An Architecture for Wireless Sensors in Distributed Management of Electrical Distribution Systems
作者:
Mikael Nordman
所属专业方向:
EE
摘要:
Abstract
This thesis presents a distributed architecture for the use of wireless sensors in the
management of electrical distribution systems. Although the general concept of using
wireless sensors for measuring quantities of power lines has previously been introduced,
the proposed solutions have not been well integrated within the power system
equipment and the automation system. This has severely compromised the applicability
of the sensors in the field. However, wireless sensors have several features that make
them an attractive instrumentation solution in the harsh environment of electrical
distribution networks. Wireless sensors do not need signal or power cables and they are
therefore easy to install and use in system refurbishment. They thus provide an
interesting and cost effective alternative that is worth studying.
In this thesis, the concept developed for enhancing the applicability and integration of
wireless sensors within the power distribution system is based on an architecture that
uses the properties of the system itself to preserve functionality and to minimise the
power consumption and thus the size and weight of the wireless sensors. In this
architecture the following essential components and aspects are identified.
First, a time synchronisation method for wireless sensors is developed. The main goal
with this method is to provide energy optimised means to determine the phase angle in
wireless sensor cells where several sensors measure phase current and one sensor
measures voltage. Secondly, a fault detection and location method with wireless sensors
is proposed. This method enables implementation of fault management using wireless
sensors that operate autonomously and asynchronously and measure only phase current.
With this approach the hardware and software implementation of a sensor can be
simplified compared to previous solutions. The combination of these developed
methods enables the construction of lighter, smaller and more flexible wireless sensors.
Consequently, the ability to integrate the sensors within components and equipment as
well as their robustness and functionality are enhanced.
Third, an architecture for the distributed management of electrical distribution networks
is developed. The primary intention with this architecture is to provide a distributed
computing platform that integrates the wireless sensors within the automation system
and manages the related functionality in a controlled and simple way. The developed
architecture also creates a generic concept for distributed and local control in the
management of electrical distribution systems. This concept will be valuable in the
future when the vast amount of data produced by sensors must be locally processed to
avoid saturating the communication and computational resources of the control centre.