摘要
为了改善移动机器人Markov定位算法中方向传感器模型的性能,提出基于高斯函数的新概率模型.该模型考虑了方向角周期性问题,对相位进行了转换,利用高斯函数对方向传感器进行了概率建模.将此模型放入Markov算法,与其他传感器组成观测模型,并进行对称环境中的单次定位仿真和复杂环境中的连续定位仿真.仿真结果表明,这种概率模型计算量小,收敛速度快,在大量测量噪声存在下工作稳定.
To improve the performance of orientation sensor model in Markov localization algorithm for mobile robots, a new probabilistic model based on Gauss distribution was proposed. Due to the periodical characteristic of orientation angle, the model changed the orientation angle and modeled the sensor by Gauss distribution. Utilizing Markov algorithm based on the proposed sensor model, localization simulations for symmetrical environment and unstructured environment were carried out. The simulation results prove that compared with the existing models, Markov algorithm with this new model is characterized by rapid convergence, low computation cost and good robustness under strong sensor noise.
出处
《浙江大学学报（工学版）》
EI
CAS
CSCD
北大核心
2005年第3期339-341,353,共4页
Journal of Zhejiang University：Engineering Science
关键词
移动机器人
Markov定位算法
传感器建模
Algorithms
Computer simulation
Convergence of numerical methods
Markov processes
Mathematical models
Position measurement
State estimation
