Carbon quota distributed information diffusion auction device and method

Abstract

The invention relates to a carbon quota distributed information diffusion auction device and method, the device comprises a seller and buyer corresponding agent module and a transaction platform module, the seller agent module submits information uploaded by a carbon quota provider to the transaction platform module, obtains an auction result from the transaction platform module and then sends the auction result to the carbon quota provider; the buyer agent module submits the information uploaded by the carbon quota demander to the transaction platform module, obtains auction time and auction results from the transaction platform module, and sends the auction time and auction results to the carbon quota demander; and the transaction platform module performs integration processing on the received information, constructs an undirected graph, generates a DFS spanning tree, adopts DFDM on the DFS spanning tree to perform auction, and manages all intermediary nodes in the DFDM so as to determine auction time and auction result information. Compared with the prior art, the method can solve the problem of insufficient circulation of the carbon transaction market, improves the income of the whole carbon transaction, and plays a role in protecting personal privacy and managing and checking data.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon trading data processing, and in particular to a carbon quota distributed information diffusion auction device and method. Background Art

[0002] Carbon trading is a general term for greenhouse gas emission rights trading. The basic principle of carbon trading is that one party to the contract obtains greenhouse gas emission reduction quotas by paying the other party, and the buyer can use the purchased emission reduction quotas to mitigate the greenhouse effect and achieve its emission reduction goals. The carbon trading market regards carbon emission rights as commodities (carbon quotas), allowing companies and individuals to freely purchase (sell) required (excess) emission rights, thereby finding the lowest cost method for emission reduction. The carbon trading market encourages companies and individuals to take energy-saving and emission reduction measures through economic incentives and other measures, and promotes the transition to a low-carbon industrial model. This approach can not only reduce the risk of global climate change, but also create opportunities for sustainable low-carbon economic growth.

[0003] Although the current carbon market is large in scale, the liquidity of the carbon market is obviously insufficient. The main reason is that in the carbon market, the main participants are major emission enterprises, while small and medium-sized enterprises, individual investors and financial institutions obviously lack the motivation to participate in emission reduction activities and the professional knowledge and capabilities required for carbon trading. This lack of motivation and technical knowledge has led to their limited participation in the carbon market and reduced the liquidity of the carbon market. Secondly, the carbon emissions monitoring, reporting and verification (MRV) mechanism needs to be improved. The effective and standardized operation of the carbon market depends on real carbon emissions data, but the current carbon market monitoring technology is not yet mature, which will affect the quality of carbon emissions data and further reduce the enthusiasm of enterprises and individuals.

[0004] In the prior art, Chinese patent CN115249092A (a method for seeking the optimal decision-making behavior of multiple subjects in the power market under the carbon trading market) discloses a method for finding the optimal decision-making behavior of multiple subjects in the power market under the carbon trading market using multi-objective linear programming. This method adopts the idea of ​​linear programming to determine the profit function and problem model of each party, and obtains the critical carbon price of each party after solving it using the target programming method, which provides a reference for the actions of the subject; however, this scheme does not consider the liquidity of carbon market information, that is, whether the enterprise will adjust its emission reduction strategy because it does not know the unit price of carbon quota, and the reference of the critical carbon price is not conducive to the seller to further increase its income, and is not conducive to the improvement of the social welfare of the transaction.

[0005] Chinese patent CN116720917A (A Distributed Carbon Trading Market Consensus Method) discloses a distributed method for supervising and managing the carbon trading market using a consensus algorithm and consensus nodes. This method divides the main participants in the carbon market into ordinary nodes (buyers and sellers) and consensus nodes (supervising parties). In a carbon transaction, more than half of the consensus nodes must agree before the trading center can make the final decision. Therefore, it can effectively ensure the fairness of carbon trading and the security and authenticity of data, and protect the privacy of buyers and sellers. However, this method does not take into account the flow of information, and it cannot attract more companies and individuals to participate in transactions; at the same time, the joint decision of consensus nodes will inevitably increase the transaction fees of third-party institutions, thereby reducing the enthusiasm of buyers and sellers to participate in transactions.

[0006] Patent CN116777623A (A carbon trading platform based on blockchain smart contracts) discloses a smart contract carbon trading platform based on blockchain encryption characteristics and with regulatory functions. The platform successfully implements a decentralized carbon trading mechanism through the distributed encryption characteristics of blockchain, improves the privacy and security of data before, during and after transactions, and is conducive to the unified supervision of my country's carbon market. However, with the introduction of blockchain, transaction costs or fees will inevitably increase, which is not conducive to acceptance by enterprises and individuals; at the same time, the invention also does not take into account the liquidity of the market and information.

[0007] In summary, the existing technologies for carbon trading and carbon markets have the following shortcomings:

[0008] 1) Existing technologies have little research on the liquidity problem of the domestic carbon trading market, that is, how to get more small and medium-sized enterprises and individuals to participate in carbon trading;

[0009] 2) At present, part of the research on the domestic carbon trading market remains on the formulation of the optimal emission reduction strategy for each enterprise and the impact of carbon prices on enterprises and individuals, without considering the improvement of the interests of carbon quota providers;

[0010] 3) Another part of the research focuses on issues such as information asymmetry and carbon emission information management in carbon trading, emphasizing the design of a fair and just trading mechanism and ignoring the issue of privacy leakage between the two parties to the transaction. Summary of the invention

[0011] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a carbon quota distributed information diffusion auction device and method, which can solve the problem of insufficient liquidity in the carbon trading market, increase the income of the entire carbon trading, and at the same time protect personal privacy and manage and verify data.

[0012] The purpose of the present invention can be achieved by the following technical solutions: a carbon quota distributed information diffusion auction device, including a seller agent module, a buyer agent module and a trading platform module, the seller agent module and the buyer agent module are respectively connected to the trading platform module for mutual communication, the seller agent module is connected to the corresponding carbon quota provider for communication, the buyer agent module is connected to the corresponding carbon quota demander for communication, the seller agent module is used to collect the information uploaded by the carbon quota provider and then distinguish and integrate it, submit the integrated information to the trading platform module, and obtain the auction result information from the trading platform module and then send it to the carbon quota provider;

[0013] The buyer agent module is used to submit the information uploaded by the carbon quota demander to the trading platform module, and obtain the auction time and auction result information from the trading platform module and then send it to the carbon quota demander;

[0014] The trading platform module is used to integrate and process the information submitted by the seller agent module and the buyer agent module by constructing an undirected graph, generating a DFS (Depth-First-Search) tree, and using DFDM (Distributed Fair Diffusion Mechanism) on the DFS tree for auction. At the same time, it manages all intermediary nodes in DFDM to determine the auction time and auction result information.

[0015] A carbon quota distributed information diffusion auction method comprises the following steps:

[0016] S1. The carbon quota provider provides auction information and uploads it to the trading platform module for publication through the seller agent module;

[0017] S2. During the auction duration, the carbon quota demander uploads the demand information to the trading platform module through the buyer agent module to join the auction;

[0018] S3. After the auction duration ends, the trading platform module integrates the auction information and demand information;

[0019] S4. The trading platform module applies a distributed fair incentive information diffusion mechanism to execute the auction and determine the allocation and payment information of the auction;

[0020] S5. Based on the auction allocation and payment information, combined with the feedback information from the seller agent module and the buyer agent module, the transaction platform module integrates the auction results and publishes them.

[0021] Furthermore, the specific process of step S1 is as follows: the carbon quota provider enters the auction through the seller agent module, and uploads the carbon quota quantity for auction, the actual unit to which the user belongs, and the expected transaction time;

[0022] After the trading platform module confirms that the uploaded information is correct, it will release auction information including the carbon quota quantity, provider unit, and duration.

[0023] Furthermore, the specific process of step S2 is as follows: during the auction duration, the carbon quota demander enters the auction through the buyer agent module and uploads its own bid, the referral object, and the real information of the unit to which the demander belongs.

[0024] Furthermore, in step S3, the trading platform module integrates the auction information and the demand information to obtain an undirected graph of the carbon quota auction, and implements a depth traversal algorithm on the undirected graph to generate a DFS spanning tree, thereby finding the strong propagation sequence and weak propagation set of the buyer user.

[0025] Furthermore, the undirected graph of carbon quotas includes:

[0026] Seller agent node S and its first-level transaction object set r s ;

[0027] Buyer agent node set N = {1, 2, 3, ..., n}, the private type of each buyer node i is θ i = {v i , r i},v i and r i They represent the buyer node i’s psychological valuation of the auction and the neighbor set respectively;

[0028] The set of intermediary nodes C corresponding to the seller and each buyer P = {P S , P 1 , ..., P i}, where P S represents the intermediary node corresponding to the seller, P i is the intermediary node corresponding to buyer i.

