Blockchain-Based Joint Auction Model for Distributed Energy in Industrial Park Microgrids

Abstract

To address the centralized trading demand within industrial parks and the scattered peer-to-peer trading demand outside industrial parks, this paper proposes a blockchain-based joint auction architecture for distributed energy in microgrids inside and outside industrial parks. By combining blockchain technology and auction theory, the architecture integrates the physical energy transactions within industrial parks with the distributed transactions in external microgrids to meet the centralized trading demand within industrial parks and the scattered peer-to-peer trading demand outside industrial parks, optimizing resource allocation and improving system resilience. In the microgrid auction mechanism for industrial parks, considering distributed energy providers (sellers) and distributed energy buyers, an auction mechanism with power transmission distance, average electricity price, and enterprise nature as its main attributes was constructed to maximize social welfare, realizing efficient energy flow in a multi-microgrid environment and enabling coordinated mutual benefits for producers and consumers within the region. Finally, a case study was conducted on the joint auction mechanism for microgrids inside and outside industrial parks, including the impacts of market dynamics and user preferences on electricity prices using different trading methods, the computational results using different trading matching methods (comparing single-attribute and multi-attribute methods), and multi-dimensional verification of user satisfaction with peer-to-peer transactions in a blockchain environment. The effectiveness of the joint trading between physical energy transactions within industrial parks and external microgrids was demonstrated, which could efficiently coordinate energy allocation inside and outside the parks and reduce the cost of energy configuration.

1. Introduction

In January 2022, the National Development and Reform Commission of China issued the “Guiding Opinions on Accelerating the Construction of a National Unified Electricity Market System”, proposing that by 2025, a power market system with collaborative operation of national and provincial (regional/municipal) markets in China and an integrated design of ancillary service markets should be initially established. By 2030, a new type of electricity market with equal competition among market entities should be basically realized. In January 2022, the National Energy Administration of China issued the “Notice on Publicly Soliciting Opinions on the <Basic Rules of the Power Spot Market (Draft for Comments)> and <Power Spot Market Supervision Methods (Draft for Comments)>”, proposing to promote the participation of emerging market entities such as energy storage, distributed power generation, load aggregators, virtual power plants, and new energy microgrids in transactions, and to establish a power spot market based on the principle of “unified market and collaborative operation”.

With the orderly development of China’s energy system reform [1] and the diversified development of microgrid energy supplies, microgrid systems have gradually become the main scenario for absorbing distributed energy due to their characteristics such as a concentrated user electricity load, large power supply, and complete power distribution equipment. Microgrids can coordinate the contradiction between the power grid and distributed energy [2], reduce the frequent impact of distributed energy on the power grid, and thus flexibly manage and dispatch distributed energy. As the most widely used application of microgrids, industrial park microgrids have enriched their own energy composition by relying on their geographical location and regional advantages and with the help of technologies such as photovoltaic power generation and solar power generation. Based on the characteristics of self-production and self-sales, industrial park microgrids have become a new entity that integrates energy production and consumption, playing an important role in fully tapping the dynamic energy potential of industrial users and effectively balancing various load requirements. However, how to manage and control the demand for centralized trading of physical energy within industrial park microgrids and the demand for peer-to-peer trading of external energy, as well as realize the coordination and mutual benefit between producers and consumers of distributed energy transactions within industrial park microgrids, and verify the effectiveness of energy transaction configurations and collaborations through multiple dimensions, are the key to achieving the reliable and stable operation of industrial park microgrids and meeting the joint regulation of distributed power resources inside and outside industrial park microgrids.

