Research On Online Health Evaluation Device Of Dry-type Transformer In Traction Power Supply System

At present,China's electrified railway technology is developing rapidly,and a large number of modern power equipments are used in traction power supply systems(TPSS).Dry-type transformers are widely used.Due to the application of modern power electronics technology,a large number of harmonics have appeared in the TPSS,and the dry-type transformers have been working under the condition of high harmonic content for a long time,and their health status has also been potentially affected.Existing drytype transformer health status assessment schemes are not yet perfect,there are problems such as too theoretical,too ideal application scenarios,etc.It is necessary to establish a complete and reliable dry-type transformer health evaluation system and an evaluation device with practical engineering.This article focuses on the related problems and limitations of dry-type transformer health assessment,starting from the failure mechanism,evaluation indicators,parameter characteristics,quantification schemes and engineering use,etc.From shallow to deep,the evaluation model and algorithm are established on the theoretical basis and experimental data,and implemented on a systematic evaluation device combining software and hardware.Based on the existing national standards and related research results,this paper analyzes the key factors that affect the health of the dry-type transformer,and plans an evaluation framework from the two aspects of the inherent parameters of the body and the external operating environment.Based on the relevant indicators,the mechanism of various types of faults of the dry-type transformer is studied to provide a criterion for the health assessment.This paper theoretically analyzes the change rules of the inherent parameters of dry-type transformers,and on this basis,a special test based on the frequency sweep method is designed for verification.A new transformer loss calculation method considering harmonic loss is proposed,which is more accurate than traditional methods.Considering the complexity of special tests,a method of online identification of inherent parameters using multiple linear regression is proposed to provide the possibility of health assessment.Based on the health assessment indicators,a set of dry transformer health assessment schemes are proposed.Quantitative dry-type transformer single index scoring rules are determined to provide a basis for fault prediction and fault identification.The method of calculating the remaining life of dry-type transformers based on the principle of thermal aging is studied,and the state maintenance plan is designed accordingly.Based on the system dynamics as the skeleton and big data as the basis,an iterative improvement scheme for the evaluation model is proposed to make the health evaluation system dynamically updated,and the evaluation results are more reliable with the increase of use time.Finally,a set of online monitoring terminals with information collection functions such as voltage,current,and temperature were independently developed.The terminal equipment follows the design principles of miniaturization,low cost,and easy installation,and has completed the circuit schematic diagram and PCB design.Combined with the relevant functional requirements,a set of host computer software with functions of communication,information display,data storage and health assessment was written.The device implements the proposed quantitative health assessment scheme,and displays various information and health assessment scores of the dry-type transformer in real time.The joint commissioning and joint test verified the engineering practicability and functional stability of the developed device.The research results show that the device can be reliable for the dry-type transformer health evaluation.It is of great significance in related fields such as intelligent monitoring,health analysis,and Internet of Things.