基于遗传优化的无人车横向模糊控制

Study on Lateral Fuzzy Control of Unmanned Vehicles Via Genetic Algorithms

以视觉导航式无人车DLUIV-1为控制对象,对其进行横向运动控制研究。建立视觉导航式无人车横向运动控制系统模型,对无人车的横向运动行为进行描述。在此基础上,分析预瞄距离对横向控制系统动态性能的影响,并建立考虑速度因素的预瞄距离计算公式。针对无人车具有非完整运动约束、高度非线性动态特性以及参数的不确定性等特点,提出基于遗传算法的无人车横向模糊控制策略,通过遗传算法对横向模糊控制器的隶属度函数参数和控制规则的自动优化,从而有效地确定出横向模糊控制器的隶属度函数和控制规则。最后通过仿真和实车试验对该横向模糊控制器进行验证和评价,仿真和试验结果表明,该横向控制器可保证无人车稳定准确地跟踪参考路径,且具有较强的鲁棒性。

Lateral motion control of vision-based unmanned vehicle DLUIV-1 is researched.Firstly,the lateral motion control model of vision-based navigation unmanned vehicle is established,which can describe the lateral motion behavior of unmanned vehicles.On this basis,the influence of look-ahead distance on lateral control system is analyzed,and the formula of look-ahead distance with the effect of velocity considered is established.Then,in view of highly nonlinear and parametric uncertain properties of non-holonomic unmanned vehicle,an efficient fuzzy control strategy based on genetic algorithm is presented,the membership functions and rules of fuzzy controller are optimized by genetic algorithm,therefore,the membership functions and rules are effectively determined.Finally,the simulation and real vehicle experiment analysis and validation of proposed control strategy are carried out.Both simulation and experimental results show that the unmanned vehicle controlled by the proposed control strategy can stably track the reference path under various conditions with high accuracy and robustness.