Continuous auction mechanism model for safety electric energy market transaction

Abstract

With the rapid development of microgrid in the electric power industry, the microgrid electric energy transaction has begun to be marketized, and the research on the microgrid electric energy trusted transaction has important theoretical research value and social value. The existing blockchain-based microgrid electric energy trusted transaction models mostly focus on energy management and scheduling control between microgrids when conducting electric energy transactions, and do not fully consider the bidding problems in the market-based transaction of microgrid electric energy, resulting in trading strategies are difficult to adapt to new market changes. In response to this problem, this paper proposes a reliable transaction approach for microgrid electric energy based on a continuous two-way auction mechanism. The proposed strategy accounts for the volatility of electricity prices in the microgrid trading market and employs the continuous two-way auction mechanism to evaluate the microgrid electricity trading tactics. In the microgrid electric energy transaction, the self-adaptive learning theory is applied to adjust the quotations of both parties, so that both parties can make reasonable quotations according to the market environment. By simulating experimental data, the findings indicate that the continuous two-way auction mechanism transaction strategy enables both parties to modify their quotations based on market transaction information, thereby displaying a high degree of flexibility in microgrid electricity’s market-oriented trading.

Keywords

Component electric power industry blockchain-based trusted transaction strategy continuous two-way auction

1. Introduction

As an important energy source for the development and progress of modern society, electric energy provides a basic guarantee for the high-quality life of human beings [1, 2]. In recent years, China’s economy has been developing rapidly, and the power industry has also continued to develop and progress, and people’s demand for electricity is increasing [3, 4]. At present, most of the power generation industry adopts the traditional centralized factory power generation technology, and the scale of power generation of the power grid is constantly expanding, which not only increases the cost of power generation and the difficulty of system operation, but also becomes more and more difficult to meet people’s demand for electricity safety [5, 6]. Therefore, the traditional centralized scheme is more and more difficult to adapt to the current user’s electricity demand [7, 8]. At the same time, as a power grid structure that can operate relatively independently, the microgrid, which can realize the combination of distributed renewable energy and user load, and can directly use local distributed renewable energy to reduce the power shock to the large power grid [9, 10].

The microgrid is a small power distribution and consumption system organically integrated by distributed power sources, loads, energy storage systems and converters. The microgrid can greatly promote access to distributed energy and renewable energy. Due to the close distance between distributed energy and load, it can greatly improve the reliability of the power supply and reduce operating costs. Therefore, microgrid systems are gaining more and more importance in the world’s attention worldwide. With the development and maturity of microgrid technology, some researchers propose a microgrid energy management system to manage the energy transaction between distributed energy and loads, but this method relies on the central management agency for management, and the transparency of transactions over-reliance on central regulatory agencies. When the central management agency is self-interested, there will be the risk of energy transaction data being tampered with. The traditional energy management model is difficult to adapt to the new microgrid energy transaction. The traditional method is to use the central management agency to manage the energy transaction entities [11, 12], which is different from the microgrid system and cannot manage the microgrid system well. Emergence may provide solutions to these problems [13, 14].

Blockchain is an emerging Internet technology [15, 16]. The data of the blockchain is stored in chronological order. Timestamps, asymmetric encryption, and distributed storage technologies are used to ensure that transaction data cannot be tampered with. Applied to the energy transaction of microgrid, it can effectively ensure the transparency and reliability of transaction data. Blockchain technology is a type of distributed ledger storage that facilitates decentralization and establishes a dependable database by being collectively maintained by multiple nodes. Blockchain platforms can be divided into three types of public and federated chains, and the difference between the two can be reflected in several aspects of user management, node access, decentralization, number of nodes, and storage methods. Using blockchain technology, local renewable energy transactions between microgrids can be better developed [17, 18].

In order to solve the bidding problem in the blockchain-based microgrid electric energy trusted transaction model, this paper studies the microgrid electric energy trusted transaction strategy based on the continuous two-way auction mechanism. The proposed strategy accounts for the volatility of electricity prices in the microgrid trading market and employs the continuous two-way auction mechanism to evaluate the microgrid electricity trading tactics. In the microgrid electric energy transaction, the self-adaptive learning theory is applied to adjust the quotations of both parties in a timely manner, so that both parties can make reasonable quotations according to the market environment. The effectiveness of the proposed trading strategy in this paper has been confirmed through the simulation of experimental data.

