基于混合遗传算法的手指肌肉力优化求解方法研究

Research on Method of Finger Muscle Force Optimization Based on Hybrid Genetic Algorithms

目的为解决手指肌肉、肌腱组织结构复杂导致肌肉力真实值很难获取的问题,采用混合遗传算法求解手指肌肉力。方法在分析手指解剖结构的基础上,基于生物力学原理,建立了手指肌肉、肌腱力和接触力的力学平衡方程,以最小化肌肉应力平方和为优化目标函数,利用遗传算法和增广拉格朗日乘子法相融合的混合遗传算法,采用典型对象的手指接触力测量数据为输入参数,优化计算求解手指的肌肉力。结果有效求出了手指的各个肌肉力,优化计算结果与其它研究的结果较为吻合;随着手指关节角减小,FDP和FDS的数值增大,而LE、LU、UI、RI的数值随之减小,FDP、FDS、LE和RI+UI+LU在数值上约为外界接触力的0.7~3.7倍。结论基于遗传算法和增广拉格朗日乘子法的混合优化算法,初始优化不依赖初值的选取,避免了经验取值,提高了算法的适用性。

Objective The real value of finger muscle force is difficult to obtain due to the complex structure of finger muscle and tendon.A hybrid genetic algorithm was utilized to calculate the finger muscle force.Methods Based on analysis of the finger anatomical structure and biomechanical principle,the mechanical equilibrium equations of finger muscle,tendon force and contact force were established.The objective function was to minimize the sum of squares of muscle stress.The hybrid genetic algorithm,which combined the genetic algorithm with augmented Lagrange method,was adopted to optimize the calculation of finger muscle force by taking the measured data of finger contact force for typical objects as the input parameters.Results The results of optimization calculation were in good agreement with those of other studies.With the decrease of finger joint angle,the values of FDP and FDS increased,while the values of LE,LU,UI and RI decreased.The values of FDP,FDS,LE and RI+UI+LU were about 0.7~3.7 times of the external contact force.Conclusion The hybrid optimization algorithm based on genetic algorithm and augmented Lagrange multiplier method does not depend on the selection of initial values,avoids empirical values and improves the applicability of the algorithm.