Swimming Human Simulation Model Research Articles

Development of a comprehensive method for musculoskeletal simulation in swimming using motion capture data

The objective of this study was to develop a comprehensive method for musculoskeletal simulation in swimming. In the developed method, the motion capture data of a swimming motion were first transformed into input data for the swimming human simulation model Swimming Human Model (SWUM). Then, the output data from SWUM were transformed into input data for the musculoskeletal simulation. In order to examine the validity of the developed method, the musculoskeletal simulations of the crawl and breaststroke by two swimmers were conducted. Based on the results of the crawl, the muscles of the upper limbs were activated significantly at the pull and push phases. The swimmer also adjusted the swimming speed, not by changing the effort of the upper limbs during the pull and push phases but by changing the length of the entry phase and the effort of the flutter kick. The results of the breaststroke showed that the muscles of the lower limbs were activated significantly at the kick phase. The swimmer adjusted the swimming speed, not by changing the effort of the upper limbs but by changing the length of the glide phase and the effort of kicking. The results from this study were consistent with the findings in previous studies, except for the period of large (nearly 180 degrees) flexion or abduction at the shoulder joint. The results also revealed the strategies of the swimmers to adjust their swimming speed. This indicates the promising potential of the musculoskeletal simulation, that is, it will be possible to estimate the strategy of a swimmer by conducting the simulation.

Preliminary Study of Backstroke by the Swimming Humanoid Robot

The swimming humanoid robot called SWUMANOID has been developed for research purpose. To date, the front crawl, the breaststroke, and the butterfly stroke have been realized by this robot. However, the backstroke has not been realized yet because of several problems. The first problem is the communication problem during an experiment. The second is that the motion of backstroke which adapted to SWUMANOID' DOFs has not been created. In this research, the creation of backstroke motion adapted to SWUMANOID is discussed. The swimming human simulation model SWUM was employed as a tool to obtain the swimming animation as well as the swimming speed and the straightness of swimming. Finally, the experimental plan was explained briefly in order to conduct an experiment in the future.

Development of a markerless optical motion capture system for daily use of training in swimming

The main objective of this research was to develop a markerless optical motion capture system that can be used for daily use in swimming training. The butterfly stroke was targeted since it is considered bilaterally symmetric in motion. The system consisted of a segmentation process to obtain the participant’s silhouettes and a matching process to estimate the pose of the participant. A variable thresholding method was used to extract the silhouettes to solve non-uniform illumination in the recorded swimming video. Prior to the matching process, the human body was modeled as a series of nine segments to help the matching process. The model was then mapped so that it aligned with the silhouettes, which were investigated by similarity of intensity value. To minimize the degree of freedom in image matching, the available joint motion in the swimming human simulation model was used as a priori information for kinematics data of the swimming motion. As a result, the rotation angle’s correlation coefficients between the references and result of the matching process were around 0.95 for trunk, thigh, shank, upper arm, forearm, hand and 0.78 for head, hip and foot. The rotation angle and the velocity of the center of mass were put into the swimming human simulation model for a dynamics analysis. The simulation results show that the velocity obtained in the experiment corresponded to the fluid force exerted on the lower and upper limbs. Consequently, the proposed system of obtaining the joint motion of the butterfly stroke is suitable for daily training and coaching.

片麻痺スイマーのための自由形アームストロークの最適化シミュレーション

For the competitive swimmers with physical disability, an ideal stroke to master has not been fully clarified yet. The objectives of this study were to solve computationally the arm stroke of the crawl swimming that maximized the swimming speed for swimmers with hemiplegia, and to investigate the features of the optimized stroke. For this objective optimizing calculations with the random search, PSO (Particle Swarm Optimization) algorithm, and the swimming human simulation model was conducted. In the optimizing calculation the design variable was the upper limb motion of the unaffected side, represented by five degrees-of-freedom (three at the shoulder joint and two at the elbow joint), and the fastest stroke was obtained as the optimum stroke. For the motions of the other body parts, the legs, trunk and the upper limb of the affected side, the motion of the actual swimmer were used. When the stroke cycle was 1.0 s or 1.1 s, the swimming speed became fastest (1.19 m/s). With the short stroke cycles there were the differences at the end of the stroke, and with the long stroke cycles the swimmer pushed the water deeper in the optimum stroke.