[0029] Furthermore, the specific process of generating the DFS spanning tree and finding the strong propagation sequence and weak propagation set of the buyer user in step S3 is as follows:

[0030] S31, running a depth-first traversal algorithm according to the seller agent S and the buyer agent set N, and generating a DFS spanning tree, wherein the starting node of the DFS spanning tree is the seller agent and the remaining nodes are all buyer agents; the edges of the tree represent the neighbor relationship of each node;

[0031] S32, the buyer agent node calculates its own status and reports it to the corresponding buyer intermediary node. The buyer agent node status is The corresponding values ​​are the parent node set of buyer i, the time when buyer i was first invited, the time when the last buyer of subtree i was invited, the shortest time to reach buyer node i, the buyer node with the highest psychological valuation in subtree i, and the psychological valuation of the buyer node with the highest psychological valuation in subtree i;

[0032] S33, seller intermediary node P s Publish the node ω with the highest psychological valuation and its status;

[0033] S34, seller intermediary node P s According to the information reported by the buyer, the strong propagation sequence C is compared and screened in turn. ω The buyer node on the , where if and only if the buyer proxy node c j When not participating in the auction, ω cannot join the auction. At this time, c j is called a strong propagation buyer, and defines the strong propagation sequence of ω as C ω , c j ∈C ω ;

[0034] If buyer i is a strong communication buyer, then the corresponding intermediary node P i dt i Report to the seller's intermediary node P s ; If buyer i is not a strong propagation buyer, he will be penalized for false positives;

[0035] In this process, each intermediate node P i The information will be verified by comparing the status of all reporting buyer nodes, thereby playing a role in information supervision;

[0036] S35, seller intermediary node P s According to dt i Arrange the key nodes in ascending order to obtain the strong relay sequence C ω And all strong propagation buyer nodes, count two strong propagation nodes c j With c j+1 All buyers on the communication path are regarded as weak communication buyers of the two strong communication buyers, and together constitute the weak communication set

[0037] Furthermore, the specific process of determining the auction allocation information in step S4 is as follows:

[0038] S41, according to the strong propagation sequence C ω , start trading from the first node after the seller, i.e. the node closest to the seller;

[0039] S42, buyer intermediary node P i Calculate the externality of the current strong propagation buyer node;

[0040] S43, compare the psychological valuation of the next buyer i+1 with the psychological valuation of the current transaction buyer i. If the next buyer i+1 needs to pay the amount of money except for itself and the weak propagation set When the externality of the buyer is higher than the psychological valuation of the previous buyer, the buyer will replace the previous buyer to join the auction;

[0041] If the next buyer needs to pay for itself and the weak propagation set When the externality is lower than the psychological valuation of the previous buyer, the current transaction buyer directly wins the auction.

[0042] Furthermore, the payment information of the auction determined in step S4 includes the initial payment and the re-payment, wherein the payment objects of the initial payment are the seller and the strong propagation buyer, and the payment rules of the relevant buyers are divided into three situations:

[0043] (1) For the seller (S), the externality amount paid by the strong communication buyer (strong communication buyer No. 1) with whom the seller has the first transaction in step S41 is all the payments that the seller can obtain in the initial payment stage, which can be expressed as:

[0044]

[0045] (2) For the buyer who finally obtains the item h The amount paid is the externality The formula is:

[0046]

[0047] (3) For non-winner strong communication buyers C j ∈C ω \c h , where '\c h 'Indicates' except c h ', will get a certain amount of communication benefits, the amount = the externality paid by the next strong communication seller except for itself and the weak communication nodes between the two - Self-paying externalities The formula is:

[0048]

[0049] Therefore, the formula for the buyer node's initial payment is generally expressed as follows:

[0050]

[0051] In the formula, (* represents any set) represents the highest psychological valuation among all nodes except the * set, that is, the externality except the * set; c j represents the jth strong communication buyer, c h Indicates the buyer who finally obtains the item; C ω is a strong propagation sequence, Between c j and c j+1 A weak diffusion set between two strong diffusion buyers;

[0052] The repayment targets are weak propagation buyers, strong propagation buyers and sellers. The repayment rules for relevant buyers are divided into three situations:

[0053] (1) For strong communication buyers C including winners j ∈C ω , the amount to be paid (p′ i )for Its value is equal to the externality of the highest psychological valuation buyer in the weak communication set divided by itself and -Externalities of all buyers except self and weakly propagated set and divided by the number of weak propagation buyers The sum of the number of sellers 1 is expressed as:

[0054]

[0055] (2) For weakly propagated buyers The amount to be paid (p′ i ) is -R i , whose value is equal to the externality between itself and the jth strong communication buyer -Externalities of all buyers except self and weakly propagated set and divided by the number of weak propagation buyers The sum of the number of sellers 1 is expressed as:

[0056]

[0057] (3) For seller S, after all buyers who meet the payment requirements have completed their payments, if the total payment amount is lower than the difference between the payment externalities of the two strong propagation sellers j and j+1, the remaining amount shall be paid by the seller in full;

[0058] The formula for the buyer node to re-pay is as follows:

[0059]

[0060] The propagation incentive is expressed as:

[0061]

[0062] In the formula, Same as during the initial payment phase; Indicates between c j-1 and c j The buyer with the highest psychological valuation in the weak communication set between two strong communication buyers.

[0063] Furthermore, the specific process of step S5 is: filter out the seller's income, the final winner, the buyer's information dissemination subsidy incentive and the amount required to be paid by the winner from the allocation information and payment information, and then publish them to the seller's agent module and the buyer's agent module, and receive the "confirmation transaction" information from both parties. After receiving the "transaction reached" signal from the two agent modules, the auction result information including the two parties to the transaction and the transaction amount will be announced.

[0064] Compared with the prior art, the present invention has the following advantages:

[0065] Based on the current unstable carbon trading market environment and insufficient information circulation, the present invention takes into account the powerful function of social network in spreading information and introduces a fair incentive information diffusion auction mechanism based on social network, so that more small and medium-sized enterprises and individuals can participate in transactions, which can effectively improve the enthusiasm and participation of small and medium-sized enterprises and individuals, and to a certain extent increase the income of the entire carbon trading.

[0066] The present invention is different from the first-order sealed auction and negotiation model commonly used at present. It adopts a fair incentive communication auction to introduce more buyers, and can effectively improve the income of each participating communication seller and social welfare on the basis of satisfying incentive compatibility and individual rationality.

[0067] The present invention fully considers the lack of data and privacy leakage in carbon trading, adds intermediary nodes to the mechanism, and the trading platform modules jointly perform the tasks of supervising the entire auction process and managing data, which can reliably maintain fairness and justice in the carbon trading process.

[0068] The present invention uses multiple trustworthy intermediary nodes and adopts a distributed mechanism based on intermediary nodes. Compared with the distribution of a single intermediary node, it better solves the situation where the seller has an information advantage in a diffusion auction, ensures that the buyer's privacy is not leaked, and indirectly increases the buyer's enthusiasm for participating in the auction.

[0069] The transaction process of the present invention is all carried out on the network platform, and the payment and allocation of the auction are also hosted by the intermediary node, which greatly facilitates the buyers and sellers and improves the efficiency of auction execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 It is a schematic diagram of the structure of the device of the present invention;

[0071] Figure 2This is a diagram of the auction device architecture constructed in the embodiment;

[0072] Figure 3 It is a schematic diagram of the method flow of the present invention;

[0073] Figure 4 A social link diagram for the carbon quota market;

[0074] Figure 5 This is a schematic diagram of the information dissemination flow of the carbon quota market;

[0075] Figure 6 Extract schematic diagram for strong propagation sequence;

[0076] Figure 7 Schematic diagram of the DFDM allocation process;

[0077] Figure 8 A schematic diagram of a DFS spanning tree generated in the embodiment;

[0078] Explanation of the markings in the figure: 1. Seller agent module, 2. Buyer agent module, 3. Trading platform module. DETAILED DESCRIPTION

[0079] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0080] Example

[0081] like Figure 1 As shown, a carbon quota distributed information diffusion auction device includes a seller agent module 1, a buyer agent module 2 and a trading platform module 3, wherein the seller agent module 1 and the buyer agent module 2 are respectively connected to the trading platform module 3 for mutual communication, the seller agent module 1 is connected to the corresponding carbon quota provider for communication, and the buyer agent module 2 is connected to the corresponding carbon quota demander for communication, the seller agent module 1 is used to collect the information uploaded by the carbon quota provider and then distinguish and integrate it, submit the integrated information to the trading platform module 3, and obtain the auction result information from the trading platform module 3 and then send it to the carbon quota provider;

[0082] The buyer agent module 2 is used to submit the information uploaded by the carbon quota demander to the trading platform module 3, and obtain the auction time and auction result information from the trading platform module 3 and then send it to the carbon quota demander;

[0083] The trading platform module 3 is used to integrate and process the information submitted by the seller agent module 1 and the buyer agent module 2 by constructing an undirected graph, generating a DFS spanning tree, and using DFDM to conduct auctions on the DFS spanning tree. At the same time, it manages all intermediary nodes in DFDM to determine the auction time and auction result information.

