摘要
针对机器人小车控制过程中的轨迹跟踪问题,以控制量为离散值的轮式小车为研究对象,提出一种新的预测控制算法。建立小车在离散状态空间下的运动学模型,并根据此模型设计预测控制算法,以克服实际过程中的不确定性。然后,为解决传统预测控制算法在应用上出现的计算量指数增长问题,基于改进模拟退火的快速寻优算法,设计一种新的预测控制策略,以同时保证小车轨迹跟踪的精确性与实时性。通过仿真实验给出了该算法下小车对不同轨迹的跟踪情况及鲁棒性测试,在与传统预测控制算法计算量的比较结果中表明,该算法能够减少计算时间且实现对轨迹有效地跟踪,并保证较高的稳定性,同时,该算法可以推广到各类控制量为离散值的预测控制问题。
A predictive control method is presented to solve the tracking problem of wheeled vehicle with discrete control value. A kinematic model of the vehicle is constructed in discrete state space, and a predictive control algorithm is designed. Predictive control strategy, based on the improved of fast optimization algorithm SA is proposed to guarantee the high tracking accuracy in real-time, and overcome the problem of the exponential raising in computing in traditional ways. The simulation results show the effectiveness of the proposed algorithm, and meanwhile, this method can also be extended to all the predictive control problems with discrete control value similarly.
出处
《控制工程》
CSCD
北大核心
2009年第6期713-716,763,共5页
Control Engineering of China
基金
北京市自然科学基金资助项目(4062030)
关键词
控制量离散
预测控制
改进SA
贪心算法
discrete control value
predictive control
improved SA algorithm
greed algorithm