ダイビングフィンの性能評価を目的としたシミュレーションモデルの構築

Scuba diving is becoming popular as a leisure activity. Although manufacturers make various types of diving fin, these developments are based on qualitative methods because of the difficulty in estimating the fin efficiency quantitatively. In this research, a simulation model to assess the performance of diving fin was constructed by extending the swimming human simulation model SWUM. In this model, the fin was modeled as a series of five rigid plates. These plates were connected by virtual springs and dampers, and fin’s elasticity was represented by springs. The swimming motion with fins was obtained by motion capture experiment, and the fin parameters such as the shape, the rigidity, and the hydrodynamic coefficient, were determined by comparing the deformation and the drag measurement with the results of simulation. Finally, the diving swimming was simulated, and the swimming speed was found to be about 0.51 m/s. The fluctuations of swimming velocity were found to be 17.8% both in the simulation and the experiment. This consistency suggested the validity of the developed simulation model.

J2340105 ダイビングフィンの性能評価を目的としたシミュレーションモデルの構築

Scuba diving is becoming popular as a leisure activity. Although manufacturers make various types of diving fin, these developments are based on qualitative methods because of the difficulty in estimating the fin effeciency quantitatively. In this research, a simulation model to assess the performance of diving fin was constructed by extending the swimming human simulation model SWUM. In this model, the fin was modeled as a series of five rigid plates. These plates were connected by virtual springs and dampers, and fin's elasticity was represented by springs. The swimming motion with fins was obtained by motion capture experiment, and the fin parameters such as the shape, the rigidity, and the hydrodynamic coefficent, were determined by comparing the deformation and the drag measurement with the result of simulation. Finally, the diving swimming was simulated, and the swimming speed was found to be about 0.85 m/s. The speed was 10% higher compared with the actual diver's speed.

J027021 タコ型推進機構のシミュレーションモデルの開発([J027-02]生物規範メカニクス・システム:生物の運動,力学及びミメティクス(2))

The simulation model for the swimming movement of octopus was developed. This model was developed by extending SWUM (swimming human simulation model). In SWUM, a human body is represented by 21 body segments. However, it was not enough to realize the smooth arm motion of octopus. Therefore, in the octopus model, the number of body segments was increased to 90 (the body and 8 arms consisted of 10 body segments for each). Each joint had 3 degrees of freedom to realize various motions. This study aimed to clarify the motion and shape for fast swimming to develop an octopus-inspired robot for underwater rescue and exploration by using this octopus model. Three arm motions (sculling, undulation, and their composition) were analyzed. As a result, the average velocity in the last cycle of each motions were 0.08 lm/s in the sculling motion, 0.023m/s in the undulation motion and 0.106m/s in the composition motion. In future, we will investigate the fluid force coefficients to improve the precision of the model and investigate the arm motion suitable for the octopus-inspired robot.

208 3D-CGに基づく競泳用水着を着用したスイマーの筋骨格解析(筋骨格モデルの応用)

The objective of this study was to develop the method of musculoskeletal simulation for a swimmer attaching a competitive swimwear. For the simulation, the body geometry and joint motion were put into the swimming human simulation model SWUM in order to calculate the distributed fluid forces acting on the whole swimmer's body. The distributed fluid forces were put into the musculoskeletal model. In addition to the fluid forces, the forces induced by the tension of the swimwear were also considered. The distribution of the forces acting on the swimmer was calculated from the stress distribution of the swimwear. Since the swimwear targeted in the present study was a half spat type, the forces by the swimwear acted on three segments; the right and left thighs, and the pelvis. Therefore, the ratio of one distributed force to these three segments had to be determined. In order to determine this ratio, the deformation of the swimwear measured in an experiment was used. As an example, the simulation for the crawl stroke was conducted. The results showed the time-varying effect of the swimwear on the muscle activity during swimming.

3D-CG Based Musculoskeletal Simulation for a Swimmer Wearing Competitive Swimwear

The objective of this study was to develop a musculoskeletal simulation for a swimmer wearing competitive swimwear. For the simulation, the body geometry and joint motion were put into the swimming human simulation model SWUM in order to calculate the distributed fluid forces acting on the whole swimmer's body. The distributed fluid forces were put into the musculoskeletal model. In addition to the fluid forces, the forces induced by the tension of the swimwear were also considered. As an example, the simulation for the crawl stroke was conducted. The results showed the time-varying effect of the swimwear on the muscle activity during swimming.

Measurement and Modeling of Unsteady Fluid Force Acting on the Trunk of a Swimmer Using a Swimmer Mannequin Robot

The objectives of this study were to measure the unsteady fluid force acting on the trunk of a swimmer using the ‘swimmer mannequin robot’ and to model the fluid forces based on the formulation of the swimming human simulation model SWUM, which was developed by the authors’ group. The swimmer mannequin robot consisted of a swimmer mannequin and a driving mechanism. The scale of the swimmer mannequin was 1/2 (half scale) and the three-dimensional shape of an athlete swimmer taking the gliding position was reproduced in detail. The driving mechanism could move the mannequin in the pitching, heaving and rolling motions. Using the swimmer mannequin robot, the trunk motions of four strokes (crawl, breast, back and butterfly) were reproduced, and the unsteady fluid forces acting on the mannequin were measured by the dynamometers installed in the robot. On the other hand, the swimmer mannequin was modeled using the swimming human simulation model SWUM. The fluid force coefficients in the model were identified so that the simulated fluid forces became as consistent as possible with the experimental ones for each stroke case. The identified coefficients were then unified into ones which can be used for all cases. It was found that the precision of the model almost did not decrease as a result of the unification. It was also found that the overall performance of the simulation using the determined fluid force coefficients to predict the time variation of the fluid forces was satisfactory.