[0084] This embodiment applies the above method to build Figure 2 The overall architecture of the auction device shown in the figure consists of three parts: the seller agent representing the carbon quota provider, the buyer agent representing the carbon quota demander, and the third-party certification agency trading platform acting as the auctioneer. During the transaction, each agent only communicates with one registered user, and both buyers and sellers need to participate in the auction through the agent.

[0085] 1) The seller agent is mainly composed of the information integration module A1 and the result broadcast module A2

[0086] The main function of the information integration module A1 is to collect the information uploaded by the carbon quota provider, and after differentiation and integration, submit the integrated information to the third-party certification agency trading platform. For example, after a power company adopts a new carbon emission reduction technology, it has some surplus carbon quotas obtained that year. After the company inputs the company's relevant information, the carbon quota for auction and the expected transaction time, the information integration module A1 will integrate the quantity and time separately and send them to the third-party platform.

[0087] The result reporting module A2 is used to provide the auction results to the seller agent, that is, the carbon quota provider. For the seller agent, the reporting results mainly include the transaction price, transaction unit, etc. In addition, after the carbon quota provider finally determines that the "transaction is reached", the result will be reported to the third-party platform by the reporting module A2.

[0088] 2) Each carbon quota demander corresponds to a buyer agent module. Carbon quota demander 1 corresponds to buyer agent module B 1 , mainly composed of information input module B1, auction notification module B2 and result broadcast module B3. Carbon quota demander n corresponds to buyer agent module B n , whose composition is similar to B 1 , namely, information input module B1, auction notification module B2, and result broadcasting module B3.

[0089] The information input module B1 is used to receive information uploaded by the buyer, that is, the party in need of carbon quotas, and submit this information to the third-party certification agency trading platform. For example, a company's demand for carbon quotas increases due to the expansion of its industrial scale. The company can input the number of carbon quotas required and the amount of carbon quotas purchased after market research as information. The information input module B1 will send the company information, company demand and its purchase amount to the third-party platform. In addition, users who participate by spreading information through other buyers also need to enter the basic information of the spreading users in order to obtain the spreading incentives and run the subsequent information diffusion auction algorithm.

[0090] The auction notification module B2 has two main functions: one is to notify the buyer's agent to participate when an auction that meets the needs of the carbon quota demander is released; the other is to notify the buyer's agent of the remaining time of the auction when the buyer's agent has already participated in an auction.

[0091] The result broadcast module B3 is used to provide the auction results to the buyer agent, that is, the carbon quota demander. For the buyer agent, the broadcast result mainly includes whether the bid is successful, the amount of the promotion incentive amount obtained or the amount required to be paid, etc. In addition, after the carbon quota demander finally determines that the "transaction is reached", the result will be reported to the third-party platform C by the broadcast module B3.

[0092] 3) The third-party certification agency trading platform C consists of an information integration module C1, an information processing module C2, an allocation module C3, a payment module C4, a supervision and management module C5, a timing module C6 and an information broadcasting module C7.

[0093] The information integration module C1 organizes and merges the information transmitted by the seller agent and the buyer agent, and corresponds the specific information of each user to each node in the algorithm; then the integrated information will be passed to the information processing module C2, which is responsible for constructing this information into an undirected graph through the order of propagation, and implementing a depth traversal algorithm on the graph and generating a depth-first spanning tree (DFS spanning tree), thereby finding the strong propagation sequence and weak propagation set of the buyer user.

[0094] The allocation module C3 and the payment module C4 together constitute the result module of the entire auction, and the auction is implemented by adopting a distributed fair incentive information dissemination mechanism on the above-mentioned DFS spanning tree to jointly realize the allocation and payment calculation of the auction.

[0095] The function of the supervision and management module C5 is mainly to manage all intermediary nodes in the distributed fair incentive information dissemination mechanism (DFDM). The functions of the intermediary nodes mainly include the following two aspects: first, the seller agent and each buyer agent correspond to a respective intermediary node, and only exchange information with their respective intermediary nodes, ensuring the privacy of both buyers and sellers; second, supervise the entire auction process including information input, and prevent information leakage or algorithm operation errors while ensuring data fairness and reliability.

[0096] The timing module C6 ensures that the seller agent, i.e. the carbon quota provider, completes the auction at the expected time. When the information of the seller agent is integrated and input into the trading platform, the module sets the time and starts the timing. After the timing ends, the information of all buyer agents stops being input, and the scale of the auction is determined.

[0097] The information broadcast module C7 sends the auction allocation and payment as the result to the seller's agent and the buyer's agent, and receives the "confirmation transaction" information from both parties. After receiving the "transaction completed" signal from the two agents, the module will record the time, scale, results and some details of the entire auction, and publish information such as the two parties who completed the transaction and the transaction amount.

[0098] In order to achieve the simplicity of transactions and better facilitate both buyers and sellers, all processes of carbon quota auctions are completed on third-party trading platforms; third-party trading platforms can be certified by the country or organization, further ensuring the privacy of both parties involved in the carbon quota auction and the fairness of information throughout the auction.

[0099] Based on the above auction device, a carbon quota distributed information diffusion auction method is implemented, such as Figure 3 As shown, the following steps are included:

[0100] S1. The carbon quota provider provides auction information and uploads it to the trading platform module for publication through the seller agent module;

[0101] S2. During the auction duration, the carbon quota demander uploads the demand information to the trading platform module through the buyer agent module to join the auction;

[0102] S3. After the auction duration ends, the trading platform module integrates the auction information and demand information;

[0103] S4. The trading platform module applies a distributed fair incentive information diffusion mechanism to execute the auction and determine the allocation and payment information of the auction;

[0104] S5. Based on the auction allocation and payment information, combined with the feedback information from the seller agent module and the buyer agent module, the transaction platform module integrates the auction results and publishes them.

[0105] In combination with the auction device architecture constructed in this embodiment, the process of applying the above method is as follows:

[0106] 1. The carbon quota provider uploads the auction information to the third-party trading platform through the information integration module A1 seller agent, and the platform publishes the auction information through the information broadcast module C7.

[0107] 2. The platform starts the timing module C6, which is the countdown of the “time available for auction”. During the countdown period, the carbon quota demander can use the corresponding buyer agent module B n The information input module B1 in uploads the demand to the third-party trading platform and disseminates the auction information to other users.

[0108] 3. When the countdown ends, the platform information integration and processing modules A1 and A2 process the collected information and obtain the undirected graph, DFS spanning tree, strong propagation sequence, weak propagation set, etc. of the carbon quota auction in turn.

[0109] 4. The platform internally operates a distributed fair incentive information dissemination mechanism through the allocation module C3, payment module C4 and supervision and management module C5 to obtain the final allocation and payment of the auction.

[0110] 5. The platform sends the results to the seller and buyer agent modules through the information dissemination module C7 to let them confirm the transaction. Finally, the platform announces the transaction parties and the auction results.

[0111] The detailed steps are as follows:

[0112] S1: Carbon quota providers provide information and the platform publishes auction information

[0113] The carbon quota provider enters the auction through the seller agent. As a seller of the distributed fair incentive information diffusion mechanism, the main information provided by the seller is the number of carbon quotas for auction. In addition, the seller is required to provide the real user unit and the expected transaction time. After the third-party trading platform confirms that the information is correct, it will publish the auction information including the number of carbon quotas, the provider unit, and the duration on the platform's official website based on the above information.

[0114] S2: During the auction duration, carbon quota demanders join the auction and provide information including demand.

[0115] During the auction, when the carbon quota demander has a certain demand for carbon quotas and is interested in the auction, he can enter the auction through the buyer's agent. As a buyer of the distributed information diffusion mechanism, the main information he needs to provide is his own bid and the referral object to invite him to join (if there is no one, you don't have to enter; if there is more than one object, you can fill in all of them). In addition, he also needs to provide real information about the user's unit. Buyers who enter the above information are deemed to have joined the auction.

[0116] S3: The platform integrates and processes auction information and prepares for the auction.

