A Smart Transformer Fleet Management System
Perception Fleet is a smart software solution designed to provide a comprehensive condition evaluation of a transformer fleet. Perception Fleet can analyze and interpret data in order to determine a transformers risk without the need for expert data analysis. Full technical analysis is also instantly available for transformer technical expert use.
This is accomplished by gathering, amalgamating, analyzing and interpreting the data held on transformers utilizing GE's online monitoring devices and offline data. The data is analyzed using intelligent algorithms for condition anomalies and characteristics to determine the transformers risk. Each transformer is then assigned a risk index, and they are ranked based on their criticality and risk of failure.
Utilising and evaluating data taken from GE's online multi gas, single gas and bushing monitoring equipment. Perception Fleet performs an automate risk analysis of transformers focusing on dissolved gas analysis (DGA) of insulation oil, tap changer insulation oil DGA, bushing capacitance, bushing power factor and partial discharge.
The transformer insulation oil properties and quality is also evaluated based on the results received from labs for manually sampled transformer oil. Perception Fleet's specific offline oil algorithm uses oil analysis techniques outlined by standards, working groups, committees and industry experts to evaluate the condition of the transformer where online monitoring is not available.
Due to the increased and varied level of information provided in lab results Perception Fleet is capable of evaluating a wealth of information beyond pure DGA, as outlined in the brochure.
The automated import facility enables not only the users but their labs to seamlessly update Perception Fleet with the latest data for a manual oil sample analysis. Perception Fleet then performs an automated analysis and evaluation of the newly imported data without the need for any operator interaction.
As well as analysing manually sampled oil data, the offline algorithm can also perform an analysis of oil information from 3rd party online monitors or software applications. The algorithm determines the data available in the imported CSV file received from the 3rd party and performs an evaluation.