当前位置: X-MOL 学术IEEE Trans. Autom. Control › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Entropy and Minimal Bit Rates for State Estimation and Model Detection
IEEE Transactions on Automatic Control ( IF 6.2 ) Pub Date : 2018-10-01 , DOI: 10.1109/tac.2017.2782478
Daniel Liberzon , Sayan Mitra

We study a notion of estimation entropy for continuous-time nonlinear systems, formulated in terms of the number of system trajectories that approximate all other trajectories up to an exponentially decaying error. We also consider an alternative definition of estimation entropy, which uses approximating functions that are not necessarily trajectories of the system, and show that the two entropy notions are equivalent. We establish an upper bound on the estimation entropy in terms of the sum of the desired convergence rate and an upper bound on the matrix measure of the Jacobian, multiplied by the system dimension. A lower bound on the estimation entropy is developed as well. We then turn our attention to state estimation and model detection with quantized and sampled state measurements. We describe an iterative procedure that uses such measurements to generate state estimates that converge to the true state at the desired exponential rate. The average bit rate utilized by this procedure matches the derived upper bound on the estimation entropy, and no other algorithm of this type can perform the same estimation task with bit rates lower than the estimation entropy. Finally, we discuss an application of the estimation procedure in determining, from the quantized state measurements, which of two competing models of a dynamical system is the true model. We show that under a mild assumption of “exponential separation” of the candidate models, detection always happens in finite time.

中文翻译:

状态估计和模型检测的熵和最小比特率

我们研究了连续时间非线性系统的估计熵概念,该概念根据系统轨迹的数量来表述,这些轨迹近似于所有其他轨迹,直至指数衰减误差。我们还考虑了估计熵的另一种定义,它使用不一定是系统轨迹的逼近函数,并表明这两个熵概念是等效的。我们根据所需收敛速度和雅可比矩阵度量的上限之和建立估计熵的上限,乘以系统维数。估计熵的下界也被开发出来。然后,我们将注意力转向具有量化和采样状态测量的状态估计和模型检测。我们描述了一个迭代过程,该过程使用此类测量来生成以所需指数速率收敛到真实状态的状态估计。该过程使用的平均比特率与估计熵的导出上限相匹配,并且没有其他此类算法可以以低于估计熵的比特率执行相同的估计任务。最后,我们讨论了估计程序在根据量化状态测量确定动态系统的两个竞争模型中的哪一个是真实模型中的应用。我们表明,在候选模型的“指数分离”的温和假设下,检测总是在有限的时间内发生。该过程使用的平均比特率与估计熵的导出上限相匹配,并且这种类型的其他算法无法以低于估计熵的比特率执行相同的估计任务。最后,我们讨论了估计程序在根据量化状态测量确定动态系统的两个竞争模型中的哪一个是真实模型中的应用。我们表明,在候选模型的“指数分离”的温和假设下,检测总是在有限的时间内发生。该过程使用的平均比特率与估计熵的导出上限相匹配,并且这种类型的其他算法无法以低于估计熵的比特率执行相同的估计任务。最后,我们讨论了估计程序在根据量化状态测量确定动态系统的两个竞争模型中的哪一个是真实模型中的应用。我们表明,在候选模型的“指数分离”的温和假设下,检测总是在有限的时间内发生。动力系统的两个相互竞争的模型中,哪一个是真实模型。我们表明,在候选模型的“指数分离”的温和假设下,检测总是在有限的时间内发生。动力系统的两个相互竞争的模型中,哪一个是真实模型。我们表明,在候选模型的“指数分离”的温和假设下,检测总是在有限的时间内发生。
更新日期:2018-10-01
down
wechat
bug