Currently, scholars have conducted in-depth research on transaction and coordination methods for distributed energy. Wang et al. proposed a distributed power transaction mechanism with the collaboration of distributed power source groups, power management platforms, and power trading platforms under the condition of complete marketization of the selling side and built a free trading system for distributed power sources to support the development of large-scale distributed power sources [3]. Aiming at the current low marketization degree of distributed power sources and imperfect public service systems, Xiong et al. designed a multilateral trading system with the participation of multiple parties, which can achieve the independence of power trading while maintaining power dispatch and help to improve the popularity of distributed energy trading [4]. Chen et al. constructed a mechanism for multi-dimensional trading and dispatch of community-integrated energy, which can improve the trading and dispatch balance of community-integrated energy and create an integrated energy community with mutual assistance and efficient operation [5]. Aiming at the problems of low efficiency and large resource loss in current distributed energy transactions, Li et al. proposed a distributed energy transaction mechanism with the proxy reputation proof consensus mechanism as its core, which can effectively motivate all parties to keep their promises spontaneously by introducing a distributed reputation control mechanism, ensuring the scalability of distributed energy transactions [6]. Based on the interconnected multi-energy system, Liu et al. proposed a day-ahead optimal operation strategy for the allocation and operation of distributed energy, providing a guarantee for the dispatch efficiency of distributed energy with intermittency and randomness [7]. Aiming at the energy coordination and dispatch problem between the park microgrid and the main distribution network, Ma et al. established an integrated dispatch model in the park-integrated energy system, distributed energy, and energy storage equipment [8]. Aiming at the internal structure and dispatchable resources of smart parks, Zhi et al. proposed a two-layer optimization dispatch model with the dispatch cost as the objective function, which compensates for the shortage of industrial parks and absorbs surplus power, providing support for the system to implement power balance [9].

Currently, most distributed energy transactions are carried out in a broad sense of distributed transactions, ensuring the feasibility of transactions by introducing reputation or using contracts, and making improvements in terms of transaction efficiency and cost. However, it is still necessary to consider how to match energy transactions between multiple entities inside and outside the industrial park microgrid, as well as between centralized transactions and distributed energy transactions, in actual distributed transaction scenarios such as industrial park microgrids. It is also important to consider how to achieve joint and efficient transactions between physical energy trading within the industrial park and external microgrids. The above-mentioned research provides a certain theoretical basis for the management and transaction of distributed energy but still faces the following problems.

There is a large difference in the frequency of distributed energy transactions between microgrids inside and outside the industrial park. Data interoperability and sharing between different platforms, as well as transaction rules and processes, are difficult to effectively integrate and regulate. There is a lack of a reasonable and effective mechanism for regulating centralized and distributed energy transactions.

The diversified sources of distributed energy lead to unclear energy transaction prices and inconsistent energy evaluation indicators within the industrial park microgrid, lacking a reasonable and effective mechanism for centralized energy transactions in the industrial park.

There is a lack of an auction and transaction matching mechanism that considers the dynamic changes in the energy trading market and user behavior preferences in the industrial park microgrid.

Blockchain technology [10], as a decentralized and trustworthy distributed ledger [11], can ensure the data privacy of power trading entities and trading information [12] by leveraging hash encryption algorithms and consensus mechanisms, providing a guarantee for distributed energy transactions in industrial park microgrids. Smart contracts are programs in a decentralized blockchain [13], special protocols that automatically execute when providing transaction verification and meeting contract-set conditions and can provide efficient processing functions for distributed energy scheduling and trading. Smart contracts are the guarantee of fairness and effectiveness in energy trading [14].

Therefore, this paper combines blockchain technology with energy trading in microgrids inside and outside industrial parks and proposes a blockchain-based joint energy trading architecture and auction trading mechanism for industrial park microgrids. First, a joint trading architecture for microgrids inside and outside industrial parks was established to solve the problems of data interoperability and sharing between different platforms, as well as the docking of trading rules and processes. Second, an auction mechanism for industrial park microgrids was designed, considering distributed energy providers (sellers) and distributed energy buyers, with the goal of maximizing social welfare, to coordinate the energy trading needs between producers and consumers in industrial park microgrids. Finally, a case study simulation was conducted on the proposed joint auction mechanism for microgrids inside and outside industrial parks, proving the effectiveness of the joint trading between energy trading entities within industrial parks and external microgrids, realizing the efficient collaborative allocation of energy inside and outside industrial parks and improving the dispatch flexibility of the distributed energy market in microgrids.

The remainder of this article is structured as follows. Section 2 introduces the related research on blockchain in microgrid power trading. Section 3 presents the blockchain-based grid transaction architecture that underpins our work. Section 4 develops a comprehensive model capturing the key elements of the microgrid electricity market within an industrial park. Section 5 builds upon this foundation, constructing an auction mechanism tailored to the distinct characteristics of industrial park microgrids. To illustrate the feasibility of our approach and validate the soundness of the proposed auction design, Section 6 walks through a concrete numerical example, providing an analysis of the results. Section 7 concludes this paper by recapping our main contributions and pointing out promising directions for further research in this area.