Simulation Analysis of the Mechanical Body Load during Walking in Water

The objective of this study was to clarify the mechanical body load during walking in water by the simulation, considering the fluid force acting on the lower extremities. First, the joint torque and force were analyzed by the simulation model of walking in water, which was developed by modifying the swimming human simulation model SWUM, and whose input data were obtained in the previous experiment. Second, the muscle force was analyzed by a two-dimensional musculoskeletal model for the lower extremities. The following findings were obtained from those analyses: The hip extensor muscles are employed more greatly during the stance phase of walking in water, compared to walking on land. The joint force in the compressing direction during walking in water is significantly lower than that of walking on land. From these results, it can be concluded that walking in water is suitable for rehabilitation. In addition, it was also found that the joint torque due to the fluid force component mainly acts in the swing phase. Therefore, the lower extremity muscles can be trained effectively by the use of the fluid force in the swing phase.

J0205-3-6 モノフィンスイミングのシミュレーション手法の開発([J0205-3]バイオメカニズム/バイオミメティクス/バイオインスパイアー(3))

The objective of this study was to develop a simulation method for monofin swimming which can consider the mechanical interaction between the monofin and swimmer. For this objective, SWUM (Swimming Human Simulation Model) was extended to the monofin swimming. The body geometry and joint motion of the swimmer were aquired from the experiment using a subject swimmer. In order to identify the bending stiffness and damping coefficient of the monofin, the static and dynamic bending tests of a monofin were conducted. In order to identify the fluid force coefficients of the monofin, the fluid force acting on the monofin was measured in the experiment. From the simulation of monofin swimming using the identified model parameters, it was found that the whole behavior of the swimmer and monofin in the simulation agrees well with that in the experiment.

Modeling Fluid Forces in the Dive Start of Competitive Swimming

The objective of this study was to construct a fluid force model for the dive start. The fluid force model is incorporated into a swimming human simulation model SWUM, which has been developed by the authors. In order to identify the fluid force coefficients in the model, dive starts performed by actual swimmers were filmed and reproduced by the simulation model. The fluid force coefficients were determined so that the swimmers’ movement in the simulation agreed with that of the experiment as much as possible. From simulations using the identified fluid force coefficients, it was found that simulation results of normal and ‘intentionally bad form’ trials for one subject agreed well with the experimental ones. This suggests that the proposed fluid force model would be valid to reproduce dive starts of various forms. For trials of another subject, the discrepancy between simulation and experiment becomes somewhat larger. This error can be reduced, however, by tuning the fluid force coefficients. This indicates that tuning the fluid force coefficients is effective to improve the accuracy of the simulation, and that the proposed model for the fluid forces will be valid for other swimmers.

Multi Agent/Object Simulation in Human Swimming

In human swimming, there are many situations in which the target of the analysis is not only the swimmer but also other swimmers or other equipment, such as synchronized swimming by multiple swimmers, swimming with fins, and so on. The objective of this study was to extend the swimming human simulation model SWUM so that it could handle the dynamic behavior of multiple rigid bodies and could freely describe the mechanical interactions among the rigid bodies. In this paper, the outline of SWUM is described first. The extension to the multi agent/object simulation and its implementation are explained next. Finally, its validity and usefulness are examined by discussing three examples of analysis. The first example was synchronized swimming by three swimmers. From the simulation results, it was found that the three swimmers gradually gathered, and finally the swimmers' upper limbs stably formed a hexagon. The second example was monofin swimming. The monofin was divided into five rigid plates in this simulation in order to represent its elastic deformation. From the simulation results, it was found that the swimmer accelerated at the moments of downward and upward kicks. The third example was the shooting motion in water polo. From the simulation results, it was found that the ball was smoothly released and flew in an appropriate direction. Since reasonable results were obtained in these examples of analysis, the validity and usefulness of the extension were confirmed.

