Power grid construction and power measurement automation systems are gaining popularity and becoming ever more commonplace in developing countries. However, adoption rates are affected by inaccurate intelligent identification systems that control remote meter reading and line loss management. Local electricity distribution networks also have inconsistent user cable wiring, different geographical topologies, and issues with cable crosstalk, which lead to inaccurate readings of users’ data coming from the intelligent identification system, and sometimes even a failure to read electricity meters. Taking into consideration the voltage required for an intelligent identification system, we propose a new transformer for an intelligent identification system. Our new transformer improves the gray correlation analysis of watt-hour meters, which overcomes the shortcomings of existing identification methods by evaluating line voltages between unidentified and identified watt-hour meters. Experimental results show that our transformer with this method can accurately measure users’ electricity meter readings and perform line loss management.