2. Related Work

Blockchain technology can build a secure, decentralized trading platform that enables energy producers and consumers to conduct transactions directly without relying on traditional intermediaries, thus reducing transaction costs and improving transaction efficiency. Many experts and scholars have also carried out in-depth research on this aspect. Mengelkamp et al. proposed seven frameworks for building an efficient microgrid energy market [15]. Li et al. proposed a decentralized framework based on blockchain and edge computing technology to meet the secure and distributed requirements of knowledge and service sharing in the manufacturing ecosystem [16]. Li et al. used blockchain technology to design a distributed peer-to-peer network architecture to promote the development of distributed peer-to-peer networks [17]. Tian et al. proposed a blockchain-based distributed energy trading model for microgrid clusters, which used particle swarm optimization for transaction matching based on the consideration of distribution network operation costs, such as network fees, solving the disadvantages of high operation costs in centralized power trading [2]. Barenji et al. proposed a consensus mechanism for power trading based on credit improvement RFAT, which introduced a credit scoring mechanism to avoid election disputes between transaction nodes, enhance the resistance to malicious power trading behaviors, and thus improve the efficiency and fairness of power trading [18]. Zhao et al. proposed a blockchain-based two-tier framework for energy transactions in multiple microgrids, providing a decentralized transaction, information transparency, and mutual trust mechanism [19]. The central node of the microgrid collects demand information and sends it to the upper market for transactions. The dual-price bidding mechanism is adopted to ensure free and fair transactions between nodes, reduce the transaction volume with the main grid, and improve the energy utilization rate. Yang et al. verified the effectiveness of blockchain in protecting distributed control systems from false data attacks [20].

Meanwhile, smart contracts, as a key component of blockchain, also play an important role in the smooth running of microgrid transactions. Smart contracts have the ability to automate the execution of contract terms without third-party intervention, ensuring the security and reliability of the transaction. Secondly, smart contracts can realize the transparency and non-tamperability of the transaction, and all participants can view the transaction details, thus reducing information asymmetry and potential fraud, thus improving the transaction efficiency, reducing the transaction cost and time, and promoting the liquidity and development of the electricity market. Some experts and scholars have also performed corresponding research on the application of smart contracts in power transactions. Wang et al. used smart contract technology to design an adaptive continuous double auction mechanism, realizing direct energy trading based on bidding strategies, used a multi-signature mechanism to protect users’ interests when transactions are revoked, and analyzed the advantages of this trading in terms of matching efficiency [21]. Pee et al. proposed a blockchain-based smart energy trading platform that utilizes smart contracts to address security issues in transactions without the need for third-party involvement [22]. Seven et al. discussed blockchain smart contracts for peer-to-peer energy trading in virtual power plants, and the proposed solution is based on a public blockchain network, where all processes of the auction are executed by smart contracts, solving the cost and security issues [23]. Liu et al. proposed a power trading mechanism based on blockchain and smart contracts for electric vehicles in V2G networks, which realizes the openness and transparency of power trading and reduces the cost of power purchasers while increasing the profit of power sellers [24]. Hu et al. introduced a blockchain-based smart contract trading mechanism for energy and power supply and demand networks, emphasizing the importance of smart contracts in realizing an efficient trading method in the electricity market [25]. Chen et al. explored a secure power trading and incentive contract model for electric vehicles based on energy blockchain technology, highlighting the advantages of smart contracts in terms of execution efficiency [26].

Finally, as a matching mechanism, the auction model provides an effective mechanism for electricity trading, achieving price discovery and optimization of resource allocation through bidding and matching demand and supply between buyers and sellers. The auction model can promote fair competition in the electricity market and ensure the transparency and efficiency of transactions. In addition, the auction model can reduce transaction costs and increase the satisfaction and trust of market participants. Similarly, the application of the auction model in electricity trading has been studied accordingly. Chen et al. proposed a predictive integration model for optimizing trading strategies in peer-to-peer markets based on continuous two-way auctions [27]. Manjunatha et al. proposed an auction-based unilateral bidding electricity market as an alternative to the bilateral contract energy trading model, focusing on the bidding behavior of bidders and achieving better profit sharing for all stakeholders [28]. Jiang et al. focused on game theory-based modeling and pricing of electricity transactions between producers and consumers in the energy blockchain environment. It was finally experimentally demonstrated that the mechanism can increase the seller’s profit and reduce the buyer’s utility loss [29]. Following the literature review of the above three aspects, the analysis and summary of research gaps are as follows: (1) difficulty in data interconnection between different trading platforms, unclear data specifications, and non-transparent trading processes; (2) the inconsistency of energy evaluation indexes in the trading process and lack of a reasonable and effective energy trading mechanism; and (3) a lack of consideration of dynamic changes in the energy trading market and user behavioral preferences in microgrids in industrial parks in the auction model currently used for aggregation mechanism. In summary, considering the demand for transaction transparency, price discovery, and resource allocation that exists in microgrid transactions, this paper combines blockchain and the auction model to propose a blockchain-based energy joint transaction architecture and auction transaction mechanism for industrial park microgrids, which ensures the safe operation of the microgrids and the energy transactions and, at the same time, improves the social welfare of the microgrids.