2. Related work

The main purpose of trusted storage and reliable communication of data at the physical level of the microgrid is for the trusted transaction of microgrid power. The requirements of electric energy trading will promote the market construction of the microgrid electric energy trading platform. In order to promote the market-oriented transaction of microgrid electric energy, Alam et al. proposed to build a demand-side management system for P2P energy transaction coordination, using load and electric energy scheduling to optimize the energy transaction strategy [19, 20]. In reference to [21], a hierarchical bidding and trading structure based on blockchain technology was introduced. Firstly, this structure incorporates multiple agents to refine the estimated cost probability distribution of other microgrids by applying Bayes’ theorem, bringing the probability closer to the actual value. To optimize the benefits of microgrids, [21] utilized the Nash equilibrium in the Cournot model to determine the most favorable offers and outputs of various bidding strategies for microgrids operating under distinct electricity demand conditions. But all transaction information in the system is obtained by the unified central entity management is easy to burden the system, and the unified central entity management system is prone to the single point of failure. Multiagent technology is beneficial to deal with the complex market environment of microgrids, and multiagent technology has been applied in many fields [22, 23]. The application of multiagent technology in the field of electricity market simulation is also increasingly widespread. Coelho et al. believe that multiagent technology can effectively promote the development of microgrid communication [24, 25]. In order to realize a complete P2P electricity market, Sorin et al. proposed a peer-to-peer electricity market with multi-bilateral economic dispatch, allowing multi-bilateral transactions based on product differentiation [26, 27]. In [28], a peer-to-peer (P2P) energy trading scheme was introduced for high penetration electricity markets with distributed energy resources (DER). This study presents a new algorithm using the primal gradient method to clear the market in a completely decentralized manner without the involvement of any central entity. Furthermore, to accommodate technical constraints in energy trading, bilateral trading’s line flow constraints are incorporated to prevent system overload or congestion. Although microgrids have not yet directly entered the power market, as the national power reform continues to progress, microgrids will participate in power market competition, enabling the more efficient and localized consumption of energy. Therefore, the marketization of microgrid electric energy trading deserves further research.

3. Trusted transaction strategy of microgrid electric energy based on continuous two-way auction mechanism

During the trusted transaction of microgrid electricity, both parties adapt their pricing based on the present market information. Therefore, before each round of transactions, both parties obtain the current optimal selling price and optimal buying price information based on the Nash equilibrium solution, and gradually adjust the quotation. Until the transaction is successful, the user’s quotation strategy is shown in Eq. (1):

$\displaystyle P_{c}(n)=\left\{{{\begin{array}[]{l}{P_{b}(n)+\mu_{c}\{{\min[{G_% {b},P_{s}(n)}]-P_{b}(n)}\},n=1}\\ {P_{b}(n)+\mu_{c}[{\alpha_{b}(t)-P_{b}(n)}],n>1}\\ {0<\mu_{n}\leqslant 1}\\ \end{array}}}\right.$ (1)

where $\mu_{c}$ represents the increased rate of user q55-ely increase their own quotations based on the best-selling price and best buying price information in the current market. In the subsequent transaction, if the transaction between the two parties in the market has not been completed, the user will appropriately increase his own quotation according to the target quotation and the current best selling price. The microgrid quotation strategy is shown in Eq. (2):

$\displaystyle P_{DG}(n)=\left\{{{\begin{array}[]{l}{P_{s}(n)+\mu_{DG}\{{P_{s}(% n)-\max[{G_{DG},P_{b}(n)}]}\},n=1}\\ {P_{S}(n)+\mu_{DG}[{P_{s}(n)-\alpha_{s}(t)}],n>1}\\ {0<\mu_{DG}\leqslant 1}\\ \end{array}}}\right.$ (2)

where $\mu_{DG}$ represents the increase rate of user quotations, and $G_{DG}$ represents the estimated price of electricity sold in the microgrid. $\alpha_{s}(t)$ represents the target electricity selling price of the microgrid.

