Fault detection method based on holohedral polytope filter

Abstract

The invention discloses a fault detection method based on holohedral polytope filter, and belongs to the field of the fault diagnosis. The method comprises the following steps: establishing a system model of a wireless sensor network system according to a state vector of the wireless sensor network system; determining observation vector according to the system model, iteratively determining the holohedral polytope at the k moment; predicting the estimated holohedral polytope at the k+1 moment according to the holohedral polytope at the k moment; acquiring the observation vector at the k+1 moment, computing a consistency state set of the wireless sensor network at the k+1 moment according to the observation vector at the k+1 moment; detecting whether the consistency state set at the k+1 moment is intersected with the estimated holohedral polytope at the k+1 moment; if detecting that the consistency state set at the k+1 moment is free from intersection with the estimated holohedral polytope at the k+1 moment, determining that the wireless sensor network system at the k+1 moment is in fault. The problem that the wireless sensor network is low in instantaneity while detecting the faultis solved; the practicability of the detection method is increased, and the fault detection efficiency and accuracy rate are improved.

Description

technical field [0001] The embodiments of the present invention relate to the field of fault diagnosis, and in particular to a fault detection method based on fully symmetrical multi-cell filtering. Background technique [0002] A wireless sensor network is a fully distributed system without a central node. Through random placement, many sensor nodes are deployed in predetermined areas. The sensor nodes placed in a predetermined area have sensors, data processing units and communication modules, and they form a network system self-organized through wireless connections. If the wireless sensor network fails, it may cause serious losses, so it is necessary to detect the fault of the wireless sensor network system in real time and accurately. [0003] Generally, faults can be detected by using the membership estimation method, which uses the state feasible set including the system model, measurement data, interference and noise boundaries to describe the real state of the sys...

AI Technical Summary

Problems solved by technology

In general, the state feasible set is an irregular convex polyhedron, which is difficult to describe with a mathematical model. Therefore, it is necessary to find a set that can approximately describe the state feasible set, such as ellipsoid, interval, parallel polyhedron, ordinary polyhedron, etc. However, ellipsoid Spheres, intervals, and parallel polyhedrons are less accurate and more conservative, and the dimensionality of ordinary polyhedrons will increase with the number of iterations, resulting in increased computational complexity

Images