[0117] When the duration ends, new buyers can no longer participate in the auction. The platform processes and models the collected information internally, and extracts strong propagation sequences and weak propagation sets from all information.

[0118] For the convenience of subsequent explanation, without loss of generality, assume that there is a carbon quota market social chain such as Figure 4 , the information dissemination flow after sorting is as follows Figure 5 . Figure 4 It represents all buyers participating in the entire auction and the information diffusion relationship between them based on the seller's selling needs.

[0119] Figure 4 , 5Each node including sellers and buyers is represented by a circle, where S represents the seller's agent node and A to M represent the buyer's agent nodes. The numbers next to the buyer's agent nodes represent their respective psychological valuations. The edges between nodes represent the information diffusion relationship between neighbors. Neighbors refer to other companies or individuals that have business relationships, social relationships, etc. with the company or individual and have certain contacts. Figure 5 for Figure 4 The information flow constructed according to a certain search order can more clearly show the propagation relationship between nodes, which facilitates the application of auction algorithms on the graph.

[0120] The auction information after platform integration is as follows:

[0121] (1) Seller agent node S and its first-level transaction object set r s (In step S2, no buyer set of referral targets to be invited to join is entered). Figure 4 , 5 For example, the corresponding nodes are the seller node S, the seller's first-level transaction object set r s ={A, B, C}.

[0122] (2) Buyer agent node set N = {1, 2, 3, ..., n}, where the private type θ of each node i i = {v i , r i} respectively represent their psychological valuation towards the auction and the set of neighbors (the set of referral objects that the buyer inputs to invite himself to join in step S2). At this time, the set θ={θ 1 ,θ 2 , ..., θ n} indicates all buyer private types. Figure 4 , 5 For example, all nodes other than the seller node S are buyer nodes, together forming the buyer set N = {A, B, C, ..., M}. The psychological valuation of each buyer is represented by the number next to it, and the neighbor set is represented by the edges between it and other nodes. For example, the psychological valuation of node D is v D =5, neighbor set r D ={A, E, F, G}, whose private type θ D ={5,{A,E,F,G}}.

[0123] (3) Assuming that the node with the highest psychological valuation is ω, we define the auction information to be available from the buyer agent node c j Propagate to ω if and only if the buyer agent node c j When not participating in the auction, ω cannot join the auction. At this time, c j is called a strong propagation buyer. Define the strong propagation sequence of ω (the set of strong propagation buyers arranged in the order of participating in the auction), and use Cω It means that at this time c j ∈C ω .by Figure 5 For example, the buyer node with the highest psychological valuation is M, with a valuation of 13. According to the definition, when the seller S does not initiate the auction, and one of the buyers C, L and M does not join the auction, M cannot receive the auction information, that is, cannot join the auction. According to the definition, M's strong propagation sequence C M ={S, C, L, M}, the strong propagation buyers are C, L, M (represented by red nodes).

[0124] (4) Define the link between two strong communication buyers c j and c j+1 The intermediate buyers are weak propagation buyers, and the set they form is a weak propagation set. Indicates. Figure 5 For example, there are buyer nodes J and I between strong propagation buyers C and L, that is, weak propagation buyers between C and L are J and I (indicated by orange nodes), and the weak propagation set M CL ={J, I}; Since there are no other buyer nodes between the strong propagation buyers L and M, there are no weak propagation buyers and weak propagation sets between L and M.

[0125] (5) The set of intermediary nodes C corresponding to the seller and each buyer P = {P S , P 1 , ..., P i}, where P S represents the intermediary node corresponding to the seller, P i is the intermediary node corresponding to buyer i.

[0126] In summary, the auction information after platform integration is shown in Table 1.

[0127] Table 1

[0128]

[0129] Based on the above information, the third-party certification agency trading platform processing steps are as follows Figure 6 As shown:

[0130] S301: Run a depth-first traversal algorithm (DFS) based on the seller agent S and the buyer agent set N, and generate a DFS tree (depth traversal tree). The starting node of the DFS generated tree is the seller agent, and the remaining nodes are all buyer agents; the edges of the tree represent the neighbor relationship of each node (if two agents are neighbors, they are said to have a neighbor relationship, and an edge is generated between the corresponding two nodes in the DFS tree).

[0131] After running the deep search algorithm (DFS), due to the existence of their respective neighbors, the originally separated seller agent node S and buyer agent node set N are constructed into a DFS spanning tree, in which there is a parent-child relationship between each node (the node that invites other agents to enter the auction is called the parent node of the invited agent node, and vice versa), which is conducive to the strong propagation sequence C ω Extraction.

[0132] S302: The buyer agent node calculates its status and reports it.

[0133] Due to the construction of the DFS spanning tree, each node has a more general judgment whether it is in C ω The buyer agent node needs to calculate the following states: The meanings of these states are shown in Table 2. Then, the buyer will report these states to their corresponding intermediary nodes P i It should be noted that in this process, the buyer intermediary node P i They will check whether the information of their respective agents is true to ensure the authenticity of the information.

[0134] Table 2

[0135]

[0136] S303: Seller intermediary node P s Publish the node ω with the highest psychological valuation and its status.

[0137] Seller intermediary node P in the supervision and management module C5 s After finding all buyers’ quotations, we compare and find the node ω with the highest psychological valuation (ω’s psychological valuation is v′ ω ) and informs ω, through its intermediate node P ω Report its partial status dt ω (the time when the user corresponding to the ω node was first invited), and then report ω to all other buyer proxy nodes. Note that the seller intermediary node P s Only the psychological valuation reported by each buyer is known, but the topological structure of the network is unknown, that is, the neighbor set of each buyer is unknown, thereby preventing excessive exposure of information. After collecting all psychological valuations, the seller intermediary node will pass the information to the allocation module C3 and the payment module C4 for subsequent use.

[0138] S304: Seller intermediary node P s According to the information reported by the buyers, we compare and select the strong propagation sequence C ω Buyer node on .

[0139] Through the above steps S302 and S303, the general buyer proxy node has known the buyer proxy node ω. If it is the key propagation buyer of ω, in order to successfully win the auction or a higher propagation incentive, it needs to inform the corresponding intermediary node P of its willingness to participate in the auction. i , so that subsequent intermediary nodes can accurately include it as a payment object.

[0140] Then, each intermediary node that receives the auction intention can identify whether the buyer belongs to the strong propagation sequence C according to the information reported by the buyer node and the information status of the ω buyer agent node after status comparison and node screening. ω If buyer i is a strong communication buyer, then the corresponding intermediary node P i dt i Report to the seller's intermediary node P s ; If buyer i is not a strong propagation buyer, he will be penalized for false positives. In particular, in this process, each intermediary node P i The information will be verified by comparing the status of all reporting buyer nodes, thereby playing a role in information supervision.

[0141] S305: Seller intermediary node P s According to dt i Arrange the key nodes in ascending order to generate a strong relay sequence.

[0142] Since there is a strict invitation order from sellers to buyer nodes in the strong propagation sequence Cω, it needs to be sorted according to the time of arrival at each buyer. s According to the time dt of the first invitation of each node i Sort in ascending order, and record the seller as S. The strong propagation sequence after sorting is C ω It is {S, 1, 2, ..., ω}, where 1, 2, etc. represent a buyer agent node.

[0143] Get the strong propagation sequence C ω After all the strong propagation buyer nodes, count the two strong propagation nodes c j With c j+1 All buyers on the communication path (one party can make the other party participate in the auction by information communication, and the communication route between the two is called the communication path) are regarded as weak communication buyers of the two strong communication buyers, and together constitute the weak communication set

[0144] S4: The third-party platform applies a distributed fair incentive information diffusion mechanism to execute the auction internally.

[0145] Allocation phase:

[0146] The strong propagation sequence C obtained by the previous step S3 ω , at this time, the auction only occurs on this sequence. That is, the seller intermediary node P s The information of the strong propagation sequence (including the position of each node in the sequence and its own psychological evaluation), the weak propagation node and other nodes are passed to the distribution module C3, and then C3 Figure 7 The steps shown determine the allocation of the auction.

[0147] S401: According to the strong propagation sequence C ω , start the transaction from the first node after the seller (the node closest to the seller).

[0148] In order to ensure the individual rationality of the mechanism, sellers need to give buyers who spread information certain incentives or subsidies to ensure that they spread information. At this time, all strong buyers before the auction winner play the role of information dissemination. Therefore, the distributed fair incentive information dissemination mechanism needs to start the auction from the first strong buyer node.

[0149] S402: The allocation module C3 provides each buyer's intermediary node P i The required buyer node information is provided by the buyer intermediary node P i Calculate the externality of the current strong propagation buyer node.

