain model based on analytical hierarchy process is determined injury gained in sports risk class, after in therefore final table 53, the risk class of two row is in full accord, illustrates that this method has reached designing requirement.

Actual motion is sick and wounded, and a situation arises, the composite algorism concrete steps of coach's expertise and the linear dynamic chain quantitative model based on analytical hierarchy process are as follows:

(1), as there being injury gained in sports to occur in actual training process, the data that this Athletes damage gathers before occurring are high-risk grade;

(2) for the situation that does not move damage, according to image data, use and dynamically to connect linear model calculation risk grade, and this athletic injury gained in sports risks and assumptions score situation is offered to 15 senior above traineres, team doctor and scientific research personnel simultaneously and carry out risk class marking.

(3) as coach, team doctor and scientific research personnel and marking result surpass 80%, be same risk class, using this result as this athletic risk rating result; Otherwise using dynamic chain quantitative model result as risk rating result.

The comparison of table 53 distinct methods exercise risk early warning result

Known by above test, use rough set & RBF network algorithm to obtain comparatively desirable result to injury gained in sports Risk-warning, so this algorithm is feasible for athletic injury gained in sports Risk-warning.

Certainly, above-mentioned explanation is not limitation of the present invention, and the present invention is also not limited only to above-mentioned giving an example, and the variation that those skilled in the art make in essential scope of the present invention, remodeling, interpolation or replacement, also belong to protection scope of the present invention.

Claims (6)

1. an Athletes damage method for prewarning risk, is characterized in that, comprises the steps:

(1) analysis of the injury gained in sports Risk-warning factor;

(2) the sports injury muscle injury Risk-warning factor dynamic chain model of causing injury is proposed;

(3) set up sportsman's Risk-warning database;

(4) use analytical hierarchy process, set up the sports injury muscle injury Risk-warning factor dynamic chain quantitative model of causing injury;

(5) according to sports events, choose sportsman and the Risk-warning factor, utilize the corresponding data in exercise risk database, set up exercise risk early warning Sample Storehouse and the test sample book data of this project;

(6) discreteness of the Risk-warning factor corresponding data in judgement exercise risk early warning Sample Storehouse, if not discrete, call SOM neural network discrete method data is carried out to discretize processing;

(7) by the synthetic decision table of the achievement data after utilizing SOM network discrete and the initial data with regard to discrete valuation;

(8) according to the yojan process of RS theory, decision table is carried out to attribute reduction and Rule Extraction, obtain one group of yojan result;

(9) decision table of one of them simplification based on selecting builds RBF neural network;

(10) according to the decision table after simplifying, choose the corresponding data in exercise risk early warning Sample Storehouse, Training RBF Neural Network;

(11) judge whether yojan result restrains, and whether meets accuracy requirement, if do not meet, modified RBF network structure be connected weights, go to step (10);

(12) utilize test sample book data, whether the RBF network that training of judgement is good obtains correct diagnostic result, if do not have, selects other yojan result from step (8), goes to step (9);

(13) preserve the RBF network training, utilize the new sportsman's risk assessment index detecting of RBF neural network assessment training.

2. Athletes damages method for prewarning risk according to claim 1, it is characterized in that, the SOM neural network discrete method of described step (6) comprises the steps:

(6.1) calculate the importance degree of each connection attribute in decision table and sort by order from big to small;

(6.2) each continuous injury gained in sports factor attribute a that causes injury _i(i=1,2 ..., | a C|) corresponding SOM network, the initial value of given SOM network output layer neuron number is 3, input layer number is 1, by the connection weights wj of network, gives [Min (a _i), Max (a _i)] random value in interval, determining the initial value α (0) of learning rate α (t), α (t) should slowly reduce with the increase of t, t representative study number of times, initial value is 1, total the study number of times T determining;

(6.3) by attribute a _iin numerical value x be input to som network;

(6.4) calculate input numerical value x and the distance being connected between weights wj;

D _j=(x-w _j) ², j=1,2,3 (formula 1)

(6.5) choose and make the minimum neuron k of distance as triumph node;

(6.6) the connection weights of triumph set of node are upgraded

W _j(t+1)=w (t)-α (t) (x-w _j) (formula 2)

(6.7) return to step (6.3), choose another data input network, until all mode in sample is offered to network;

(6.8) renewal learning rate:

(formula 3)

(6.9) make t=t+1, return to step (6.2), be repeated to t=T;

(6.10) incompatibility of decision table after calculating discretize, calculates end if meet the requirement of incompatibility threshold values, otherwise turns next step;

(6.11) select the attribute of importance sorting first, its SOM network output layer neuron number is adjusted into 4, the SOM network output layer neuron number that all the other attributes are corresponding is constant, again discretize decision table, and go to step (6.10), until the maximum cluster numbers of each SOM network is adjusted into 4, the maximum cluster numbers of each SOM network is defined as 4, if its SOM network clustering number of the attribute of importance degree maximum has reached requirement, select the attribute that importance degree takes second place to adjust, by that analogy;

(6.12) complete final discretize operation.

