Sensor data faults - code

Ash

2012 9 28

0 个回答

4 次查看（30 天）

0 个投票

Hello,

I am developing a fault analyzer for data from wireless sensors. At the moment I am using random data I have created in Matlab at the moment. As part of my self study, I intend to detect the following data faults.

1. Outliers

2. Spikes

Defined as the rate of change much greater than expected over a short period of time where the signal may pr may not return to normal afterwards

3. Stuck-at fault Defined as a serioes of sensor data values that experience zero or almost zero variation for a period of time greater than expected

4. Noisy signal.

Detected by observing the signal variance and by correlating sensor data with its neighbours

I kindly request you to assist me with a matlab code that detects each of these data faults.

Is it possible to build a fault analyzer that runs at once and reads the data file to detect or the mentioned faults in one go, one after the other?

I will be very grateful for your help.

Thank you

Kind regards

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Kobra Ghasemi 2013-6-22

Hi you can use many approaches for detect these fault such as limit checking for case 2 or filtering sensor data in case of 4 or use pca (principle component analysis) for case3.

请先登录，再进行评论。

请先登录，再回答此问题。

Follow Question