Adaptive Dynamic Programming based Control Scheme for Uncertain Two-Wheel Robots

<div>The paper addresses the problem of effectively controlling a two-wheel robot given its inherent non-linearity and parameter uncertainties. In order to deal with the unknown</div><div>and uncertain dynamics of the robot, it is proposed to employ the adaptive dynamic programming, a reinforcement learning based technique, to develop an optimal control law. It is interesting that the proposed algorithm does not require kinematic parameters while finding the optimal state controller is guaranteed. Moreover, convergence of the optimal control scheme is theoretically proved. The proposed approach was implemented in a synthetic</div><div>two-wheel robot where the obtained results demonstrate its</div><div>effectiveness.</div>

Adaptive Dynamic Programming based Control Scheme for Uncertain Two-Wheel Robots

<div>The paper addresses the problem of effectively controlling a two-wheel robot given its inherent non-linearity and parameter uncertainties. In order to deal with the unknown</div><div>and uncertain dynamics of the robot, it is proposed to employ the adaptive dynamic programming, a reinforcement learning based technique, to develop an optimal control law. It is interesting that the proposed algorithm does not require kinematic parameters while finding the optimal state controller is guaranteed. Moreover, convergence of the optimal control scheme is theoretically proved. The proposed approach was implemented in a synthetic</div><div>two-wheel robot where the obtained results demonstrate its</div><div>effectiveness.</div>

“…This Area Is almost the only Classical Country in Our Possession”: Baron B. B. Kampengauzen’s Memoir “On the Antiquities in the New Russia Region” (1817)

This publication introduces into the scholarship Baron B. B. Kampengauzen’s (Campenhausen, 1772–1823) memoir “On the Antiquities in the New Russia Region.” The state controller of the Russian Empire Kampengauzen visited the New Russia Region and the Crimea in summer 1816. In result of this trip, he prepared a long memoir discussing possible transformation of the country. Kampengauzen compiled the part addressing the antiquities of New Russia in a traditional way of the observations of the kind. In the beginning, he stated the general history of the country; later on, he called the reader’s attention to the topical problems of New Russia, discussed the current status of the ancient sites, and expressed his own recommendations for the protection and research of antiquities. This memoir is especially valuable since one of its first readers was the Russian Emperor Alexander I.

Verifiable RNN-Based Policies for POMDPs Under Temporal Logic Constraints

Recurrent neural networks (RNNs) have emerged as an effective representation of control policies in sequential decision-making problems. However, a major drawback in the application of RNN-based policies is the difficulty in providing formal guarantees on the satisfaction of behavioral specifications, e.g. safety and/or reachability. By integrating techniques from formal methods and machine learning, we propose an approach to automatically extract a finite-state controller (FSC) from an RNN, which, when composed with a finite-state system model, is amenable to existing formal verification tools. Specifically, we introduce an iterative modification to the so-called quantized bottleneck insertion technique to create an FSC as a randomized policy with memory. For the cases in which the resulting FSC fails to satisfy the specification, verification generates diagnostic information. We utilize this information to either adjust the amount of memory in the extracted FSC or perform focused retraining of the RNN. While generally applicable, we detail the resulting iterative procedure in the context of policy synthesis for partially observable Markov decision processes (POMDPs), which is known to be notoriously hard. The numerical experiments show that the proposed approach outperforms traditional POMDP synthesis methods by 3 orders of magnitude within 2% of optimal benchmark values.

SYNTHESIS AND ANALYSIS OF THE CONCENTRATION CONTROL SYSTEM IN THE CHEMICAL REACTOR ON THE BASIS OF THE STATE CONTROLLER

The paper deals with the synthesis of a singlechannel control system of concentration vector in a liquidphase chemical reactor using a state controller. Settings of the state controller are determined using the modal control method. An integral component was introduced into the structure of the state controller to eliminate the static control error and ensure the robustness of the control system.