Simulation Analysis of the Effect of Trunk Undulation on Swimming Performance in Underwater Dolphin Kick of Human

The objectives of this study were to investigate the effect of trunk undulation on the swimming performance in underwater dolphin kick, and to clarify the ideal trunk undulation form. The reference swimming motion of an elite swimmer was firstly acquired from the video analysis, and input into the swimming human simulation model SWUM, which had been developed by the authors. The trunk motion was next optimized by the simulation for three objective functions: maximizing swimming speed, maximizing propulsive efficiency, and minimizing fluid force acting on hands. The following findings were obtained: In the case of maximizing swimming speed, the whole body forms a ‘C’ shape due to the in-phase trunk undulation. The swimming motion of maximizing propulsive efficiency and the reference swimming are considerably similar to each other. In both cases, the trunk moves as a seesaw with a node; whereas, the lower limbs form a traveling wave in the absolute space. The values of propulsive efficiency are around 0.2 in the cases of maximizing propulsive efficiency, minimizing fluid force on hands, and the reference swimming. The swimming motion in the case of minimizing fluid force on hands is almost the same as that of maximizing propulsive efficiency. The trunk undulation with the appropriate amplitudes and phases, especially bending at the chest, is important in realizing the swimming motion which maximizes propulsive efficiency.

A-42 水泳人体シミュレーションモデルSWUMのマルチエージェント/オプジェクトシミュレーションヘの拡張(FD水泳)

A-42 水泳人体シミュレーションモデルSWUMのマルチエージェント/オプジェクトシミュレーションヘの拡張(FD水泳)

0210 水中ウォーキングにおける筋力強化に最適なフォームの提案 : 第1報,シミュレーションによる最適フォームの選定(J19-1 人間・ロボットの動作解析,J19 ヒューマンダイナミクス)

The objective of this study was to propose the optimal gait for muscle training during walking in water. First, the simulation method, which has been developed by the authors, was briefly described. The method consists of the inverse dynamics analysis part, which is a variation of the authors' swimming human simulation model SWUM, and the two-dimensional musculo-skeletal analysis part. Second, the evaluation index for the muscle training and the penalty index for the load on the joint were defined. Third, various gaits were analyzed by the simulation method. From the analysis, two gaits accompanying the large joint motions at the hip and knee were selected as the optimal gaits, by comparing the evaluation and penalty indices.

Mechanical Study of Standard Six Beat Front Crawl Swimming by Using Swimming Human Simulation Model

There are many dynamical problems in front crawl swimming which have not been fully investigated by analytical approaches. Therefore, in this paper, standard six beat front crawl swimming is analyzed by the swimming human simulation model SWUM, which has been developed by the authors. First, the outline of the simulation model, the joint motion for one stroke cycle, and the specifications of calculation are described respectively. Next, contribution of each fluid force component and of each body part to the thrust, effect of the flutter kick, estimation of the active drag, roll motion, and the propulsive efficiency are discussed respectively. The following results were theoretically obtained: The thrust is produced at the upper limb by the normal drag force component. The flutter kick plays a role in raising the lower half of the body. The active drag coefficient in the simulation becomes 0.082. Buoyancy determines the primal wave of the roll motion fluctuation. The propulsive efficiency in the simulation becomes 0.2.

Development of Swimming Human Simulation Model Considering Rigid Body Dynamics and Unsteady Fluid Force for Whole Body

The purpose of this study is to develop a swimming human simulation model considering rigid body dynamics and unsteady fluid force for the whole body, which will be utilized to analyze various dynamical problems in human swimming. First, the modeling methods and their formulations for the human body and the fluid force are respectively described. Second, experiments to identify the coefficients of the normal drag and the added mass are conducted by use of an experimental setup, in which a limb model rotates in the water, and its rotating angle and the bending moment at the root are measured. As the result of the identification, the present model for the fluid force was found to have satisfactory performance in order to represent the unsteady fluctuations of the experimental data, although it has 10% error. Third, a simulation for the gliding position is conducted in order to identify the tangential drag coefficient. Finally, a simulation example of standard six beat front crawl swimming is shown. The swimming speed of the simulation became a reasonable value, indicating the validity of the present simulation model, although it is 7.5% lower than the actual swimming.

An attempt to clarify the fastest stroke in freestyle swimming by simulation (Summary)

In this paper, clarification of the fastest stroke in freestyle swimming by simulation was attempted. That is, the front crawl swimming was optimized for swimming speed and propulsive efficiency using Genetic Algorithm (GA). First, GA was integrated into our swimming human simulation model SWUM, which was developed for analyzing swimming motion. The design variables in the optimization were joint angles of the upper limbs. In addition, in order to obtain reasonable results and to make the computation more efficient, three constraints were imposed on the optimizing algorithm. Then, the fastest and most efficient strokes were numerically solved. Both the obtained strokes have their own features distinguished from the conventional strokes. From the analyses of the joint torque and fluid force, it was found that the fastest stroke is explained as 'full power during whole stroke', while the most efficient as 'full power in efficient moment'