fault2A_xts: Process Data under a System Drift Fault

In gabrielodom/mvMonitoring: Multi-State Adaptive Dynamic Principal Component Analysis for Multivariate Process Monitoring

Process Data under a System Drift Fault

Description

Three-feature, three-state simulated process data including observations under normal operating conditions and observations after a positive drift in values for each feature in the system.

Usage

fault2A_xts

Format

An xts data matrix with 10080 rows and four columns, corresponding to one week worth of data recorded at a 1-minute interval. The columns under normal conditions are defined in the help file for normal_switch_xts. The fault is a drift on each feature by s / 10 ^ 3, where s is the observation index. The fault starts at row 8500, and the four columns under the fault state are defined here:

state :

the state indicator for the multivariate system, with three levels

x :

x(t) = t + drift + error

y :

y(t) = t ^ 2 - 3t + drift + error

z :

z(t) = - t ^ 3 + 3t ^ 2 + drift + error

where t is a 10080-entry vector of autocorrelated and non-stationary hidden process realizations. The states alternate each hour and are defined as follows:

State1 –

As presented

State2 –

Rotated by (yaw = 0, pitch = 90, roll = 30) and scaled by (1 * x, 0.5 * y , 2 * z).

State3 –

Rotated by (yaw = 90, pitch = 0, roll = -30) and scaled by (0.25 * x, 0.1 * y , 0.75 * z).

See the vignette for more details.

Source

Simulated in R.

gabrielodom/mvMonitoring documentation built on Nov. 23, 2023, 6:39 p.m.