We present a novel method to generate solutions with required probability distribution for the turning time of a four-state first-order Markov Chain. We do this by using the stable distribution function and numerical simulation and show that the obtained chi-square probability distribution satisfactorily models the data. We also compare the results obtained using the proposed method with those obtained when using a known random number generator. We conclude that the proposed method is robust, simple, and highly accurate.
This paper presents a new class of switched systems with the objective to drive the system state from zero to a target set. Unlike well literatures that presented the result regarding switched systems where the state approaches to the targets, this paper presents a generic way to control a switched system such that a finite time and a minimum distance to the target set is guaranteed. The comparison between our method and the well known zero trace control framework is also presented.
The system of resource allocation is a general class of problems faced by all distributed systems. In this paper, we propose a distributed algorithm for solving the resource allocation problem based on the concept of distributed convex optimization. The main algorithm for solving the resource allocation problem is a divide-and-conquer algorithm. At each iteration, each degree of freedom is allocated among a set of selected agents. For each agent, the resource allocation problem is solved using a subroutine for distributed convex optimization, which is closely related to the numerical convex optimization algorithm. The main algorithm is proved to be optimal asymptotically when the number of agents tends to infinity.
This paper proposes a control method for autonomous navigation of a battery operated car that is a subject to wheel slip. In literature, methods like planning all the road dynamics and wheel slip forces using time-series input-series models are extensively being used for autonomous navigation of a battery operated car. These conventional methods mostly require a huge number of data for models and this becomes a problem for practical applications. Also, these methods are not necessarily adaptive to road conditions, since they require inputs only at a fixed time and at a fixed frequency. d2c66b5586