轮式车辆转向机构的遗传算法优化设计

Genetic Algorithm Based Optimization Design of Steering Mechanism in Wheeled Vehicles

在对车辆前轮转向机构进行运动分析的基础上,给出了参数优化设计的数学模型,提出了一种改进的遗传算法,并利用Matlab遗传算法工具箱对转向机构进行了优化设计,得到了满意的设计参数.讨论了遗传算法的一些关键技术:如改进的二进制编码方法,基于惩罚项的适应度计算,遗传算子的构造,以及遗传终止法则等.计算结果表明:与复合形法相比,遗传算法使目标函数值减少了6.6%,取得了较好的优化效果,从而更有利于减小转向车轮的滑移量和轮胎的磨损量,同时也验证了遗传算法解决多约束、多变量类型优化问题的可行性和优越性.

On the basis of kinematic analysis for steering mechanism in a certain wheeled vehicle, a mathematic model of optimal design is provided. An improved genetic algorithm （GA） is adopted to implement the conceptual optimization design and optimal parameters satisfying design requirement are obtained using the GA toolbox of Matlab. Some critical problems on GA are discussed in this research, such as an improved binary encoding method, calculation of fitness based on penalty function, construction of genetic operators, and the criteria for heredity termination, etc. Finally a comparison of optimization results between complex method and genetic algorithm is carried out, which shows the value of objective function has decreased 6. 6% under GA. This will help decrease the slippage of steering wheel and wear extent of tyre. At the same time, it also verifies the feasibility and superiority of genetic algorithm in solving the nonlinear optimization problems with many restraints and variables.