91 / 2024-09-17 15:12:47

Proposal of a Multi-energy internet based on CPS： Entropy assessment and improvement

Energy Internet,Information entropy,Thermodynamic entropy,Information physical system,Multi-energy complementation

The technological and industrial revolutions driven by digitization are accelerating the digital transformation of the energy sector, fostering the rapid development of the "Internet + Energy" paradigm. Jeremy Rifkin initially outlined the concept of an energy internet in his book The Third Industrial Revolution, forecasting the imminent emergence of a distributed network based on renewable energy sources. In response, the United States established the FREEDM Systems Center to develop next-generation power systems capable of plug-and-play integration for distributed devices, positioning these systems as prototypes for the energy internet.



However, most current studies concentrate on constructing the framework of the Energy Internet while overlooking that its core efficiency lies in slowing down the increase of entropy. Traditionally, methods to reduce entropy growth have relied on thermodynamic principles to enhance the quality and efficiency of energy systems. With advancements in information physics technology, the vast amounts of data generated by the operation of energy systems are now considered a form of information that can effectively reduce the entropy increase in new energy systems. Inspired by this, this paper proposes a distributed system for marine multi-energy complementarity and an information physical system, each developed from the perspectives of thermodynamic entropy and information entropy. The two are then integrated to build a more efficient Energy Internet.



The wind-light-storage multi-energy complementary form from the perspective of thermodynamic entropy means that due to the intermittent and unstable characteristics of photovoltaic power generation and wind power generation affected by natural factors, a hydrogen energy storage device is introduced, and the excess electricity in the system's peak power generation can be used for electrolytic hydrogen production, and hydrogen can be converted into water power generation through proton exchange membrane fuel cells at the power generation trough. The essence of optimizing the multi-energy complementary system is to reduce the entropy increase of the system.



From the perspective of thermodynamic entropy, a wind-solar-storage multi-energy complementary system addresses the intermittent and unstable nature of photovoltaic and wind power generation, which are influenced by natural factors. This approach involves integrating a hydrogen energy storage device that uses excess electricity generated during peak periods for electrolytic hydrogen production. The stored hydrogen can then be converted back into electricity during low-generation periods through proton exchange membrane fuel cells. The core objective of optimizing such a multi-energy complementary system is to minimize the increase in entropy within the system.



Cyber-Physical Systems (CPS) integrate computing, networking, and physical environments into complex, multidimensional systems. Incorporating CPS into integrated energy systems facilitates the creation of highly automated embedded systems, such as solar and wind tracking systems. Measurement devices link energy and control systems, employing feedback control through information. This enhances process reversibility and orderliness, thereby reducing entropy production.



To achieve global optimization, this paper applies the principle of synergetic theory, utilizing the correlation balance method to establish a synergistic relationship among the factors involved in the energy system's processes. Through the combined actions of the three subsystems—material flow, energy flow, and information flow—the system generates minimal entropy (thermodynamic entropy) within an ordered structure. The synergistic structure of the system adopts a step-control approach.



The energy internet merges integrated energy systems with the internet, using advanced sensors to create an "Internet of Things" foundation. This facilitates the coordinated management of energy production, transmission, storage, and sharing, and allows for data analysis and machine learning. Shifting from a centralized to a distributed model, it transforms the global power grid into an energy-sharing network, enabling the development of a CPS-based energy internet. Additionally, traditional energy internet models often use fixed load and climate parameters, leading to potential deviations during operation. This paper addresses this by incorporating the quantification of informational entropy related to uncertainty into the optimization of energy internet design.



Subsequently, this paper develops an entropy model for distributed energy systems. In this model, the system prioritizes using generated power to meet load demands, with any excess electricity being first directed to battery storage, and the remaining surplus fed into the grid. It is assumed that energy losses in the batteries occur solely during the charging phase. The study quantifies entropy production associated with solar panels, wind turbines, cooling machines during the cooling season, and the processes of battery charging and discharging. Additionally, the model accounts for negentropy induced by information, leading to the calculation of the system's total entropy increase, DS. Furthermore, the paper employs the on-site energy fraction (OEF) as a performance indicator to assess the energy balance of the distributed energy system, specifically evaluating how well the distributed energy sources match the load requirements.



Finally, system configuration optimization is conducted to minimize the total entropy increase. To address the multi-objective optimization problem of minimizing entropy increase (minDS) and maximizing the on-site energy fraction (OEF), this paper employs a multi-objective decision-making approach based on the Linear Programming Method with Multidimensional Preference (LINMAP). This method is used to solve the optimization problem of distributed energy system capacity configuration.