[0150] In order to ensure the incentive compatibility of the mechanism, the buyer intermediary node P i It is necessary to calculate the externality of the current strong propagation buyer proxy node, that is, the highest bid among all other network nodes except the selected set. This step can be divided into two parts: ① The intermediary node of the first strong propagation node calculates the externality. At this time, due to insufficient external information of the network, the allocation module C3 will report the information required by the intermediary node (the buyer's psychological valuation of the non-strong propagation sequence) to it for its calculation; ② The intermediary node of the subsequent strong propagation node calculates the externality, and only needs to obtain the externality information of the previous strong propagation node and the weak propagation information between the two nodes.

[0151] S403: The auction proceeds in sequence or ends.

[0152] According to the size of the psychological valuation of the next buyer i+1 and the psychological valuation of the current transaction buyer i, this step can be divided into two cases, A and B.

[0153] S403A: The next buyer i+1 needs to pay except for itself and the weak propagation set When the externality of buyer 2 is higher than the psychological valuation of the previous buyer, it will replace the previous buyer to join the auction. For example, when the psychological valuation of buyer 2 is higher than the externality of buyer 1, it will replace buyer 1 to join the auction and become the buyer currently allocated carbon quotas.

[0154] S403B: The next buyer needs to pay except for itself and the weak propagation set When the externality is lower than the psychological valuation of the previous buyer, the current transaction buyer directly wins the auction and inputs the information into the payment module C4.

[0155] The allocation process is performed by the allocation module C3 and the intermediate node P i , P s Execution. Buyer intermediary node P i While acting as an auctioneer, a distributed computing mechanism is also introduced. If there is no buyer intermediary node, the information of all nodes is required every time the externality is calculated; after the introduction of the buyer intermediary node, since the previous buyer intermediary node has retained the information of the previous externality calculation after the first calculation, the next buyer intermediary node only needs to obtain the information between two strong propagation nodes when performing calculations (that is, only the intermediary nodes corresponding to the strong propagation nodes need to exchange information), which greatly reduces the scope of information acquisition, reduces the duplication of information acquisition work, and also directly reduces the risk of privacy leakage.

[0156] Payment stage:

[0157] The payment process is closely related to the distribution results, and can be divided into two stages according to the distribution order: the initial payment and the re-payment stage. The initial payment solves the payment of the main amount, which occurs between strong communication buyers; the re-payment solves the payment amount less than the former, which occurs between all communication buyers, but can effectively ensure the enthusiasm of buyers to participate in the mechanism.

[0158] 1) Initial payment stage:

[0159] In the initial payment stage, the main payment recipients are sellers and strong communication buyers. The payment rules for relevant buyers are divided into three situations:

[0160] (1) For the seller (S), the externality amount paid by the strong communication buyer (strong communication buyer No. 1) with whom the seller has the first transaction in step S401 is all the payments that the seller can obtain in the initial payment stage. It can be expressed as:

[0161]

[0162] (2) For the buyer who finally obtains the item (c h The amount paid by ) (who must be one of the strong communication buyers) is its externality The formula is:

[0163]

[0164] (3) For non-winner strong communication buyers (c j ∈C ω\c h , where '\c h 'Indicates' except c h ') will get a certain amount of communication benefits, the amount = the externality paid by the next strong communication seller except for itself and the weak communication nodes between the two - Self-paying externalities Expressed as:

[0165]

[0166] The formula for the buyer node's initial payment is as follows:

[0167]

[0168] in (* represents any set) represents the highest psychological valuation among all nodes except the * set, that is, the externality except the * set; C j represents the jth strong communication buyer, c h Indicates the buyer who finally obtains the item; C ω is a strong propagation sequence, Between c j and c j+1 A weakly spreading set between two strongly spreading buyers.

[0169] 2) Repayment stage:

[0170] Two-strong transmission of seller payment externalities and If there is a difference, the difference will be distributed to all buyers who meet the payment requirements during the payment phase. First, determine the number of people who need to pay (Weakly propagated set + the jth strong propagation buyer), and then decide the payment amount based on the difference between the buyers.

[0171] In the re-payment stage, the main payment objects are weak propagation buyers, strong propagation buyers and sellers. The relevant buyer re-payment rules can be divided into three situations:

[0172] (1) For strong communication buyers including winners (c j ∈C ω ), the amount to be paid (p′ i )for Its value is equal to the externality of the highest psychological valuation buyer in the weak communication set divided by itself and -Externalities of all buyers except self and weakly propagated set and divided by the number of weak propagation buyers The sum of the number of sellers 1 is expressed as:

[0173]

[0174] (2) For weakly propagated buyers The amount to be paid (p′ i ) is -R i , whose value is equal to the externality between itself and the jth strong communication buyer -Externalities of all buyers except self and weakly propagated set and divided by the number of weak propagation buyers The sum of the number of sellers 1 is expressed as:

[0175]

[0176] (3) For the seller (S), after all buyers who meet the payment requirements have completed their payments, if the total payment amount is lower than the difference between the payment externalities of the two strong-propagation sellers j and j+1, the remaining amount shall be paid in full by the seller.

[0177] The formula for the buyer node to re-pay is as follows:

[0178]

[0179] The specific expression of propagation incentive is:

[0180]

[0181] In the formula, c j , Same as during the initial payment phase; Indicates between c j-1 and c j The buyer with the highest psychological valuation in the weak communication set between two strong communication buyers.

[0182] Like the allocation process, the payment process is executed by the payment module C4 and the intermediary node. While acting as an "auctioneer", the intermediary node also introduces a distributed computing mechanism. If there is no buyer intermediary node, the information of all nodes is required every time the externality is calculated; after the introduction of the buyer intermediary node, since the previous buyer intermediary node has retained the information of the previous externality calculation after the first calculation, the next buyer intermediary node only needs to obtain the information between the two strong propagation nodes (that is, only the intermediary nodes corresponding to the strong propagation nodes need to exchange information), which greatly reduces the scope of information acquisition, reduces the duplication of information acquisition work, and also directly reduces the risk of privacy leakage.

[0183] S5: The third-party certification agency trading center integrates the auction information and publishes the auction results.

[0184] After steps S3 and S4, the carbon quota auction has ended, and all the allocation and payment information is stored in the information broadcast module C7. Then, the third-party platform will filter out the required information, such as sellers' earnings, final winners, buyer information dissemination subsidy incentives, and the amount of money the winner needs to pay, and publish this information to the respective seller agents and buyer agents so that the agents can "confirm the transaction" as soon as possible. In addition, the information broadcast module C7 will also publicize the sellers and final winners to achieve fairness and justice in the auction information.

[0185] The following is a detailed description of the intermediary nodes and the designed auction mechanism:

[0186] 1. The role of intermediary nodes

[0187] During the entire auction process, this solution uses multiple intermediary nodes managed by the supervision and management module C5 and designs a new distributed processing mechanism to solve the problem of privacy leakage in carbon trading auctions and achieve privacy protection for both buyers and sellers. Intermediary nodes play a key role in protecting privacy at the following levels:

[0188] In determining the strong propagation buyer sequence and the weak propagation buyer set, the seller intermediary node P s Only the psychological valuation of the buyer node is obtained, and the buyer intermediary node P i The distributed mechanism obtains information such as the neighbor set of the reporting buyer respectively; compared with the traditional centralized mechanism that does not introduce intermediary nodes, the distributed mechanism does not need to obtain the topological structure of the entire network, which reduces the spread of private information between buyers and greatly reduces the risk of information leakage.

[0189] In the payment and allocation phase, each time the externality is calculated, only a specific buyer intermediary node P is required i Calculations are performed between nodes and some specific node information is obtained (for example, when determining the allocation, it is only necessary to obtain the psychological valuations of two strong communication nodes and the weak communication buyers between them); compared with the traditional centralized mechanism without the introduction of intermediary nodes, the distributed mechanism does not require the psychological valuation information of all buyers for each externality calculation, which effectively prevents the exposure of buyer information; compared with the distributed mechanism of a single intermediary node, multiple intermediary nodes can better prevent the problem of excessive concentration of information.

[0190] During the entire auction process, all intermediary nodes will conduct a certain degree of verification of the information each time they receive or obtain node information, ensuring the authenticity of the information and the fairness of the auction.

[0191] 2. Designing an auction mechanism that includes individual rationality and incentive compatibility

[0192] 2.1 Individual Rationality

[0193] In the auction mechanism, the individual benefits of both buyers and sellers are non-negative. In the mechanism of this scheme, this property can ensure that carbon quota demanders actively participate in the auction and spread auction information, solving the problem of low participation of small and medium-sized enterprises and individuals.

