An alpha-uniformized Markov chain is defined by the concept of equivalent infinitesimalgenerator for a semi-Markov decision process (SMDP) with both average- and discounted-criteria.According to the relations of their...An alpha-uniformized Markov chain is defined by the concept of equivalent infinitesimalgenerator for a semi-Markov decision process (SMDP) with both average- and discounted-criteria.According to the relations of their performance measures and performance potentials, the optimiza-tion of an SMDP can be realized by simulating the chain. For the critic model of neuro-dynamicprogramming (NDP), a neuro-policy iteration (NPI) algorithm is presented, and the performanceerror bound is shown as there are approximate error and improvement error in each iteration step.The obtained results may be extended to Markov systems, and have much applicability. Finally, anumerical example is provided.展开更多
采用频率测量实现目标定位具有成本低、可靠性高的特点,仅利用到达频率差(frequency difference of arrival,FDOA)测量,提出了一种静态目标位置的精确定位方法。针对所建立的频率测量方程的高度非线性这一问题,通过引入辅助变量,将其转...采用频率测量实现目标定位具有成本低、可靠性高的特点,仅利用到达频率差(frequency difference of arrival,FDOA)测量,提出了一种静态目标位置的精确定位方法。针对所建立的频率测量方程的高度非线性这一问题,通过引入辅助变量,将其转化为矩阵形式的伪线性方程;然后利用半正定松弛(semi-definite relaxation,SDR)方法将非凸的加权最小二乘(weighted least square,WLS)问题松弛为半正定规划(semidefinite programming,SDP)问题,从而进一步精确估计未知变量;最后对所提出方法的均方根误差(rootmean-square error,RMSE)进行了分析,以验证其性能。仿真结果表明,在较低的高斯噪声水平下,所采用的半正定松弛方法的性能能够达到克拉美罗下界(Cramer-Rao lower bound,CRLB),且该算法对几何形状具有较高的鲁棒性;此外,在使用较少数量的传感器时,其RMSE性能要优于两阶段加权最小二乘(two-stage weighted least square,TSWLS)法。展开更多
文摘An alpha-uniformized Markov chain is defined by the concept of equivalent infinitesimalgenerator for a semi-Markov decision process (SMDP) with both average- and discounted-criteria.According to the relations of their performance measures and performance potentials, the optimiza-tion of an SMDP can be realized by simulating the chain. For the critic model of neuro-dynamicprogramming (NDP), a neuro-policy iteration (NPI) algorithm is presented, and the performanceerror bound is shown as there are approximate error and improvement error in each iteration step.The obtained results may be extended to Markov systems, and have much applicability. Finally, anumerical example is provided.
文摘采用频率测量实现目标定位具有成本低、可靠性高的特点,仅利用到达频率差(frequency difference of arrival,FDOA)测量,提出了一种静态目标位置的精确定位方法。针对所建立的频率测量方程的高度非线性这一问题,通过引入辅助变量,将其转化为矩阵形式的伪线性方程;然后利用半正定松弛(semi-definite relaxation,SDR)方法将非凸的加权最小二乘(weighted least square,WLS)问题松弛为半正定规划(semidefinite programming,SDP)问题,从而进一步精确估计未知变量;最后对所提出方法的均方根误差(rootmean-square error,RMSE)进行了分析,以验证其性能。仿真结果表明,在较低的高斯噪声水平下,所采用的半正定松弛方法的性能能够达到克拉美罗下界(Cramer-Rao lower bound,CRLB),且该算法对几何形状具有较高的鲁棒性;此外,在使用较少数量的传感器时,其RMSE性能要优于两阶段加权最小二乘(two-stage weighted least square,TSWLS)法。