Considering that the predictions of users and microgrids for their own consumption and purchase of electricity cannot be completely accurate, after the microgrid market transaction is closed, prediction errors may occur, resulting in excess or shortage of electricity. In this case, the users and the microgrid are settled separately with the large grid transaction due to the shortage or excess of electricity caused by forecast errors. For the user, when the user’s actual consumption of electricity is less than the transaction electricity, the excess electricity is sold to the grid. When the user actually consumes more electricity than the traded electricity, the user needs to purchase electricity from the grid to supplement their own electricity. Therefore, considering the user’s prediction error, the user’s actual expenditure is shown in Eq. (3):

$\displaystyle P_{\textit{ctrue}}=\left\{{{\begin{array}[]{l}{\overline{P_{b}}% \ast w,\Delta w<0}\\ {\overline{P_{b}}\ast w+P_{\textit{ctDG}}\ast\Delta w,\Delta w>0^{\prime}}\\ \end{array}}}\right.$ (3)

where $P_{b}$ represents the average price at which users purchase electricity in the microgrid; $w$ indicates the total electricity purchased by the user; $\Delta w$ refers to the load prediction error; $\Delta w<0$ indicates that the load user’s power consumption is lower than the anticipated value, and $\Delta w>0$ denotes that the actual power consumption of the load is greater than the predicted value; $P_{\textit{ctDG}}$ is the user’s purchase price from the grid. If the user’s prediction is entirely precise, the expected expenditure $E(P)$ of trading electric energy in the microgrid electric energy market is $P_{\textit{ctrue}}=\overline{P_{b}}\ast({w+\Delta w})$ .

For the microgrid, when the microgrid actually emits more electricity than it sells, the microgrid trades excess electricity with the grid. When the electricity actually emitted by the microgrid is less than the electricity sold, it is necessary to purchase the excess electricity from the grid to make up for the lack of its own output. The actual benefit of the microgrid is shown in Eq. (4):

$\displaystyle P_{\textit{DGtrue}}=\left\{{{\begin{array}[]{*{20}c}{\overline{P% _{DG}}\ast q+P_{\textit{DGbDG}}\ast\Delta q,\Delta q<0}\\ {\overline{P_{DG}}\ast w+P_{\textit{DGsDG}}\ast\Delta q,\Delta q>0}\\ \end{array}}}\right.,$ (4)

where $\overline{P_{DG}}$ is the average selling price of electricity in the market transaction; $q$ denotes the amount of electricity sold by the microgrid in the electricity trading market; $\Delta q$ is the prediction error of the distributed power supply, $\Delta q<0$ is the actual electricity generated by the microgrid is lower than the predicted value, and $\Delta q>0$ suggests that the actual power generated by the microgrid is higher than the predicted value; $P_{\textit{DGbDG}}$ represents the electricity purchase price of the microgrid to the grid, and $P_{\textit{DGsDG}}$ is the electricity sale price of the microgrid to the grid. When the prediction of the output of the microgrid is completely accurate, the anticipated profit from trading the same quantity of electricity in the market is $E({P_{DG}})=\overline{P_{DG}}{\ast}({q+\Delta q})$ .

In the microgrid electric energy transaction, after receiving the market quotation information, the adaptive learning theory is used to adjust the quotations of both parties in time, so that the two parties can make reasonable quotations according to the market environment. The two parties of the microgrid electric energy transaction adjust the aggressiveness $r$ and parameter $\theta$ according to the market information, and adjust them through short-term learning and long-term learning respectively. In the continuous two-way auction mechanism, the degree of aggressiveness $r$ is a parameter that reflects the intensity of participants’ bidding behavior. The value of $r$ falls within the range of [ $-$ 1, 1], where a value closer to 1 denotes a higher willingness to bid, a value closer to $-$ 1 indicates a lower inclination to bid, and when $r$ equals 0, it signifies a neutral stance towards the transaction. The value of $\theta$ is affected by market price fluctuations. When the market price fluctuates greatly, the agent reduces $\theta$ to quickly converge the target price to near the equilibrium price. The short-term learning is shown in Eq. (5):

