A novel blended state estimated adaptive controller is designed for voltage and current control of microgrid against unknown noise. The state of a microgrid can deteriorate due to disturbances and packet losses. Therefore, it is necessary to achieve the true state of the system to enhance the control requirement and automation of the microgrid. To achieve the true state, this study proposes an unscented kalman filter algorithm. This algorithm is developed using an unscented-transformation and sigma-points measurement technique capable of minimizing the mean and covariance of a nonlinear cost function to estimate the true state of microgrid system. The effectiveness of the proposed algorithm is compared with an extended kalman filter. A model-reference modified adaptive PID controller is also developed and used in conjunction with the system to control its voltage and current. The performance of the developed controller is also evaluated against a certain number of load dynamics. Results indicate the controller provides more stable and high-tracking performance with the inclusion of the UKF in the system.