Electric power transformer diagnostics using neural-based observer

An essential step toward the development of an intelligent substation is to provide self-diagnosing capability at the equipment level. Transformers, circuit breakers and other substation equipment should be enabled to detect their potential failures and make life expectancy prediction without human interference. This paper focuses on the development of an online equipment diagnostics using artificial intelligence and a nonlinear observer to prevent catastrophic failures in substation equipment, thus providing preventive maintenance. Key elements of the system are a nonlinear observer, system identifier, and fault detector that use a uniquely designed neuro-fuzzy inference engine. Experimental results from application of this system to a distribution transformer are presented.