Optimal passive filter planning considering probabilistic parameters using cumulant and adaptive dynamic clone selection algorithm

Power quality in distribution systems may deteriorate due to increased time-varying and nonlinear loads. The harmonic current is one of the probabilistic and nonlinear pollution sources causing voltage distortion. The harmonic currents and bus loads in a power system are non-deterministic in nature. The single-tuned passive filter is considered an economical but efficient approach in reducing harmonics. This paper presents a new method for studying passive filter planning using cumulants and the Adaptive Dynamic Clone Selection Algorithm (ADCSA). The harmonic voltages, calculated by cumulants, correspond to those whose values of cumulative probabilities are 0.95. ADCSA was employed because of its ability to search the global optimum. The number (integer) of capacitor bank and tuned resonance frequency (real number) of the single-tuned passive filter are considered unknowns. The simulation results from an 18-bus system with multiple time-varying loads and harmonic currents show the applicability of the proposed method.

► The probabilistic nature of both harmonics and nonlinear loads is modeled using cumulants.
► The proposed method obtains the optimal filter size considering cumulative probability of 0.95.
► The number of capacitor and the resonance frequency of passive filter are obtained at the same time.