[0194] Theorem 1: The distributed fair incentive information dissemination mechanism (DFDM) satisfies individual rationality.

[0195] Proof: After the DFDM mechanism is implemented, only the income of strong and weak propagation buyers may not be 0. The income of other buyers is 0. And R i is defined as the re-payment amount, obviously R i ≥0.

[0196] For buyer node i=c j ∈C ω \c h (Non-winner strong communication buyer), according to the payment formula (1-3) (2-1) we can know its final profit Due to c j Always precedes in the information dissemination sequence With c j+1 , so the subset Externality Set Equivalent to Therefore,

[0197] For buyer node i=c h (Winning buyer), according to the allocation rule, the equivalent amount of the auction item won is defined as From the payment formula (1-2) (2-1), we can know that the final profit at this time is

[0198] For buyer nodes (Weakly propagated buyer), according to the payment formula (2-2), its final profit can be known

[0199] In summary, for the three types of strong or weak communication buyers, their benefits are not less than 0; for other buyers, their benefits are 0, so the DFDM mechanism satisfies individual rationality.

[0200] 2.2 Incentive Compatibility

[0201] In the auction mechanism, the user's real bid (psychological valuation) is the dominant strategy and the benefits obtained by the user who adopts the real bid are non-negative. In the mechanism of this scheme, this property can ensure that the psychological estimates and neighbors proposed by the buyer's agent are true and reliable, that is, there is no problem of buyers participating in the auction but not disseminating, which solves the problem of insufficient liquidity in the traditional carbon trading market.

[0202] Theorem 2: The distributed fair incentive information dissemination mechanism (DFDM) satisfies incentive compatibility.

[0203] Proof: All buyers in the DFDM mechanism are divided into four types: non-winner strong propagation buyers, weak propagation buyers, winners and other buyers, and the four need to be analyzed separately.

[0204] (1) For buyer node i = c j ∈C ω \c h (Non-winner strong spread buyer).

[0205] When its neighbor set When fixed, since the neighbor set has not changed, according to the allocation rule, it cannot falsely report psychological valuation Become a weak communication buyer, that is, it is still a strong communication buyer.

[0206] From the payment formula (1-3) (2-1), we can know its final profit It has nothing to do with its own quotation, that is, the income does not change; if the node passes the psychological valuation by high Become the new winner, according to the definition of re-payment, will not change, but due to the externality paid by the next strong propagation buyer Will be greater than the true psychological valuation of the new winner Its final profit will be reduced, that is,

[0207] When its psychological evaluation Fixed, false reporting of neighbor sets (neighborhood set can only be underreported).

[0208] 1.1 The node is still a non-winner strong propagation buyer. As the number of neighbors of the node decreases, the selection set of the next node to pay the externality may also decrease, that is, The corresponding externality is reduced, i.e. The amount of his own expenditure and re-payment amount The final profit may decrease.

[0209] 1.2 The node becomes a weak propagation buyer with non-negative benefits. At this time, its utility becomes Due to reduced revenue It must be non-negative, so the final return decreases.

[0210] 1.3 The node becomes the new winner. Similar to the situation of false psychological valuation, there is That means the final profit decreases.

[0211] 1.4 The node becomes a non-strong or weak propagation buyer, and its profit is always 0, and the final profit will also decrease.

[0212] (2) For buyer nodes (Weakly spread buyer).

[0213] When its neighbor set When fixed, since the neighbor set has not changed, according to the allocation rule, it cannot falsely report psychological valuation Become a strong communication buyer, that is, it is still a weak communication buyer. The final profit can be known from the payment formula (2-2) It has nothing to do with its own quotation, that is, the profit has not changed; if the node passes the high psychological valuation Become the new winner, and its final profit becomes Due to the externality of its expenditure Must be greater than its true psychological valuation Therefore, the final profit will decrease.

[0214] When its psychological evaluation Fixed, false reporting of neighbor sets (neighborhood set can only be underreported). Due to weak propagation, the final revenue of the buyer is It has nothing to do with the objects that the node propagates, that is, no matter how much the number of its neighbors decreases, it will not affect its final benefit.

[0215] (3) For buyer node i = c h (Winner Buyer).

[0216] When its neighbor set When fixed, the winner’s final income can be obtained according to the proof in Theorem 1: It has nothing to do with its own bid, that is, under the premise that the distribution has not changed, no matter how the winner lies about the psychological valuation, the final benefit will not change; if the winner reports a lower bid and becomes a non-winner strong communication buyer, according to the distribution rules, the winner's true psychological valuation must not be lower than the externality paid by the next strong communication node, and its final benefit That is, the final profit is reduced; if the winner reports a lower bid and becomes a weak propagation buyer, according to the allocation rule, the winner's real psychological valuation must not be lower than the externality paid by the next strong propagation node, that is, Therefore, the final revenue of the node is That is, the final profit will not increase.

[0217] When its psychological evaluation Fixed, false reporting of neighbor sets (The neighbor set can only be underreported). According to the allocation rule, the winner’s neighbors do not affect the allocation, that is, the allocation result remains unchanged. According to the payment formula (1-2) (2-1), the winner’s utility is It has nothing to do with the neighbors, so the final payoff will not change if the winner lies about his neighbors.

[0218] (4) For other buyers who do not belong to the above three (using c n express).

[0219] When its neighbor set When fixed, if the buyer is not the child node of the winner, he can change the allocation result by raising his bid. , it will become a strong propagation buyer, and the payment will be the initial highest valuation on the network. According to the allocation rules, the payment must be greater than its real psychological valuation, so the final profit is reduced; if the buyer is a child node of the winner, no matter how it bids, the allocation result will not change, and the final profit will still be 0.

[0220] When its psychological evaluation Fixed, false reporting of neighbor sets (the neighbor set can only be underreported). Since the psychological estimate has not changed, the allocation result will not change, that is, its final benefit is still 0.

[0221] In summary, any buyer node that truthfully reports its private type (psychological valuation and neighbor set) will gain the most benefits, and false reporting will not increase benefits, so the DFDM mechanism satisfies the incentive compatibility characteristics.

[0222] In addition, to verify the effectiveness of this solution, this embodiment uses the transaction of a local exchange under the national carbon quota market on a specific day as an example for application description. Taking the power generation industry as an example, in order to respond to the call of the low-carbon policy, a power plant upgraded a 600MW thermal power generating unit to a supercritical generating unit, that is, by increasing the boiler pressure and temperature, the power generation efficiency was improved while reducing the coal consumption per unit of electricity. After the adjustment and upgrade, the annual carbon dioxide emissions of the thermal power plant decreased by 30%, and the number of carbon quotas allocated that year remained 600 tons.

[0223] After pledging the National Certified Emission Reductions (CCER), all 600 tons of carbon quotas met the CCER standards, and the company planned to sell all the carbon quotas at one time before the end of the year.

[0224] First, this solution is used for online auction. In order to maximize profits, the company adopts online distributed fair incentive information diffusion auction. The specific auction process is as follows:

[0225] S1: The seller provides the auction information and the third-party platform publishes it.

[0226] The company input its information into the official website of a certain energy exchange (a nationally certified third-party trading platform) and commissioned the exchange to conduct an online information diffusion auction. After the exchange verified the information, it published the information on its official website.

[0227] The carbon quota suppliers (sellers) and auction information are shown in Table 3 below:

[0228] Table 3

[0229] Carbon quota supplier Auction announcement time Auction Deadline CCER Trading Volume Average opening price Power Generation Company 2023.12.1 2023.12.31 600 tons 79 yuan / ton

[0230] S2: Buyers participate in the auction within the specified time.

[0231] After step S1, users who have carbon quota demand and are able to purchase 600 tons of CCER (carbon quota) will provide their information and join the auction. At the same time, they can also spread the information to users who have more urgent demand and higher bids to benefit from it.

[0232] S3: Third-party platform integrates information.

[0233] At the end of the specified time, the third-party platform first integrates the collected buyer information to obtain the following summary table 4 (where the bids of the carbon quota demanders are input by the users themselves after market research, and the neighbors are obtained after sorting out the information input by the buyers, which represents the relationship of buyer information diffusion).

