Critical Bus Voltage Mapping using ANFIS with regards to Max Reactive Power in PV buses

The aim of this paper is to apply Adaptive Network based Fuzzy Inference System (ANFIS) on a critical bus, to map the relation between voltage and power load with respect to the maximum reactive power injections into the system. Throughout different reactive power injection scenarios we have different voltage profiles for each bus of the system, therefore it is possible to use ANFIS on data collected for voltage and power load predictions. The ANFIS algorithm is applied to a limited amount of data in order to showcase the benefits of the neuro-fuzzy model while predicting bus voltage with good accuracy. It is important to mention that the voltage prediction is done within a restricted range of power loads, even though voltage reaches its critical point at the maximum load. The academic IEEE 14-bus system is employed with all its limits considered, so the results may be reproduced.