| 摘要: |
| 【目的】提出一种基于改进型Bug算法的自动导航避障方法，为提高林间移动平台作业时的避障能力提供参考。【方法】通过6个超声波传感器将林间移动平台四周分为9个探测区域，根据各区域内的车-障距离确定相应的避障策略；改进传统Bug算法，将避障过程分成路径追踪、避行和绕行3个阶段；在避行阶段，建立避障转向角索引图，同时根据转向延迟试验获得转向时机的提前量，将该提前量与索引图结合，使林间移动平台在合适的时机使用最大转向角避障；在绕行阶段，通过设计带自调节函数的模糊控制器，控制林间移动平台绕障碍物边缘行驶。最后根据林间环境特点，设计大、小2种障碍物试验，分别对避障时机、绕行偏差和总路程等参数进行统计与分析。【结果】林间环境大、小障碍物避障试验表明：林间移动平台提前避障距离分别为93.9和81.7 cm，平台前方的2个传感器的探测距离在索引图范围内，可以精准实现提前避障；大、小障碍物避障总路程与理论路程之差分别占理论值的0.64%和0.42%，平均绕行偏差分别为-15.2和-7.9 cm，绕行标准差分别为7.7 和12.5 cm，可以实现稳定绕障。【结论】所建立的基于改进型Bug算法的自动导航避障方法，避障路径冗余小且行驶稳定，满足了林间环境运输及其他移动作业的要求。 |
| 关键词: 林间移动平台 移动作业 超声波避障 自调节模糊控制 自动导航 Bug算法 |
| DOI： |
| 分类号: |
| 基金项目:浙江省基础公益研究计划项目（LGN18F030001，LGN18C200017） |
|
| Obstacle avoidance method for inter-forest mobile platform based on improved Bug algorithm |
|
CAI Cuncheng,YAO Lijian,CHAI Shanpeng,et al
|
| Abstract: |
| 【Objective】In this paper,an automatic navigation obstacle avoidance method based on improved bug algorithm was proposed to improve obstacle avoidance ability of forest mobile platform.【Method】Six ultrasonic sensors were used to divide periphery of the forest mobile platform into nine detection areas,and the corresponding obstacle avoidance strategy was determined in accordance with the distance between mobile platform and obstacle in each area.Through the upgrading of traditional Bug algorithm,the obstacle avoidance process was divided into three stages of tracking,avoiding and detouring.In the avoidance stage,the index map of obstacle avoidance steering angle was established,and the steering timing advance was also acquired by steering delay test.Then,by combining timing advance with index map,forest mobile platform was able to use the maximum steering angle to avoid obstacles at the right time.In the bypass stage,a fuzzy controller with self adjusting function was designed to control the platform to travel around the edge of the obstacle.Finally,according to the characteristics of forest environment, two sizes of obstacles were designed to statistically analyze the distance between forest mobile platform and obstacle,detour deviation as well as total distance.【Result】The experiment of avoiding large and small obstacles in forest environment showed that the distances between forest mobile platform and obstacle avoidance in advance were 93.9 and 81.7 cm,respectively.The detection distance of the two ultrasonic sensors in front of the platform was within the scope of the index map,and it was feasible to accurately avoid obstacles in advance.The differences between obtained and theoretical total distances of obstacle avoidance for large and small obstacles were 0.64% and 0.42% of the theoretical values,respectively.With average detour deviations of -15.2 and -7.9 cm,and standard deviations of 7.7 and 12.5 cm,stable obstacle avoidance was realized in this study.【Conclusion】The established automatic navigation obstacle avoidance method based on the improved bug algorithm had short obstacle avoidance path redundancy and stable driving condition,which met the requirements of forest environment transportation and other mobile operations. |
| Key words: forest mobile platform mobile operation ultrasonic obstacle avoidance self-adjusting fuzzy control automatic navigation Bug algorithm |
