In this article, a new kind of neural network model named multi-scale convolutional echo state network (MCESN) is proposed for solar irradiance prediction, which integrates the strong feature extraction capability of convolutional neural network (CNN) and the fast yet efficient prediction ability of echo state network (ESN). Firstly, the feature information at different time scales of solar irradiance (one dimensional series) data are extracted and selected by multi-scale CNN (MCNN) in the pre-training stage. Then, the trained features extracted above are concatenated and passed to ESN module as the input signal, which can be further encoded into high-dimensional state space; Meanwhile, the target solar irradiance value is fitted and predicted by ESN in the prediction phase. Finally, the effectiveness of MCESN is evaluated by hourly solar irradiance prediction. In experiment, RMSE, MAE, MAPE and R are chosen as four metrics to evaluate the performance of the proposed model. Simulation results demonstrate that the proposed MCESN perform better than classical ESN, MCNN, backpropagation (BP) random forest (RF), long short time memory (LSTM) and deep ESN (DESN) algorithms.

This article proposed an optimal-consensus-based even t-triggered control strategy to coordinate multiple distributed generators (DGs) in an alternative current (ac) microgrid. The proposed control strategy integrated the average bus voltage recovery and the proportional reactive power sharing by turning the coordinated control problem into a convex optimization problem, which realized the bus voltage control and proportional reactive power sharing of the DGs in one controller so that the controller structure is significantly simplified. Accordingly, the controller provided an asynchronous event-triggered mechanism considering the finite availability of communication resources, which reduces communication burdens and avoids Zeno behaviors. The sufficient conditions are demonstrated using the Lyapunov function for maintaining the stability of the proposed control strategy. Several simulations are exhibited using the MATLAB/Simulink platform and the results are reviewed to validate the effectiveness of the proposed control strategy for enhancing the ac microgrid robustness and reducing communication burdens.

The standard condition for achieving exponential convergence of distributed resource allocation is the strongly convex objective functions, which is hard to be guaranteed in many practical cyber-physical systems. To study the resource allocation problem in a more general setting, we provide a new condition which only requires that the gradient-based map satisfies the metric subregularity. This condition is weaker than the standard strongly convex condition and is imposed to the objective functions. Based on such a relaxed condition, two new kinds of distributed allocation algorithms are proposed under continuous-time and event-triggered communications, respectively. The exponential convergence of our proposed algorithms are verified by rigorous theoretical analyses and some economic dispatch examples in the industrial cyber-physical system.