2
1. Problem Definition:
Hydraulic systems consist of motors, pumps, valves, actuators and hydraulic fluids. If not
maintained properly, repairs can be expensive, depending upon the application i.e. heat
degradation, air contamination, water contamination, fluid filtration and quality, etc. Condition
monitoring of hydraulic systems has gained increasing importance in industrial, energy and mobile
applications as a requirement of condition-based maintenance with several benefits such as
reduction of downtime and maintenance costs. Therefore, the problem that we will be exploring is
the condition monitoring of hydraulic systems by Tizian Schneider and Nikolai Helwig from
Saarland University in Saarbruecken, Germany.
2. Statement of Objectives:
The data set was experimentally obtained with a hydraulic test rig. This test rig consists of a
primary working and a secondary cooling-filtration circuit which are connected via the oil tank.
The data set contains raw processor sensor data which can be used to monitor the condition of the
hydraulic test rig.
The purpose of the project will therefore be to investigate and propose condition
monitoring of the hydraulic system by incorporating machine learning techniques acquired over
the semester with a focus in deep learning. Additionally, the measuring data from onboard sensors,
should be analyzed. The best use of the collected data would be to use it in precautionary
maintenance i.e. by analyzing the correlation of features extracted from raw sensor data and the
known fault characteristics of experimental obtained data, the most significant features specific to
a fault case can be identified.
3. Big data analytics methods and tool(s)
Tools used: Python and RapidMiner
Data preprocessing: The data was preprocessed using python by entreating big data techniques:
i. Impute missing values
ii. Standardization
iii. Test-Train Split
iv. Feature Engineering
v. Encoding
vi. Cross Validation
Processing methods: The machine learning techniques that will be explored as follows:
