Determination of transfer capability using ANFIS with system condition separability

The transfer capability is the power quantity that can be transferred between considered points without any power system violations. With the repeated power flow (RPF), the calculation time is based on the scale of power system and its operating conditions. By nature, the transfer capability always changes continually as the power system operates. Thus, the power flow-base calculation is not suitable for updating the transfer capability in the real-time application. This research uses the adaptive neuro-fuzzy inference system (ANFIS) for the real-time calculation of the transfer capability. ANFIS is the intelligent method which can be trained by sample data and can calculate the output by the feed-forward algorithm. This paper proposes the input variables for ANFIS by the idea; ANFIS should be able to separate the operating conditions by these input variables. This concept also takes into account SVC and TCSC installed in the system. The proposed input variables are tested by the modified IEEE 30-bus test system. The calculation error between the checking data, from the RPF method, and the output of ANFIS is considered.