3. Platform Architecture

To address the problems of data interoperability and sharing between different platforms, as well as the docking of trading rules and processes during the trading process between energy entities within industrial parks and external microgrids, this paper innovatively proposes a joint trading architecture for microgrids inside and outside industrial parks using blockchain technology. This architecture can combine the physical energy trading within industrial parks with the distributed trading of external microgrids. Blockchain technology and auction theory solve the problems of data interoperability and sharing between different platforms, as well as the docking of trading rules and processes. At the same time, they satisfy and balance the centralized trading demands within industrial parks and the scattered peer-to-peer trading demands outside industrial parks, optimizing resource allocation and helping to improve system resilience. In this architecture, the application of smart contracts improves the execution process of distributed energy transactions, thereby helping both parties reach their trading goals without disclosing private trading data, further enhancing the security of the entire system. The structural model of the proposed system is shown in Figure 1, and the entire system consists of three levels: the blockchain network layer, the microgrid layer, and the industrial park layer:

The blockchain network layer includes distributed ledgers, consensus protocols, smart contracts, and encryption. The distributed ledger contains the current state and serves as a register for transactions. The consensus protocol allows multiple nodes to reach an agreement on the order, legitimacy, and structure of updates to the ledger state for a batch of transactions through endorsement, ordering, and peer-to-peer validation. Smart contracts validate distributed transaction programs executed on nodes and automatically execute a specific business logic [30]. Identity data pools, auction proof pools, privacy credential pools, etc., are used to construct the business logic of smart contracts. The business logic in distributed energy transactions is abstracted into code and written into smart contracts, which are automatically executed by the smart contracts, facilitating accountability and supervision, and protecting users’ rights and interests [31].

The microgrid layer is a platform that ensures the completion of distributed energy transactions in multiple parks within a specific scope. Transaction participants, such as distributed energy resource owners and consumers, can query their data upload process, submit their supply and demand requirements, and verify each other’s identification information and transaction specifications. In order to maintain fairness in transactions and maximize the interests of all parties, the microgrid layer adopts a multi-attribute auction mechanism, creating a balanced trading environment. This ensures that factors such as the power transmission distance, electricity price, and peak-to-valley electricity ratio are fully considered in transactions, thus ensuring that both parties can obtain satisfactory results and further promoting the healthy development of the distributed energy trading market.

The industrial park layer provides functional services to the platform, including transaction supervision services, middleware process services, blockchain interface services, etc., and offers different functions according to the different stages of the transaction in which the participants are involved. The main function of the supervision service is to monitor whether the transaction participants fulfill the transaction agreement and detect any emergencies. The middleware process service and blockchain interface service provide technical support for the smooth progress of transactions. (1) Before starting a transaction, transaction participants need to undergo identity verification to confirm their transaction qualifications. At this stage, the middleware process service is responsible for handling related affairs, such as identity confirmation, transaction condition confirmation, and information exchange. (2) During the transaction stage, the middleware process service handles transaction requests, such as submitting and modifying energy orders, generating matching lists, etc. The blockchain interface service retains transaction records while processing transaction requests, ensuring transaction transparency and data backup, enabling quick traceability and investigation when problems arise. (3) During the transaction process, the supervision service of the industrial park layer begins to play its role. It is responsible for supervising whether the participants perform accurate status and demand reporting in accordance with the transaction agreement, such as paying fees on time and providing electricity on schedule, and conducts transaction acceptance to confirm the transaction results. In addition, it regularly reviews and detects any emergencies, such as abnormal price fluctuations, supply–demand imbalances, unexpected power outages, etc., and promptly handles and prevents these problems.

