论文部分内容阅读
针对仿人机器人NAO实际行走距离受环境因素(如脚底打滑)和自身因素(如关节的反作用力)的影响,导致其与理论期望距离有一定偏差的情况,提出了一种基于自适应神经模糊推理系统的行走距离预测方法.通过获得NAO行走距离的相关数据,采用网格分割法,构建自适应模糊推理系统来预测机器人实际行走的距离.训练时采用网格分割、BP算法和最小二乘算法的组合优化.对训练后的系统进行了分析与测试,仿真实验证明了方法的有效性.
In view of the fact that the actual walking distance of humanoid robot NAO is affected by environmental factors (such as foot slip) and its own factors (such as the joint reaction force), which leads to a certain deviation from the theoretical expected distance, a method based on adaptive neuro-fuzzy Inference system of walking distance prediction method.By obtaining the NAO walking distance of the relevant data, the grid segmentation method, the construction of adaptive fuzzy inference system to predict the robot’s actual walking distance.Training using grid segmentation, BP algorithm and least squares Algorithm optimization and combination.After the training system is analyzed and tested, the simulation experiment proves the effectiveness of the method.