[0234] Table 4

[0235] Carbon quota demand side Bid (unit: Yuan / ton) Neighbor Company A 79 S,D Self-employed B 75 S,D,C Company C 76 S,E,F,G,B Company D 80 H, J, A, B Company E 82 K,C Enterprise F 78 K,L,G,C Self-employed 81 F,C Enterprise H 80 I,D Enterprise I 79 J,H Enterprise J 78 I,D Company K 83 N,E,F Self-employed 81 F Self-employed 80 O,I Enterprise N 84 P,Q,K Company O 79 M Company P 80 Q,N Company Q 79 P,N

[0236] Based on the above integrated information, the third-party platform runs a deep traversal algorithm on the information to obtain the DFS spanning tree as follows: Figure 8 As shown. Each node in the DFS tree represents the seller (S) and the buyer (A, B, ...) participating in the auction; the edge between the nodes represents the neighbor relationship between the two nodes, that is, there is an invitation-invitee relationship between the two buyers; the number next to the buyer's node represents his psychological valuation of the auction. In particular, the set of intermediary nodes (represented by blue nodes) is independent of the auction system and is the supervision and management module C5, which collects auction information and calculates the distribution payment.

[0237] After obtaining the DFS spanning tree, the seller and buyer intermediary node set (supervision and management module C5) determines the strong propagation sequence C according to the position and calculation status of each buyer agent node on the tree shown in Table 2. ω ={S,C,K,N}. In this sequence, excluding the seller S, the remaining buyers C, K, and N are strong propagation buyers (indicated in red in the figure), among which N is the buyer with the highest psychological valuation ω; then the intermediary node determines the weak propagation buyer set M between C and K ck={E, F, G} (indicated by orange in the figure). Since there is no buyer node between the proxy nodes K and N, there is no weak propagation set.

[0238] S4: The third-party platform determines the auction allocation and payment.

[0239] The supervision and management module C5, the allocation module C3 and the payment module C4 jointly determine the allocation and payment of this auction according to the operating mechanism of the distributed fair incentive information diffusion mechanism.

[0240] Allocation process:

[0241] According to Section 2.4 above, the highest buyer is N, so the allocation of auction objects is only in the strong propagation sequence C ω , that is, it is carried out on S, C, K, and N.

[0242] First, the transaction occurs between the seller S and the buyer C. At this time, C "owns these carbon quotas", that is, at this time, C spends the amount and temporarily owns the auction object, that is, the right to control the carbon quota. According to the auction rules, it is necessary to continue to judge whether the next strong propagation node K is qualified to win the auction. After excluding the buyer C and all nodes that can only be propagated through C, the highest psychological valuation is 80 yuan / ton.

[0243] Next, the transaction takes place between buyer C and buyer K. The entire network excludes buyer K, all nodes that can only be propagated through K, and the weak propagation set M. ck After that, the highest psychological valuation is 81 yuan / ton of L. This valuation is higher than C's psychological valuation of 76 yuan / ton, so at this time, according to the auction rules, C hands over these carbon quotas to K and obtains the amount paid by K.

[0244] Finally, the transaction occurs between buyer K and buyer N. After excluding buyer N and all nodes that can only be transmitted through N, the highest psychological valuation is 83 yuan / ton (not higher than K's psychological valuation of 83 yuan / ton). That is, according to the auction rules, N cannot win the auction object at this time, and the carbon quota is finally obtained by K.

[0245] Payment process:

[0246] During the allocation process, since buyer N did not participate in the auction, and the final winner was buyer K, the initial allocation occurred among S, C, and K. According to the initial allocation rules, seller S received 48,000 yuan (80 yuan / ton as the unit price, 80*600=48,000); the final winner K needs to pay 49,200 yuan (82 yuan / ton as the unit price, 82*600=49,200) to remove its own externalities and obtain the 600 tons of carbon quota; the non-winner strong communication buyer C can obtain the communication incentives of removing buyer K and the weak communication set M ckExternalities (81*600=48600) - Externalities other than itself (80*600=48000) = 600 yuan.

[0247] Since the externality obtained by buyer C is 48,600 yuan (81 yuan / ton as the unit price, 81*600=48,600), and the externality paid by buyer K is 49,200 yuan (82 yuan / ton as the unit price, 82*600=49,200), the difference between the two is 600 yuan, so this 600 yuan will be paid to the weak propagation set M through the repayment process. ck There are buyers E, F, G, buyer K and seller S in the.

[0248] First, determine the number of buyers participating in the repayment stage as 4 (E, F, G, K), and then calculate the repayment amount for the four in turn. For the strong propagation node K, if it does not participate in the auction, the final winner will become E. At this time, the externality unit price paid by buyer E is 81 yuan / ton, and the externality difference is 0, and the repayment amount is 0 / 4=0 yuan; for the weak propagation node E, if it does not participate in the auction, the final winner is still K, but its externality unit price becomes 81 yuan / ton, so the repayment amount is also 0 yuan; for the weak propagation nodes F and G, their non-participation in the auction does not affect the final result, so the repayment amount obtained by F and G is 600 / 4=150 yuan. Finally, the externality difference remains 600-150*2=300 yuan, all of which are obtained by the seller S.

[0249] S5: The third-party platform publishes the auction results.

[0250] The final auction result obtained from step S4 is shown in Table 5 below:

[0251] Table 5

[0252] User Information User Type Auction revenue / yuan Auction expenditure / yuan Spread incentives / yuan Company S Seller 48300 / / Company C Non-Winning Buyer / / 600 Enterprise F Non-Winning Buyer / / 150 Self-employed Non-Winning Buyer / / 150 Company K Winner / 49200 /

[0253] The third-party platform transmits the above auction results to the buyer agent nodes C, F, G, K and the seller agent node through the information broadcast module C7, and announces their income or required payment on the corresponding user interface. The user clicks "Confirm Transaction" to indicate that the auction is completed. Other buyers' interfaces only show that they did not win the auction, and will not reveal too much other information. After the auction is completed, the third-party platform will upload the specific information of this auction to a Shanghai Stock Exchange database and update it for data storage and use.

[0254] Then, the traditional auction plan is used for comparison. If the traditional auction method is used, the buyers participating in the auction will not spread the auction information to other buyers they know in order to maximize their own interests. Therefore, only A, B, and C will receive the notification and participate in the auction. The profit at this time is 79*600=47,400 yuan (A's psychological valuation is the highest, which is 79 yuan / ton).

[0255] By adopting the distributed fair incentive information dissemination mechanism proposed by the present invention, the buyer finally obtains a profit of 48,300 yuan, which is 900 yuan higher than the 47,400 yuan when only A, B, and C participate without dissemination, thus expanding the seller's profit to a certain extent.

[0256] At the same time, because the present invention designs a new distributed mechanism containing intermediary nodes during the auction process, the information required for calculating the externalities of the system or node each time during the allocation and payment process is greatly reduced. Specifically, the original need to know the psychological bids and neighbor information of all nodes is reduced to only the information between some specific nodes (C nodes and K nodes in this case), which effectively reduces the risk of information leakage. In addition, the expansion of multiple intermediary nodes prevents the information between buyers from being overly concentrated (in this case, information exchange and calculation are only performed between the two intermediary nodes C and K), effectively protecting information security. In addition, the intermediary nodes can also verify the authenticity of the buyer's information during the auction process, playing a role in supervising and managing information while protecting privacy.

[0257] Finally, the results are compared. By comparing the above traditional auction method with the method proposed by the present invention, it can be clearly seen that:

[0258] (1) The new auction mechanism proposed in the present invention can effectively increase the seller's profit; at the same time, in terms of social welfare, it has a more obvious improvement effect than the traditional auction method.

[0259] (2) The new mechanism proposed in this invention provides certain incentives to the dissemination nodes, thereby encouraging more individual users and small enterprises to participate while improving the efficiency, and to a certain extent solves the shortcoming of weak liquidity of carbon quota trading.

[0260] (3) Different from the previous centralized auction mechanism, the present invention introduces a distributed mechanism by adopting multiple intermediary nodes, which effectively protects the privacy of sellers and buyers.

[0261] (4) The intermediary nodes in the present invention also play a role in data supervision and management, and compared with the traditional carbon quota auction mechanism, they can prevent the occurrence of data falsification to a certain extent.

[0262] In summary, this scheme introduces a fair incentive information diffusion mechanism to model each enterprise, individual and their neighbors (the meaning of "neighbors" here is: other enterprises or individuals with certain contacts that have business relationships, social relationships, etc. with the enterprise or individual) as a social network graph, and adopts measures such as dissemination incentives to effectively improve the enthusiasm and participation of small and medium-sized enterprises and individuals, which can increase the income of the entire carbon trading to a certain extent; in response to the problem of lack of information management system, the present invention, based on the auction mechanism, draws on the form of distributed mechanism design, adds trustworthy intermediary nodes, and at the same time plays a role in protecting personal privacy and managing and verifying data.