4. Modeling of Industrial Park Microgrid Participating in Electricity Market

Within industrial parks, electricity resource holders and demanders widely exist as prosumers. Multiple prosumers can spontaneously form microgrids based on their own energy characteristics and demands, leveraging cluster advantages to provide diversified energy services for the region, jointly constituting an organization interaction architecture based on individuals and groups. Each prosumer, as an economic entity at the individual level, achieves autonomous operation using the energy management system at the prosumer level. The characteristic of this trading model is that the transactions do not go online, with fewer restrictions.

Outside industrial parks, microgrids integrate the scattered resources of many internal prosumers for overall configuration. However, at the same time, due to the centralization and monopoly of institutions, the security of transaction data cannot be guaranteed, and centralized scheduling is not as efficient as decentralized trading between users through rules. Moreover, due to the opacity of transactions, if rent-seeking, selling user information and other behaviors occur, user rights and interests are harmed, and their willingness to participate in market transactions is affected.

Therefore, by analyzing the above two trading characteristics, we establish an auction mechanism for industrial park microgrids to ensure the smooth progress of these two types of transactions. Among them, blockchain technology is used to ensure the data security of transactions outside industrial parks, and smart contracts are utilized to optimize transaction processes and improve efficiency and security. As for transactions within industrial parks, the trading model balances the profits of both parties to increase the enthusiasm for transaction participation.

4.1. Model Building

Combining the background of industrial park microgrid transactions and based on time-of-use electricity prices, this paper focuses on the flexible interaction between microgrid trading and main grid coordination in microgrid electricity consumption. It proposes an event-driven aggregation method based on on-chain and off-chain transactions in industrial parks. Under the premise of meeting the internal demands of various prosumers within the industrial park, each prosumer in the industrial park participates in P2P energy trading with multiple aggregators through contracts. Multiple aggregators are responsible for matching transactions between users to find the optimal capacity of different distributed energy resources, maximizing the balance between local demand and supply. The objective function is:

$$P_{\min }=C_1+C_2$$

(1)

where $C_1$ is the transaction cost for prosumers participating in the industrial park microgrid market, and $C_2$ is the resource consumption cost for prosumers participating in the industrial park microgrid market.

4.2. Constraints

When prosumers in an industrial park participate in electric energy trading, they must comply with the power balance constraint, which means that the sum of the grid’s power supply and photovoltaic power generation should be equal to the sum of the energy storage resource power and the load resource power:

$$P_g(t)+P_{pv}(t)=P_{es}(t)+L_{au}(t)$$

(2)

where $P_g(t)>0$ is the power supplied by the grid to the microgrid at time t; $P_g(t)<0$ is the power sold by the microgrid to the grid at time t; $P_{pv}(t)$ is the power generated by photovoltaics at time t; and $P_{es}(t)$ is the power of energy storage resources at time t.

5. Construction of the Industrial Park Microgrid Auction Model

5.1. Construction of the Multi-Attribute Auction Trading Model

The auction model is described as follows. In the above market modeling, the prosumers when $P_g(t)<0$ are defined as the set of sellers in the industrial park microgrid, $P_{seller}=\left\{\begin{array}{cccc}{P}_{s1},& P_{s2},& \cdots ,& P_{sn}\end{array}\right\}$, which means there are n sellers.

The prosumers when $P_g(t)>0$ are defined as the set of buyers in the industrial park microgrid $P_{buyer}=\left\{\begin{array}{cccc}{P}_{b1},& P_{b2},& \cdots ,& P_{bm}\end{array}\right\}$, where there are m buyers. At the same time, the electricity to be sold is defined as $E=\left\{\begin{array}{cccc}{E}_1,& E_2,& \cdots ,& E_k\end{array}\right\}$, which means there are K units of electricity to be sold.

To ensure the accuracy of the model, this paper sets the electricity trading scenario in the microgrid trading process as n prosumers actively cooperating, with m power response parties interested in their cooperation, and makes the following assumptions:

Independence: The private value $v$ of each prosumer is not affected by the estimates of others, that is, it is assumed that there is no collusion among prosumers.

Risk neutrality: All prosumers are neutral decision-makers, and each prosumer makes optimal decisions with the goal of maximizing their own cooperative profits.

Non-cooperativeness: All prosumers independently decide their own bidding strategies, and there are no cooperative agreements with constraints.