$\displaystyle\left\{{{\begin{array}[]{l}{r({t+1})=r(t)+\beta_{1}[{\delta(t)-r(% t)}]}\\ {\delta(t)=({1+\lambda})r_{\textit{shout}}(t)}\\ {-0.05\leqslant\lambda\leqslant 0.05,0<\beta_{1}\leqslant 1}\\ \end{array}}}\right.$ (5)

$r(t)$ signifies the aggressiveness of the t-round transaction, $\beta_{1}$ denotes the short-term learning speed, $\delta(t)$ denotes the t-round transaction’s aggressiveness for the user’s quotation near the optimal buying price or the microgrid’s quotation near the optimal selling price. $r_{\textit{shout}}(t)$ represents the aggressiveness corresponding to the optimal quotation of the t-round transaction. When the agent has not completed the transaction in the t-round transaction, the optimal quotation of the t-round transaction is the transaction price. When the agent is a user and the transaction is completed in the t-round, the optimal quotation is the electricity purchase price or the electricity selling price in round t. $\lambda$ refers to the ideal progress factor. A value of $-$ 0.05 for $\lambda$ implies that the agent enhances its profit by decreasing its aggressiveness, whereas a value of 0.05 indicates that the agent improves the likelihood of transaction success by increasing its aggressiveness. $\beta_{1}$ means short-term learning speed. When the user’s target price is greater than the transaction price or the previous round of electricity purchase price, the user needs to reduce the aggressiveness, otherwise increase the aggressiveness. When the target price of the microgrid is less than the transaction price or the last round of electricity sales price, the microgrid agent needs to reduce the aggressiveness, otherwise increase the aggressiveness. Long-term learning is shown in Eq. (6):

$\displaystyle\left\{{{\begin{array}[]{l}{\theta({t+1})=\theta(t)+\beta_{2}[{% \theta^{\prime}(\alpha)-\theta(t)}]}\\ {\alpha=\frac{\sqrt{\frac{1}{n}\mathop{\sum}\nolimits_{i=m-n+1}^{n}({p_{i}-p^{% \prime}})^{2}}}{p^{\prime}}}\\ {0<\beta_{2}\leqslant 1}\\ \end{array}}}\right.$ (6)

where $\alpha$ is the normalized value of the standard deviation of the transaction prices of $n$ recent transactions, reflecting the market price fluctuations; $\beta_{2}$ is the long-term learning speed. $\theta^{\prime}(\alpha)$ is a function of $\alpha$ , and the optimal $\theta({t+1})$ can be found through historical market price fluctuations. The functional relationship is shown in Eq. (7):

$\displaystyle\theta^{\prime}(\alpha)=({\theta_{\max}-\theta_{\min}})\left({1-% \frac{\alpha-\alpha_{\min}}{\alpha_{\max}-\alpha_{\min}}}\right)e^{2\left({% \frac{\alpha-\alpha_{\min}}{\alpha_{\max}-\alpha_{\min}}-1}\right)}+\theta_{\min}$ (7)

where $\theta_{\min}$ and $\theta_{\max}$ are the limits of $\theta$ ; $\alpha_{\min}$ and $\alpha_{\max}$ are the minimum and maximum values of $\alpha$ respectively.

4. Experimental results and analysis

Table 1
Microgrid basic quotation information

Microgrid	Electricity sold /MW	Target selling price /(USD $\cdot$ MW-1)	Minimum electricity price /(USD $\cdot$ MW-1)
1	20	67	60
2	30	71	60
3	10	65	60
4	20	69	60
5	30	70	60
6	20	68	60

Assuming that there are 6 microgrids and 8 users in the microgrid electric energy trading market conducting electric energy transactions, the continuous two-way auction mechanism is utilized by both transaction parties to engage in bidding transactions. Table 1 displays the fundamental quotation details of the microgrid, including the electricity available for sale, the desired selling price, and the minimum selling price before the initial quotation. Table 2 exhibits the fundamental quotation information of users, which includes the amount of electricity that each user needs to purchase, as well as the target power purchase price and the maximum power purchase price before the first quotation.

