Modeling Of The Training Load -Performance Relationship In Sanda Competitors

Objective: According to the hypothesis of“super-compensation”and“capability reserve”,“training load– performance”relationship was modeled for understanding the athletes’personal characters of adaption to training. According to the outlet of modeling, four“taper”planes of pre-competition were simulated, the time of reaching peak performance and the value of peak performance were discussed.Methods: the model was added parameter of extreme limit of adaption based on the Banister’s model of training load performance relationship. The data for parameters estimating come from the training data of six Sanda competitors. The competitors’age was 18.2±0.75, and the training period was 4.6±0.49. during the test, the training plan was made by the coacher.Model’s input variable was the external training load. For the team’s training manners was regulate, the external training loads of the training manners were calibrated by quantifying the inner training load of the training manners using TRIMP method. After calibration, the training day’s external training load was calculated by the calibrated training load of training manners and the training contents of everyday’s. During the calibrating, the subjects warred the Suunto T6 meters through the whole training courses about two weeks. After the calibrating course, all the training manners’training load values and average heart rates. At the same time, every training manners’RPE values were recorded by questionnaire. Maximum heart rates for TRIMP calculation were the highest heart rate during 4 rounds’real competitions. The resting heart rate to test the day resting heart rate shall prevail. The average heart rate to a mean heart rate training content, including the time between the intermittent rests. The average heart rate of the training manners were compared with 10 points RPE values.Model’s output variable is the performance. Performance was quantified the ration of external load and internal load, whole training session treated as a physical test, as the output of the previous inputs. Method of determining the external load is acquired as above; the internal load was quantified by the level of heart rate recovery after the physical training session of each day. The level of heart rate recovery was adjusted according to the load center of gravity.According to "super-compensation" and "capability reserve" hypothesis, the training load-performance relationship was modeled. Parameters were estimated using Simulink Parameter Estimation toolbox in nonlinear least squares parameter estimation, the adapt and fatigue and the two intermediate state variables need to be limited. Goodness of fit of the models was tested by the determine coefficient ( R 2). The residuals’normality test was the Shapiro-Wilk test; the significant level P> 0.5 means the distribution of residual is normality. After model’s parameters obtained, the performance responses to 4 taper training plans were simulated in six subjects for investigation to the maximum value of performance status and the time required for reaching the peak performance. Before the load of the 4 taper training plans, the initial load of the simulated training was 88( AU) over 28 days for the performance initial stable conditions. Taper training plans include liner decrease and nonlinear decrease, each type had over load or not before decreasing. for the non-overload training, directly simulate the taper process, for overload training before reduction, 130% of the average load during prevail training sessions for 1 week training simulated. All the plans’final training load were about 36 (AU). The means of the maximum performance value and the time to reach the peak performance and the period of maintaining higher physical state (95% of added value than before reduction) were compared in one-way ANOVA.Results: the average heart rate of the training manners and the RPE values were significantly highly correlated; the correlation of TRIMP of the training manners and the adjusted heart rate level was 0.833±0.068 (P <0.05), non-linear fitting of heart rate recovery was R 2 =0.91±0.031(P <0.01). Estimate the model parameters fitting degree, only one subject’s residual normality test is refused to be true, compared to similar studies, the results of model’s fitting is better. Tapering the simulation found that the reduction of overload before the load reached after the exponential reduction of the highest level of physical condition, the linear reduction of overload physical state after the maximum time required for the longest, while it maintains a high Physical state of the longest.Conclusion: The model of the relationship of the training load and performance is improved based on Banister’s "training load - performance" model by adding the adaption’s limitation parameter, the model becomes nonlinear, model’s fitting results compared prevail research was reasonable. Modeling results indicated that adaption hypothesis is acceptable, but does not exclude the other adaptation mechanisms hypothesis. The simulation of the taper using the established model shows that the linear reduction of load volume achieved the lower performance than nonlinear reduction, but maintains longer period of high state of performance, this makes the coacher and the competitors easily to catch the high training status in competition. At the end of the paper, based on the hypothesis of modeling, the suggestion of dynamic diagnosis and classification of the performance variation was proposed.