By constructing transaction evaluation indicators for prosumers, a transaction evaluation model for prosumers based on multi-attribute assessment is established. At the same time, the fuzzy comprehensive evaluation method is used to quantitatively score each indicator factor in the prosumer indicator matrix.

This paper selects the power transmission distance, average electricity price, peak–valley–flat electricity consumption ratio, and enterprise nature as key indicators for power transaction evaluation [32,33], providing a basis for power transaction evaluation. Among them, the power transmission distance indicator considers the loss of power in the transmission process, preventing high costs from being incurred during transmission. The enterprise node rating is the rating of prosumer enterprises within the industrial park, used to assess risks and carbon emissions. The average electricity price is the ratio of the total electricity expenditure to the total electricity consumption of prosumers during the statistical period (

Table 1. The main evaluation properties of the model.

	Indicators
Matching criteria Q	$\mathrm{Power} \mathrm{transmission} \mathrm{distance} q_1$
	$\mathrm{Average} \mathrm{electricity} \mathrm{price} q_2$
	$\mathrm{Enterprise} \mathrm{node} \mathrm{rating} q_3$
	$\mathrm{Customer} \mathrm{satisfaction} q_4$

).

The scoring function based on the above parameters is:

$$S_i\left(q_1,q_2,\cdots ,q_m\right)={\displaystyle \sum _{i=1}^m{l}_i{w}_i{q}_i^{s_i}}$$

(3)

where $w_i$ is the weight of each indicator $q_i$, and the value is public, $l_i$ is the normalization parameter used to unify the dimensions, and $s_i$ is the power of the electricity provider’s attributes, $0<s_i<1$ indicating the decreasing characteristic of the score.

The transaction process based on the above model construction is as follows:

Step 1: Integrate the transaction evaluation matrix, collect the evaluation data Q_i, and sort the evaluation data according to the established standards, as shown below:

$$Q_i=\left(q_1,q_2,\cdots ,q_m\right)$$

(4)

In the formula, $i=1,2,\cdots ,n$ represents the number of participating sellers, and $m$ is the number of indicators in the evaluation matrix.

Step 2: By determining the reference sequence [34], normalize the indicators with clear numbers such as the interval distance and fuzzy evaluation indicators in the above indicators to obtain a normalized prosumer evaluation set. The process is shown as follows:

$$Q_0(i)=\left(q_0(1),q_0(2),\cdots ,q_0(m)\right)$$

(5)

Step 3: According to the above normalization processing results [35], eliminate the difference in the order of magnitude between each indicator through the indicator processing parameters to prevent the impact on the transaction results. The specific calculation formula is as follows:

$$q_i(j)=\frac{h}{l}{q}_0(j)$$

(6)

In the formula, $h$ and $l$ represent the indicator processing parameters, and $q_0(j)$ represents the indicator to be processed.

Step 4: Determine the weight coefficient of each indicator. In this model, the weight coefficient is formulated by the electricity demander according to their own needs. The weight matrix is shown as follows:

$$W=\left(w_1,w_2,\cdots ,w_m\right)$$

(7)

Step 5: Combine the normalized prosumer evaluation set and the indicator weight coefficient to score each prosumer. All trading nodes with evaluation scores reaching the standard value $S$ can participate in the reverse auction and conduct reverse bidding. During the bidding process, the evaluation score can be used as a component to participate. The scoring function is shown as follows:

$$S(i)=q_i(j)w(j)$$

(8)

5.2. Blockchain Transaction Process

Traditional power transactions are often characterized by complex processes, dependence on centralized institutions, opaque information, low efficiency, and high costs. However, the introduction of blockchain technology can effectively address these challenges. The decentralization and immutability of blockchain ensure the transparency and openness of transaction information, reduce intermediary links, and improve the efficiency and security of transactions. The traditional electricity market involves numerous participants, including power plants, suppliers, and consumers. Coordinating and managing transactions among these participants is often time-consuming and complex. By leveraging blockchain, direct peer-to-peer transactions can be realized, simplifying the process and reducing operational costs. Furthermore, the smart contract component of blockchain enables the automatic execution of transactions based on preset conditions, minimizing human intervention and potential misuse while enhancing the overall operational efficiency of the market. Moreover, the real-time data logging and traceability provided by blockchain technology contribute to improving the transparency of energy distribution and usage. Figure 2 illustrates the overall process of electricity trading incorporating blockchain technology, with specific details outlined in the following steps.

