MetaPINNs:Predicting soliton and rogue wave of nonlinear PDEs via the improved physics-informed neural networks based on meta-learned optimization

Efficiently solving partial differential equations(PDEs)is a long-standing challenge in mathematics and physics research.In recent years,the rapid development of artificial intelligence technology has brought deep learning-based methods to the forefront of research on numerical methods for partial differential equations.Among them,physics-informed neural networks(PINNs)are a new class of deep learning methods that show great potential in solving PDEs and predicting complex physical phenomena.In the field of nonlinear science,solitary waves and rogue waves have been important research topics.In this paper,we propose an improved PINN that enhances the physical constraints of the neural network model by adding gradient information constraints.In addition,we employ meta-learning optimization to speed up the training process.We apply the improved PINNs to the numerical simulation and prediction of solitary and rogue waves.We evaluate the accuracy of the prediction results by error analysis.The experimental results show that the improved PINNs can make more accurate predictions in less time than that of the original PINNs.

Interest-Driven Model for Human Dynamics

Empirical observations indicate that the interevent time distribution of human actions exhibits heavy-tailed features. The queuing model based on task priorities is to some extent successful in explaining the origin of such heavy tails, however, it cannot explain all the temporal statistics of human behavior especially for the daily entertainments. We propose an interest-driven model, which can reproduce the power-law distribution of interevent time. The exponent can be analytically obtained and is in good accordance with the simulations. This model well explains the observed relationship between activities and power-law exponents, as reported recently for web-based behavior and the instant message communications.

Information Feedback Strategies in a Signal Controlled Network with Overlapped Routes

We investigate the effects of four different information feedback strategies on the dynamics of traffic, travelers＇ route choice and the resultant system performance in a signal controlled network with overlapped routes. Simulation results given by the cellular automaton model show that the system purpose-based mean velocity feedback strategy and the congestion coefficient feedback strategy have more advantages in improving network utilization efficiency and reducing travelers＇ travel times. The travel time feedback strategy and the individual purposed-based mean velocity feedback strategy behave slightly better to ensure user equity.

Heterogeneity of Some Cooperation/Competition Properties

We show that the heterogeneity index, which was proposed by Hu and Wang [Physica A 387 （2008） 3769], can be used to describe the disparity of the cooperation sharing or competition gain distributions, which is very important for understanding the dynamics of a cooperation/competition system. An analytical relation between the distribution parameters and the heterogeneity index is derived, which is in good agreement with the empirical results. Our theoretical and empirical analyses also show that the relation between the distribution parameters can be analytically derived from the so-called Zhang-Chang model [Physica A 360 （2006） 599; 383 （2007） 687）. This strongly recommends a possibility to create a general dynamic cooperation/competition model.

A Realistic Cellular Automaton Model for Synchronized Traffic Flow

A cellular automaton model is proposed to consider the anticipation effect in drivers＇ behavior. It is shown that the anticipation effect can be one of the origins of synchronized traffic flow. With anticipation effect, the congested traffic flow simulated by the model exhibits the features of synchronized flow. The spatiotemporal patterns induced by an on-ramp are also consistent with the three-phase traffic theory. Since the origin of synchronized flow is still controversial, our work can shed some light on the mechanism of synchronized flow.

Prisoner＇s Dilemma on Co-Evolving Networks without Strategy Update

We investigate the game theory in a structured population with the assumption that the evolution of network structure is far faster than that of strategy update. We find that the degree distribution for the finM network consists of two distinct parts： the low degree part which is contributed to by defectors and a broadband in the regime with high degree which is formed by cooperators. The structure of the final network and the final strategy pattern have also been numerically proved to be independent of the game parameters.