Table 2

User base quotation information

User	Electricity sell /MW	Target selling price /(USD $\cdot$ MW-1)	Minimum electricity price /(USD $\cdot$ MW-1)
1	20	65	90
2	20	63	90
3	10	70	90
4	10	67	90
5	20	63	90
6	10	68	90
7	20	67	90
8	10	65	90

Figure 1.

Microgrid quotation per round.

Figure 2.

User quotation per round.

In the microgrid electric energy trading market, the two parties finally completed 8 transactions through 4 rounds of quotations. Each round of quotation of the microgrid is shown in Fig. 1, and each round of quotation of users is shown in Fig. 2. The two parties to the transaction in the order of the transactions in the market are Microgrid 3-User 3, Microgrid 1-User 4, Microgrid 1-User 6, Microgrid 4-User 7, Microgrid 6-User 1, Microgrid 2-User 8, Microgrid 2-User 5 and Microgrid 5-User 2. Microgrid 3-User 3 is the first to reach a transaction with each other. This is because when making an offer, Microgrid 3 sells electricity at the lowest price, while User 3 purchases electricity at the highest price and exceeds the selling price of Microgrid 3. Therefore, the transaction is concluded first. In the end, the transaction was concluded on Microgrid 5-User 2. This is because the two parties to the transaction have higher requirements for improving their own interests when adjusting the quotation through the adaptive learning theory. The adaptive learning parameters set to pay more attention to the transaction price, resulting in the previous three rounds of quotations failed to be traded due to price issues.

5. Conclusion

This paper studies the bidding problem in the trusted transaction model of microgrid electric energy based on blockchain. The existing blockchain-based microgrid electric energy trusted transaction models mostly focus on energy management and scheduling control between microgrids when conducting electric energy transactions, and do not fully consider the bidding problems in the market-based transaction of microgrid electric energy, resulting in trading strategies that are difficult to adapt to new market changes. Therefore, this paper proposes a trusted trading strategy for microgrid electric energy based on a continuous two-way auction mechanism. This strategy can take into account the fluctuation of electric energy prices in the microgrid electric energy trading market, and use the continuous two-way auction mechanism to analyze the microgrid electric energy trading strategy. The experimental results show that, compared with the existing trading strategies, the trading strategies are more suitable for the marketization of microgrid electricity trading, and serve as a reference for developing microgrid electricity trading platforms.

Footnotes

Acknowledgments

This work was supported by the Electric Power artificial Intelligence Engineering Center Project of the Shanghai Municipal Commission of Science and Technology (grant no. 19DZ2252800).

References

Liu

Yang

, et al. Lithium-ion battery charging management considering economic costs of electrical energy loss and battery degradation. Energy Conversion and Management. 2019; 195: 167-179.

İnci

Büyük

Demir

, et al. A review and research on fuel cell electric vehicles: Topologies, power electronic converters, energy management methods, technical challenges, marketing and future aspects. Renewable and Sustainable Energy Reviews. 2021; 137: 110648.

Chen

Wang

Chen

Zhao

. Networked microgrids for grid resilience, robustness, and efficiency: a review. IEEE Transactions on Smart Grid. 2021; 18-32.

Abu-Rayash

Dincer

. Analysis of the electricity demand trends amidst the COVID-19 coronavirus pandemic. Energy Research & Social Science. 2020; 68: 101682.

Gandoman

Jaguemont

Goutam

, et al. Concept of reliability and safety assessment of lithium-ion batteries in electric vehicles: Basics, progress, and challenges. Applied Energy. 2019; 25: 113343.

Duan

Tang

Dai

, et al. Building safe lithium-ion batteries for electric vehicles: A review. Electrochemical Energy Reviews. 2020; 3: 1-42.

San

Zhang

Guo

Hua

Xin

Blaabjerg

. Large-disturbance stability for power-converter-dominated microgrid: a review. Renewable and Sustainable Energy Reviews. 2020; 109859.

Bahrami

Chen

Wong

VWS

. Deep reinforcement learning for demand response in distribution networks. IEEE Transactions on Smart Grid. 2020; 12(2): 1496-1506.

