Operation Strategy Optimization of Energy Storage Virtual Synchronous Machine Using Big Data Analysis Technology

Abstract

With the increasing combination of energy with the power grid, energy storage systems play a crucial role in maintaining grid stability and ensuring reliable power supply. In recent years, the concept of energy storage virtual synchronous machine (VSM) has emerged as an effective and flexible method for mimicking the behavior of synchronous generators in energy storage systems. This paper presents a new approach to optimize the operation strategy of energy storage VSM using big data analytics techniques. The proposed method utilizes data analysis to enhance the performance and efficiency of energy storage systems, contributing to overall grid reliability and resilience. Neural networks are employed as an efficient big data analytics technique, and in this study, they are applied to evaluate and optimize the operation strategy of energy storage VSM. Firstly, this paper evaluates the operation strategy of energy storage VSM using a deep belief network (DBN) as the underlying network. Secondly, an improved whale optimization algorithm (WOA) is proposed. IWOA initializes the initial population by employing a mapping strategy, laying the foundation for global search. Furthermore, IWOA introduces a nonlinear strategy to balance the global and local exploration capabilities, while avoiding premature convergence based on diversified mutation strategies. IWOA demonstrates significant improvements in search speed and convergence accuracy, with a strong ability to escape local optima. Thirdly, this paper utilizes IWOA to initialize the weights and thresholds of the DBN network, constructing IWOA-DBN. The network is then utilized to evaluate and optimize the operation strategy of energy storage VSM based on the evaluation results. The feasibility and correctness of IWOA-DBN are validated through systematic experiments.