Step 1: In this blockchain system, all transaction participants have their own identity verification and corresponding keys, which are stored in the microgrid seller set and microgrid buyer set, respectively. The information of each node is published on the chain, and only nodes that have passed the verification can initiate and receive requests. First, the microgrid sellers $\left\{P_{s1},P_{s2},\cdots ,P_{sn}\right\}$ and buyers $\left\{P_{b1},P_{b2},\cdots ,P_{bm}\right\}$ in the microgrid blockchain are judged through the market access rules on the smart contract to determine whether the producers and consumers meet the access conditions. Only market members who meet the conditions can enter the market to participate in transactions.

Step 2: When the quotation deadline arrives, each producer and consumer member submits relevant initial parameters (sales volume and minimum transaction price) for the transaction cycle based on their own forecasted electricity generation and consumption, and their own needs. Finally, the smart contract calculates the relevant scores of each member and sorts them.

Step 3: Through a series of smart devices at the edge of the industrial park, the latest demand forecast model is constructed using the day-ahead data. The clearing result is calculated in real time with power generators based on the forecast result, and finally, the clearing result is published to the market participants in real time.

Step 4: When the transaction is completed, the power generation, consumption, and transaction data of each producer and consumer are broadcast within the blockchain, and the legality of the transaction is automatically verified through scripts. If it is legal, the relevant information is recorded.

5.3. Blockchain Transaction Security and Privacy Analysis

• Auction Data Privacy

The data provider encrypts the auction data of all transaction participants using a symmetric encryption algorithm before uploading it to the blockchain network. This ensures that no malicious node, whether internal or external, can obtain the unencrypted metadata without the decryption key. Only authorized nodes can access the unencrypted auction data, preventing smart contracts and other participating entities from obtaining any information about the content of the auction data without proper authorization.

2.

Auction Process Privacy

During the power aggregation process, the blockchain employs a pre-set access mechanism to audit the auction participants, ensuring the legitimacy of all participants. Upon completion of the audit, the blockchain network generates a virtual identity for each participant, concealing their real identity and protecting their privacy. Furthermore, user privacy remains intact during interactions with the smart contract, which only has access to data related to the transaction process and not to other attributes associated with the user’s identity. This process guarantees the security and privacy of the auction process, providing reliable privacy protection for the participants.

3.

Traceability

In this scheme, once the two parties involved in the power transaction reach a consensus, the business process in which they carry out data sharing is recorded on the blockchain. If one party engages in illegal behavior that leads to a transaction problem, they will be held accountable.

4.

Resistance to Single Point of Failure

This scheme utilizes a distributed storage structure, and all access control within the blockchain network is carried out in a point-to-point manner among the nodes. This effectively mitigates the risk of a single point of failure and improves the efficiency of microgrid transactions.

6. Example Analysis

Since the electricity demand side has the decision-making power to choose the electricity provider, the purpose of designing this trading mechanism is to maximize the expected utility of the electricity demand side.

The expected utility calculation formula for the electricity demand side is:

$$U_i=\left\{\begin{array}{cc}{U}_b\left(q_1^{*},q_2^{*},\cdots ,q_m^{*}\right)& \underset{¯}{\theta }\le \theta \le \widehat{\theta }\\ 0& \widehat{\theta }\le \theta \le \overline{\theta }\end{array}\right\}$$

(9)

For a data demander with cost type $\theta$, the utility of winning the bid is:

$$U_i=U_b\left(q_1^{*},q_2^{*},\cdots ,q_m^{*}\right)P(\underset{¯}{\theta }\le \theta \le \widehat{\theta })+0\cdot P(\widehat{\theta }\le \theta \le \overline{\theta })$$

(10)

In the formula, $P(\underset{¯}{\theta }\le \theta \le \widehat{\theta })={(1-F(\theta ))}^{n-1}$.

Supposing $\theta$ follows an independent and identical distribution on $\left[\underset{¯}{\theta },\overline{\theta }\right]$, and substituting the previous equation into (11), we obtain:

$$U_i=U_b\left(q_1^{*},q_2^{*},\cdots ,q_m^{*}\right){(1-F(\theta ))}^{n-1}$$

(11)

Since $U_i$ is a random variable belonging to the ex-ante estimation with a corresponding probability density of $f\left(\theta \right)$, each data demander may become the type $\theta$ and win. There is a total of $n$ bidders. Therefore, the total expected utility of the data demander is:

$$U_t=n{\displaystyle {\int }_{\underset{¯}{\theta }}^{\overline{\theta }}{U}_b}\left(q_1^{*}(t),q_2^{*}(t),\cdots ,q_m^{*}(t)\right)\cdot {(1-F(\theta ))}^{n-1}\cdot f(t)dt$$

(12)

In the trading scenario, it is assumed that electricity suppliers submit relevant materials based on their actual costs and matching rules.