Fan

Liu

, et al. Battery technologies for grid-level large-scale electrical energy storage. Transactions of Tianjin University. 2020; 26: 92-103.

10.

Yun

Cui

. Online preventive control method for static voltage stability of large power grids. IEEE Transactions on Power Systems. 2020; 35(6): 4689-4698.

11.

Ghaemi

Salehi

Hamzeh Aghdam

. Risk aversion energy management in the networked microgrids with presence of renewable generation using decentralised optimisation approach. IET Renewable Power Generation. 2019; 13(7): 1050-1061.

12.

Dong

Cheng

. Energy management optimization of microgrid cluster based on multi-agent-system and hierarchical Stackelberg game theory. IEEE Acces. 2020; 8: 206183-206197.

13.

Andoni

Robu

Flynn

, et al. Blockchain technology in the energy sector: a systematic review of challenges and opportunities. Renewable and Sustainable Energy Reviews. 2019; 43-174.

14.

Kousi

Tang

Metcalfe

, et al. Emergence and future of exsolved materials. Small. 2021; 17(21): 2006479.

15.

Zarrin

Wen Phang

Babu Saheer

, et al. Blockchain for decentralization of internet: Prospects, trends, and challenges. Cluster Computing. 2021; 24(4): 2841-2866.

16.

Alsharari

. Integrating blockchain technology with internet of things to efficiency. International Journal of Technology, Innovation and Management. 2021; 1(2): 01-13.

17.

Cheng

Niu

Zhang

. Prediction defaults for networked-guarantee loans. 24th International Conference on Pattern Recognition. 2018; pp. 361-366.

18.

Cheng

Chen

Wang

Xiang

. Efficient Top-k Vulnerable Nodes Detection in Uncertain Graphs. IEEE Transactions on Knowledge and Data Engineering. 2023; 35: 1460-1472.

19.

Alam

St-Hilaire

Kunz

. An optimal p2p energy trading model for smart homes in the smart grid. Energy Eciency. 2017; 1475-1493.

20.

Xiang

Wen

Cheng

Zhang

Qin

Qian

Lin

. General graph generators: Experiments, analyses, and improvements. The VLDB Journal. 2021; 31: 897-925.

21.

Guo

Min

, et al. Trusted transactions in micro-grid based on blockchain. Energies. 2019; 12(10): 1952.

22.

Niu

Chen

Cheng

Zhang

. Learning from web data with self-organizing memory module. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020; pp. 12846-12855.

23.

Xiang

Cheng

Shang

Zhang

Liang

. Temporal and Heterogeneous Graph Neural Network for Financial Time Series Prediction. Proceedings of the 31st ACM International Conference on Information & Knowledge Management. 2022.

24.

Coelho

Cohen

Coelho

Liu

Guimares

. Multi-agent systems applied for energy systems integration: state-of-the-art applications and trends in microgrids. Applied Energy. 2019; 820-832.

25.

Xiang

Cheng

Zhang

Wang

Zhang

. Efficient Learning-based Community-Preserving Graph Generation. 2022 IEEE 38th International Conference on Data Engineering (ICDE). 2022; pp. 1982-1994.

26.

Sorin

Bobo

Pinson

. Consensus-based approach to peer-to-peer electricity markets with product differentiation. IEEE Transactions on Power Systems. 2019; 994-1004.

27.

Zhu

Cheng

Luo

Yang

Luo

Qian

Zhou

. SI-News: Integrating social information for news recommendation with attention-based graph convolutional network. Neurocomputing. 2022; 494: 33-42.

28.

Khorasany

Mishra

Ledwich

. A decentralized bilateral energy trading system for peer-to-peer electricity markets. IEEE Transactions on industrial Electronics. 2019; 67(6): 4646-4657.

Continuous auction mechanism model for safety electric energy market transaction

Abstract

Keywords

1. Introduction

2. Related work

3. Trusted transaction strategy of microgrid electric energy based on continuous two-way auction mechanism

Table 1 Microgrid basic quotation information

Footnotes

Acknowledgments

References

Table 1
Microgrid basic quotation information