To improve the accuracy of midterm power load forecasting, a forecasting model is proposed by combing kernel principal component analysis (KPCA) with back propagation neural network. First, the dimension of the input space is reduced by KPCA, then input the data set to the neural network model, optimized by particle swarm optimization. The monthly average of daily peak loads is forecasted to modify the daily forecast values and output the daily peak load in the end. Using the data provided by European Network on Intelligent Technologies to test the model, the mean absolute percent error of load forecasting model is only 1.39%. The feasibility and validity of the model have been proven.

The key technology of cascaded multilevel inverter with hybrid energy sources lies in the power distribution among different chains. A power distribution control strategy between the energy storage elements and the capacitors is proposed to achieve fault tolerant control. In cascaded multilevel inverter with hybrid energy sources, the chains with energy storage elements can operate in four quadrants while the chains with capacitors can only operate in two quadrants. With the proposed control method, the active power distribution is realized by active voltage vector superposition and the reactive power distribution is achieved by initial operation point selection. This enhances both system reliability and availability while enabling continuous operation in four quadrants. Meanwhile, the stable operation region and the range of the initial operation point have been derived by vector analysis to avoid over-modulation. An experimental prototype based on a single phase link with two series chains has been built in the laboratory. Simulation and experimental results are provided to demonstrate the effectiveness of the proposed control method.