The following are the relevant trading attributes of each data owner:

Supplier 1:

$$(q_{11},q_{12},q_{13},q_{14})=(298,0.7,A,81)$$

(13)

Supplier 2:

$$(q_{21},q_{22},q_{23},q_{24})=(337,0.5,C,78)$$

(14)

Supplier 3:

$$(q_{31},q_{32},q_{33},q_{34})=(289,0.6,B,75)$$

(15)

Supplier 4:

$$(q_{41},q_{42},q_{43},q_{44})=(244,0.45,D,72)$$

(16)

Supplier 5:

$$(q_{51},q_{52},q_{53},q_{54})=(391,0.55,D,77)$$

(17)

Supplier 6:

$$(q_{61},q_{62},q_{63},q_{64})=(362,0.65,C,80)$$

(18)

In the case study, this paper verifies the supply-side participants in two dimensions: single-attribute transaction matching and multi-attribute transaction matching. It also verifies the effectiveness of the model by comparing the final pricing, user evaluation scores, and user satisfaction.

The following details the comparison of the final pricing at each microgrid node using the different matching methods (Figure 3).

In the process of continuously increasing the number of microgrid nodes in the scenario, compared with simply matching trading partners based on a single attribute, the multi-attribute auction can always maintain prices within a lower range. However, as the relative distance between nodes expands, it becomes difficult to simultaneously satisfy the requirements of distance and better prices, leading to an overall upward trend in the final transaction prices.

The following details the comparison of the multi-attribute evaluation values at each microgrid node using the different matching methods (Figure 4).

The matching method aiming at the lowest quotation has large fluctuations in the evaluation values due to the lack of correlation attribute evaluation. The shortest distance allocation method is affected by the expanding distance, and the influence of other factors also increases, resulting in consistently higher evaluation scores. In contrast, the attribute evaluation of the multi-attribute auction always stays close to the optimal solution.

The following details the comparison of user satisfaction at each microgrid node using the different matching methods (Figure 5).

By comprehensively considering the quotations from electricity providers and the multi-attribute evaluation values, we find that matching using the multi-attribute reverse auction method can maintain the highest user satisfaction in most cases. However, as the number of microgrid nodes continues to increase, this advantage gradually weakens, and the deviation also gradually expands. Despite this, the multi-attribute reverse auction method still shows a certain advantageous trend.

7. Conclusions

Under the current microgrid architecture of inside and outside industrial parks, there exist both centralized trading demands that focus on energy allocation and peer-to-peer trading demands that emphasize trading efficiency. To achieve rapid resource allocation and mutual benefits for prosumers within the park, the main contributions of this study include the following three aspects.

We innovatively proposed a joint trading architecture for microgrids inside and outside industrial parks. By integrating blockchain technology and auction theory, the architecture organically combines physical energy transactions within industrial parks with transactions in external microgrids, solving the problems of data interoperability and sharing between different platforms, as well as the docking of trading rules and processes. The architecture meets the centralized trading demands within industrial parks and the scattered peer-to-peer trading demands outside the parks, optimizing resource allocation and improving system resilience.

We innovatively proposed an auction mechanism for industrial park microgrids. Considering distributed energy providers (sellers) and distributed energy buyers, with the goal of maximizing social welfare, an auction mechanism was constructed with the power transmission distance, average electricity price, peak–valley–flat electricity consumption ratio, and enterprise nature as the main attributes. This mechanism realizes energy flow in a multi-microgrid environment, enabling coordinated mutual benefits for prosumers within the region.

We conducted a case study on the joint auction mechanism for microgrids inside and outside industrial parks. The analysis included the impact of market dynamics and user preferences on electricity prices, the computational results using different trading matching methods (comparing single-attribute and multi-attribute methods), and multi-dimensional verification of user satisfaction with peer-to-peer transactions in a blockchain environment. The results demonstrate the effectiveness of joint trading between physical energy transactions within industrial parks and external microgrids, efficiently coordinating energy allocation inside and outside the parks, and reducing energy allocation costs.