Research profile of ESR-B

Opis zdjęcia, a w nim ważne dla nas słowa kluczowe
Jose Gregorio Ferreira
Nationality: Spain
e-mail: jgferreira@pl.edu.pl

Research Topic:

1. Data capturing, conditioning and analysis using advanced signal processing methods for the determination of feature relationships for the purpose of diagnostics of electrical/mechanical interactions.


Research Partner:
ABB-NO - ABB AS (Norway)

Description of work
Task 1.1. Selection of electromechanical object and preparation for data capturing.
The main aim in this task is to select the electromechanical object from which the measured signals will be obtained. These signals will be analysed to determine relations between specific symptoms which appear in signals and object state including various failures. Then the number of signal and acquisition technique should be determined.

Task 1.2. Data conditioning and processing techniques.
After the measuring system is created, the quality of signals should be examined. According to advanced techniques chosen conditioning methods should be applied to improve the signal quality. For further analysis the measured data should be recorded for different state of the object.

Task 1.3. Review of data analysis methods for electromechanical interactions.
The aim is do a review of different methods and select the most appropriates to obtain the signals from the object. In selection process it is recommended to take into consideration used conditioning methods in acquisition process.

Task 1.4. Application of analysis methods to features determination.
In this task characteristic signal features have to be determined. These features should correctly characterise only a given state of the object independently from other states.

Task 1.5. Comparison between different methods used for diagnosis of exemplary system.
It is necessary to make a comparison between different methods used to analysed the measured data. The outcome is to determine the most effective method to select and assess the characteristic features of the object.

Task 1.6. Report of the project task realisation.
All researches have to be sum up by writing comprehensive report containing main results and achievements.

2. Diagnostic algorithm development for the assessment of machinery with industrial electric drives.

Description of work
Task 1.1. Review of intelligent methods used for electric drives diagnostic.
The aim is to do the comprehensive description containing the most important intelligent methods which are commonly known and can be used for electric drives diagnostic. The concept of using various intelligent method for object diagnosis should be proposed.

Task 1.2. Electric drive object investigation.
Learning about the functioning of a Direct Torque Control drive. Motor Parameter identification, Speed and Torque Control. Programming and controlling the drive.

Task 1.3. Data capturing and processing
Acquisition of drive-generated signals (e.g., estimated speed, torque) using OPC channels and built-in data loggers. Identification of differences in error signatures between direct-on-line and drive-powered motors (literature search and experimental verification). The signals which will be used for diagnostic process have to be selected and data for different state of the object should be registered. The data have to be preliminary processed to separate characteristic features which will be used to create effective diagnostic algorithms.

Task 1.4. Application of selected intelligent methods.
Recognise the possibility of using various types of neural network for diagnostic assessment have to be analysed. Then the most appropriate type, structure and parameters of neural network should be chosen. Such a network have to be tested using data measured in various state of the electric drive. Create the possible method and algorithm of diagnostic assessment using fuzzy logic method. The characteristic signals have to be selected and processed to create a belonging functions and inference rules describing a drive state. Exemplary models of diagnostic assessment system have to be implemented. Apply selected method of pattern recognition for assessment of the electric drive state. Pattern recognition method should be analysed and the most effective selected. To create a diagnosis algorithm the signal processing and characteristic feature separating techniques have to be chosen.

Task 1.5. Comparison of used methods for diagnostic.
It is necessary to make a comparison between different intelligent methods used to diagnose chosen electric drive. The outcome is to determine the most effective method to assess the electric drive condition.

Task 1.6. Report of the project task realisation.
All researches have to be sum up by writing comprehensive report containing main results and achievements.