Study on a class of Sensor Fault Detection for Systems
with Multiplicative Noise
Abstract
Systems with multiplicative noise have largely promoted the technology of
modern signal processing. In recent years, with the accuracy of modeling and data
acquisition required in the field of underwater target tracking and petroleum seismic
exploration, estimation theory of systems with multiplicative noise is increasingly
showing its usefulness and attracting more and more attention of scholars at home and
abroad. However, the noises of the systems are portrayed more accurate and fine,
which makes the model of the systems more complex than ever before. So the theory
and application of systems with multiplicative noise are not perfect in many respects.
Sensor fault detection is an important branch of modern control theory, which comes
up an endless stream of theoretical results in last decades. The main focus of this
paper is on resolving the problem of a class of sensor fault detection for systems with
multiplicative noise. At the same time, the paper makes an in-depth study in detecting
targets at sea with high-frequency surface wave radar and inquires into an algorithm
for the detection based on systems with multiplicative noise while interfered by
complex and uncertain disturbances in the marine environment. Simulation and
comparative experiment results illustrate the effectiveness of the above algorithms.
Here are the main works of this paper:
1. As to the necessity and complexity of sensor fault detection in the field of
underwater target tracking and petroleum seismic exploration, this paper proposes a
new sensor fault model based on systems with multiplicative noise and builds an
appropriate fault indicator to manage the fault detection. Different from the traditional
methods restricted with the Gaussian distribution characteristics of the innovation, the
new methodology is more flexible, which can improve the accuracy of fault detection
as well.
2. Due to the great impact made by a variety of strong noises on high-frequency
surface wave radar, real-time and accuracy of ship targets detection at sea are
seriously affected. This paper makes an analysis of two clutters, ionospheric clutter
and sea clutter, which affect surface wave radar mostly and introduces a model of
radar target detection with multiplicative noise and proposes a new method of
real-time detection for ship targets based on the theory of fault detection.
