Autonomous system (mathematics)

About: Autonomous system (mathematics) is a research topic. Over the lifetime, 1648 publications have been published within this topic receiving 38373 citations.

PN Inference Based Autonomous Sequential Restoration of Distribution System Under Natural Disaster

Abstract: Potential challenge during a disaster is the wide range of disturbance caused to the power system network. The article attempts to develop a systematic procedure to progressively restore the network by recognizing the alive network. During disaster there is manpower shortage and poor road connectivity around the affected area. Hence, an autonomous priority-based restoration policy is required for critical load restoration. The main idea of the proposed article is to develop an autonomous system, which is decision-based restoration system by identifying the network component still available after disaster followed by reconfiguration confirming the proposed load pickup procedure. Because of the inherent capability of implementing the condition-based state transition of switches and parallel processing, a petri net (PN) based methodology has been used in the article to integrate alive parts, which further assists in identifying the best switching schedule of circuit breakers for autonomous load restoration. The primary contribution of the article is to develop an encapsulated algorithm using PN as a modeling platform of components as well as empowering the same by introducing the game-theory-based concept in the formulation of different microgrids integrated with distributed generators during extreme weather events. The verification of effective autonomous restoration is then realized by using the available delivery capability and design verification of the framed microgrids, and the comparison with different restoration strategies adopted by other authors. The studies are carried out with the IEEE 34, 37, and 123-bus feeder test systems for implementation and validation of the proposed article along with the comparison with other published works.

Hybrid energy harvesting for autonomous sensors in building automation

Abstract: For researches on energy harvesting in building automation two independent power sources have been examined. Therefor long term power profiles have been analysed, which show the available energy per transducer over time. These profiles have been investigated according to their ability to supply an energy autonomous system and to improve the availability and reliability of operation. To reduce the failing time of a sensor system by using two independent sources within a hybrid energy harvesting supply, is one of the general aims for this contribution. An energy management circuit has been developed and compared to the state of art solution of combining different energy sources. The additional losses of this method have been simulated. This contribution tends on a method to improve the reliability of an energy autonomous system using two different and independent energy sources and an optimized combination of both within an adapted energy management strategy.

Deep CNN-based autonomous system for safety measures in logistics transportation

Abstract: The careless activity of drivers in logistics transportation is a primary reason inside the vehicle during road accidents. This research aims to reduce the number of accidents caused by a failure of the driver in logistics transportation by incorporating an autonomous system. We propose a convolutional neural network -based architecture to recognize and classify different positions which cause road accidents. The proposed system is evaluated with the State Farm Distracted Driver Database, which included examples illustrating ten different driving positions like reaching behind and talking to the passenger, making up, safe driving, talking on the phone, clothing, checking right/left hand, right/left hand, and running the radio. The proposed approach has also been tested against recent algorithms and evaluated. Our model has obtained 98.98% accuracy compared to other types of approaches with different descriptors and classification techniques

Heterogeneous multistability in a novel system with purely nonlinear terms

Abstract: In this paper, the dynamic properties of the new 3-Dimensional autonomous system without linear terms are investigated. The nonlinear features of the explored model are highlighted in terms of bifu...

Continuous Probabilistic Mapping by Autonomous Robots

Abstract: In this paper, we present a new approach for continuous probabilistic mapping The objective is to build metric maps of unknown environments through cooperation between multiple autonomous mobile robots The approach is based on a Bayesian update rule that can be used to integrate the range sensing data coming from multiple sensors on multiple robots In addition, the algorithm is fast and computationally inexpensive so that it can be implemented on small robots with limited computation resources The paper describes the algorithm and illustrates it with experiments in simulation and on real robots