3. Athletes damages method for prewarning risk according to claim 1, it is characterized in that, described step (8) adopts the clearly demarcated matrix method based in rough set to carry out Decision Table Reduction.

4. Athletes damages method for prewarning risk according to claim 3, it is characterized in that, the described clearly demarcated matrix method based in rough set comprises the steps:

(8.1) read in exercise risk early warning Sample Storehouse data, as decision table;

(8.2) calculate the clearly demarcated Matrix C of this decision table _d;

(8.3) in clearly demarcated matrix, all values are non-zero, 1, element c _ij, set up corresponding logical expression L _ij, wherein

(formula 4)

I, j are clearly demarcated entry of a matrix element subscript;

(8.4) by all expression formula L that extracts _ijdo conjunction computing, obtain a conjunctive normal form L,

(formula 5)

(8.5) form that above-mentioned conjunctive normal form is converted to disjunctive normal form is the yojan result number that i representative is finally obtained;

(8.6) output attribute yojan result, the corresponding attribute reduction result of each conjunct in disjunctive normal form, the conditional attribute set of the set of properties comprising in each conjunct after into about letter.

5. Athletes damages method for prewarning risk according to claim 1, it is characterized in that, described step (9) comprises the steps:

(9.1) choose basis function

Select Gaussian function as the activation function of hidden layer unit, suppose that input vector is n dimension, have m output node, the number of hidden layer node is k, and the output of hidden layer node is expressed as:

(formula 6)

J={1 wherein, 2 ..., k}, R _j(x) be j hidden node to output, the n dimension input vector that x is network, c _jbe the kernel function center vector of j hidden layer node, it has identical dimension, σ with input vector x _jbe the width of j hidden layer node Gaussian function, || x-c _j|| represent the distance measure between input vector and central point, network output is defined as:

(formula 7)

I=1 wherein, 2 ..., m; y _ibe the output of i output layer node, w _ijbe that j hidden layer node is to the connection weights of i output layer node;

(9.2) hidden layer design

The hidden layer of design RBF network, mainly to determine hidden layer number of unit and their excitation function, adopt Gaussian function, Gaussian function is determined by cluster centre and cluster width, the classification of training mode is known, it is injury gained in sports Risk-warning grade, therefore can adopt a kind of simple mode to determine, be totally three of the corresponding Gaussian functions of each class, be k=3, the width parameter of Gaussian function can be got the average of three class samples and its central point distance, and the average that three class samples can be got respectively in the center of Gaussian function is as the central point of unit separately;

(9.3) radial basis function center, width and to the renewal of the weights between output layer

Radial basis function center and other parameter all experience learning process, conventionally adopt error correction learning process, and application gradient descent method is specific as follows:

Suppose to have N sample input, to all input samples, definition error function:

(formula 8)

E wherein _qfor error, be defined as follows:

(formula 9)

D wherein _qfor sample x _qthe value of required type, d corresponding to the sample of high risk class in this project _q=3.

6. Athletes damages method for prewarning risk according to claim 1, it is characterized in that, described step (10) comprises the steps:

Step 1: determine hidden layer node number according to known class number, k=3 in this project, asks for the mean value of all input vectors of this classification as such center initial value c _j(1), the width initial value σ using the mean value of all input vectors of every class and centre distance as such _j(1), a sample of each classification input, asks for initial w _1j(1), set allowable error ε, ε is an a small amount of setting learn efficiency η that is greater than zero ₁=η ₂=η ₃=1/t, t is iterative loop variable, sets loop variable t and is initially 1, maximal value is MaxT;

Step 2: input t training sample, obtain the actual output of network y (t);

Step 3: calculate the error ξ (t) between actual output and desired output, if error ξ (t) < is ε, this sample does not need to adjust network parameter, jumps to step 6, otherwise enters next step;

Step 4: press formula

formula

formula

Calculate and press formula

formula

formula calculate w _1j(t+1), c _j(t+1), σ _j(t+1);

Step 5: upgrade new network parameter w _1j(t+1), c _j(t+1), σ _j(t+1), t=t+1; If t>MaxT, shows and do not restrain, enter step 6, otherwise forward step 2 to;

Step 6: whole learning process finishes, preserves each parameter of current network.