The emergence of coordinated and consistent macro behavior among self-interested individuals competing for limited resources represents a central inquiry in comprehending market mechanisms and collective behavior. Traditional economics tackles this challenge through a mathematical and theoretical lens, assuming individuals are entirely rational and markets tend to stabilize through the price mechanism. Our paper addresses this issue from an econophysics standpoint, employing reinforcement learning to construct a multi-agent system modeled on minority games. Our study has undertaken a comparative analysis from both collective and individual perspectives, affirming the pivotal roles of reward feedback and individual memory in addressing the aforementioned challenge. Reward feedback serves as the guiding force for the evolution of collective behavior, propelling it towards an overall increase in rewards. Individuals, drawing insights from their own rewards through accumulated learning, gain information about the collective state and adjust their behavior accordingly. Furthermore, we apply information theory to present a formalized equation for the evolution of collective behavior. Our research supplements existing conclusions regarding the mechanisms of a free market and, at a micro level, unveils the dynamic evolution of individual behavior in synchronization with the collective.

In recent years, in order to achieve the goal of “carbon peaking and carbon neutralization”, many countries have focused on the development of clean energy, and the prediction of photovoltaic power generation has become a hot research topic. However, many traditional methods only use meteorological factors such as temperature and irradiance as the features of photovoltaic power generation, and they rarely consider the multi-features fusion methods for power prediction. This paper first preprocesses abnormal data points and missing values in the data from 18 power stations in Northwest China, and then carries out correlation analysis to screen out 8 meteorological features as the most relevant to power generation. Next, the historical generating power and 8 meteorological features are fused in different ways to construct three types of experimental datasets. Finally, traditional time series prediction methods, such as Recurrent Neural Network (RNN), Convolution Neural Network (CNN) combined with eXtreme Gradient Boosting (XGBoost), are applied to study the impact of different feature fusion methods on power prediction. The results show that the prediction accuracy of Long Short-Term Memory (LSTM), stacked Long Short-Term Memory (stacked LSTM), Bi-directional LSTM (Bi-LSTM), Temporal Convolutional Network (TCN), and XGBoost algorithms can be greatly improved by the method of integrating historical generation power and meteorological features. Therefore, the feature fusion based photovoltaic power prediction method proposed in this paper is of great significance to the development of the photovoltaic power generation industry.