Claims

1. A carbon quota distributed information diffusion auction device, characterized in that: It includes a seller agent module, a buyer agent module and a trading platform module. The seller agent module and the buyer agent module are respectively connected to the trading platform module for mutual communication. The seller agent module is connected to the corresponding carbon quota provider for communication, and the buyer agent module is connected to the corresponding carbon quota demander for communication. The seller agent module is used to collect the information uploaded by the carbon quota provider and then distinguish and integrate it, submit the integrated information to the trading platform module, and obtain the auction result information from the trading platform module and then send it to the carbon quota provider; The buyer agent module is used to submit the information uploaded by the carbon quota demander to the trading platform module, and obtain the auction time and auction result information from the trading platform module and then send it to the carbon quota demander; The transaction platform module is used to integrate and process the information submitted by the seller agent module and the buyer agent module by constructing an undirected graph, generating a DFS spanning tree, and using DFDM to conduct auctions on the DFS spanning tree, while managing all intermediary nodes in DFDM to determine the auction time and auction result information.

2. A carbon quota distributed information diffusion auction method, applied to a carbon quota distributed information diffusion auction device as claimed in claim 1, characterized in that: The following steps are involved: S1. The carbon quota provider provides auction information and uploads it to the trading platform module for publication through the seller agent module; S2. During the auction duration, the carbon quota demander uploads the demand information to the trading platform module through the buyer agent module to join the auction; S3. After the auction duration ends, the trading platform module integrates the auction information and demand information; S4. The trading platform module applies a distributed fair incentive information diffusion mechanism to execute the auction and determine the allocation and payment information of the auction; S5. Based on the auction allocation and payment information, combined with the feedback information from the seller agent module and the buyer agent module, the transaction platform module integrates the auction results and publishes them.

3. A carbon quota distributed information diffusion auction method according to claim 2, characterized in that: The specific process of step S1 is as follows: the carbon quota provider enters the auction through the seller agent module and uploads the carbon quota quantity for auction, the actual unit to which the user belongs and the expected transaction time; After the trading platform module confirms that the uploaded information is correct, it will release auction information including the carbon quota quantity, provider unit, and duration.

4. A carbon quota distributed information diffusion auction method according to claim 2, characterized in that: The specific process of step S2 is as follows: during the auction duration, the carbon quota demander enters the auction through the buyer agent module and uploads its own bid, the referral object, and the real information of the unit to which the demander belongs.

5. A carbon quota distributed information diffusion auction method according to claim 2, characterized in that: In step S3, the trading platform module integrates the auction information and the demand information to obtain an undirected graph of the carbon quota auction, and implements a deep traversal algorithm on the undirected graph to generate a DFS spanning tree, thereby finding the strong propagation sequence and weak propagation set of the buyer user.

6. A carbon quota distributed information diffusion auction method according to claim 5, characterized in that: The undirected graph of carbon quotas includes: Seller agent node S and its first-level transaction object set r s ; The buyer agent node set N = {1, 2, 3, ..., n}, the private type of each buyer node i is θ i = {v i , r i },v i and r i They represent the buyer node i’s psychological valuation of the auction and the neighbor set respectively; The set of intermediary nodes C corresponding to the seller and each buyer P = {P S ,P1,…,P i }, where P S represents the intermediary node corresponding to the seller, P i is the intermediary node corresponding to buyer i.

7. A carbon quota distributed information diffusion auction method according to claim 6, characterized in that: The specific process of generating the DFS spanning tree and finding the strong propagation sequence and weak propagation set of the buyer user in step S3 is as follows: S31, running a depth-first traversal algorithm according to the seller agent S and the buyer agent set N, and generating a DFS spanning tree, wherein the starting node of the DFS spanning tree is the seller agent and the remaining nodes are all buyer agents; the edges of the tree represent the neighbor relationship of each node; S32, the buyer agent node calculates its own status and reports it to the corresponding buyer intermediary node. The buyer agent node status is The corresponding values ​​are the parent node set of buyer i, the time when buyer i was first invited, the time when the last buyer of subtree i was invited, the shortest time to reach buyer node i, the buyer node with the highest psychological valuation in subtree i, and the psychological valuation of the buyer node with the highest psychological valuation in subtree i; S33, seller intermediary node P s Publish the node ω with the highest psychological evaluation and its status; S34, seller intermediary node P s According to the information reported by the buyer, the strong propagation sequence C is compared and screened in turn. ω The buyer node on the , where if and only if the buyer proxy node c j When not participating in the auction, ω cannot join the auction. At this time, c j is called a strong propagation buyer, and defines the strong propagation sequence of ω as C ω , c j ∈C ω ; If buyer i is a strong communication buyer, then the corresponding intermediary node P i dt i Report to the seller's intermediary node P s ; If buyer i is not a strong propagation buyer, he will be penalized for false reporting; In this process, each intermediate node P i The information will be verified by comparing the status of all reporting buyer nodes, thereby playing a role in information supervision; S35, seller intermediary node P s According to dt i Arrange the key nodes in ascending order to obtain the strong propagation sequence C ω And all strong propagation buyer nodes, count two strong propagation nodes c j With c j+1 All buyers on the communication path are regarded as weak communication buyers of the two strong communication buyers, and together constitute the weak communication set 8. A carbon quota distributed information diffusion auction method according to claim 7, characterized in that: The specific process of determining the auction allocation information in step S4 is as follows: S41, according to the strong propagation sequence C ω , start trading from the first node after the seller, i.e. the node closest to the seller; S42, buyer intermediary node P i Calculate the externality of the current strong propagation buyer node; S43, compare the psychological valuation of the next buyer i+1 with the psychological valuation of the current transaction buyer i. If the next buyer i+1 needs to pay the amount of money except for itself and the weak propagation set When the externality of the buyer is higher than the psychological valuation of the previous buyer, the buyer will replace the previous buyer to join the auction; If the next buyer needs to pay for itself and the weak propagation set When the externality is lower than the psychological valuation of the previous buyer, the current transaction buyer directly wins the auction.

9. A carbon quota distributed information diffusion auction method according to claim 8, characterized in that: The payment information of the auction determined in step S4 includes the initial payment and the re-payment. The initial payment is made to the seller and the strong propagation buyer. The payment rules of the relevant buyers are divided into three situations: (1) For seller S, the externality amount paid by the strong propagation buyer in the first transaction with him in step S41 is all the payments that the seller can obtain in the initial payment stage, which can be expressed as: (2) For the buyer who finally obtains the item h The amount paid is the externality Expressed as: (3) For non-winner strong communication buyers c j ∈C ω \c h , where '\c h 'Indicates' except c h ', will get a certain amount of dissemination benefits, amount (p i ) is equal to the externality paid by the next strong communication seller except for itself and the weak communication nodes between them - Self-paying externalities Expressed as: Therefore, the formula for the buyer node's initial payment is generally expressed as follows: In the formula, represents the highest psychological valuation among all nodes except the * set, that is, the externality except the * set, * represents any set; c j represents the jth strong communication buyer, c h Indicates the buyer who finally obtains the item; C ω is a strong propagation sequence, Between c j and c j+1 A weak diffusion set between two strong diffusion buyers; The repayment targets are weak propagation buyers, strong propagation buyers and sellers. The repayment rules for relevant buyers are divided into three situations: (1) For strong communication buyers including winners c j ∈C ω , the amount to be paid (p i ')for Its value is equal to the externality of the highest psychological valuation buyer in the weak communication set divided by itself and -Externalities of all buyers except self and weakly propagated set and divided by the number of weak propagation buyers The sum of the number of sellers 1 is expressed as: (2) For weakly propagated buyers The amount to be paid (p i ′) is -R i , whose value is equal to the externality between itself and the jth strong communication buyer -Externalities of all buyers except self and weakly propagated set and divided by the number of weak propagation buyers The sum of the number of sellers 1 is expressed as: (3) For seller S, after all buyers who meet the payment requirements have completed their payments, if the total payment amount is lower than the difference between the payment externalities of the two strong propagation sellers j and j+1, the remaining amount shall be paid by the seller in full; The formula for the buyer node to re-pay is as follows: The propagation incentive is expressed as: In the formula, c j , Same as during the initial payment phase; Indicates between c j-1 and c j The buyer with the highest psychological valuation in the weak communication set between two strong communication buyers.

10. A carbon quota distributed information diffusion auction method according to claim 9, characterized in that: The specific process of step S5 is: filter out the seller's income, the final winner, the buyer's information dissemination subsidy incentive and the amount the winner needs to pay from the allocation information and payment information, and then publish them to the seller's agent module and the buyer's agent module, and receive the "confirm transaction" information from both parties. After receiving the "transaction reached" signal from the two agent modules, the auction result information including the two parties to the transaction and the transaction amount will be announced.

Images