Optimizing the actuation of musculoskeletal model by genetic algorithm to simulate the vertical jump

In human body movement simulation such as vertical jump by a forward dynamic model, optimal control theories must be used. In the recent years, new methods were created for solving optimization problems which they were adopted from animal behaviors and environment events such as Genetic algorithm, Particle swarm and Imperialism competitive. In this work, the skeletal model was constructed by Newton-Euler equation of motion. This 2D model has 4 rigid segments that include foot, shank, thigh and HAT (Head, Arm and Trunk) and all joints were assumed to be revolute and ideal. Also 20 effective muscles in vertical jump were constructed as joint actuator. The ground reaction force was simulated by a spring-damper element. Additionally, joints ligament were constructed to simulate the joint out of range motion. The Genetic algorithm was used to generate the best muscle excitation for maximum height in vertical jumping and the generated muscles excitations were converted to muscles activations. The muscles activations were applied to muscles model to generate muscles force. The maximum height of jump was considered as a criteria function of optimization problem. The designed genetic algorithm could control the musculoskeletal and simulate the vertical jump movement. The result showed that the height of center of mass was equal to 121.67 cm after 533 iterations. It is looks to be able to obtain better result provided to increase the iteration or combining clever algorithms together.