A method and system for forming a regulatory entity alliance in commodity trading

Abstract

The present invention discloses a method and system for forming an alliance of regulatory entities in bulk commodity transactions. The method and system are used in the context of bulk commodity e-commerce transactions to form specific regulatory tasks for risks existing in transactions, to form an alliance of regulatory entities that can provide stable and efficient regulatory resources for executing regulatory tasks, to form regulatory tasks based on basic information of a single transaction in bulk commodity transactions and the type of regulatory services required, to model regulatory resources and regulatory entities by learning historical risky transaction events; to solve task allocation problems based on resource matching for specific regulatory tasks, to find regulatory entities that can execute the tasks and have the highest regulatory benefits, and all regulatory entities that collaborate to execute regulatory tasks can be regarded as a regulatory alliance, so as to achieve the purpose of supervising risk events in transactions to reduce losses.

Description

Technical Field

[0001] The present invention belongs to the technical field of bulk electronic commodity trading, and in particular, relates to a method and system for forming a regulatory entity alliance in bulk commodity trading. Background Art

[0002] Commodities refer to physical goods that have commodity attributes and can enter the circulation area and be used for industrial and agricultural production and consumption in large quantities. With the advancement of Internet + Industry 4.0 and Internet of Things technology, the scale of e-commerce transactions continues to expand. There are a large number of risk events in the complex and numerous transaction processes, which may cause serious losses to users and industries. For example, using false warehouse receipts to repeatedly pledge to financial institutions to finance loans and achieve commercial arbitrage will increase the financial risks of trading users and financial institutions, and even affect the overall industrial structure in serious cases. In response to this situation, the relevant regulatory agencies can cooperate in supervision to prevent risks and save losses. For different types of risk events that occur in different transactions, the regulatory resources required in the supervision process are also different. The regulatory resources are provided by the regulatory agencies, and the regulatory agencies can provide different regulatory resources according to their respective functional divisions. For example, in the financing risk event in the above example, the Financial Regulatory Bureau needs to conduct administrative law enforcement supervision, and the warehouse party needs to verify the warehouse receipt. The administrative law enforcement capabilities and warehouse receipt verification capabilities here can be regarded as regulatory resources, and the Financial Regulatory Bureau and the warehouse can be regarded as regulatory entities.

[0003] Since the types of risks in commodity trading are complex and changeable, the types of resources required to regulate different transaction risks vary greatly. The single vertical regulatory model in traditional commodity e-commerce transactions can only locate a single type of regulatory subject, with low efficiency in the allocation of regulatory subjects and lack of collaboration between regulatory subjects. Therefore, we hope that the system can implement an alliance formation method for specific regulatory tasks. This method solves the task allocation problem through resource matching, and can find suitable regulatory subjects to form an alliance to collaborate and complete the regulatory task, so as to achieve global benefit optimization. Summary of the invention

[0004] Technical problem: The purpose of the present invention is to propose a method and system for forming an alliance of regulatory entities in commodity trading. The system establishes regulatory tasks for risk events in commodities, uses the alliance formation method to find regulatory entities that can complete the regulatory tasks, and collaborates to build a reliable and efficient regulatory alliance.

[0005] Technical solution: Based on the above technical problems, a regulatory alliance formation system for commodity trading is proposed, including task input module, alliance formation module, and alliance evolution module. The regulatory tasks initiated by system users require a variety of different regulatory resources. The alliance formation module will match resources to form an initial regulatory alliance; at the same time, we hope that the regulatory alliance that provides these resources is stable and efficient, so the alliance evolution module will be optimized in combination with the existing regulatory alliances in the system.

[0006] A method and system for forming an alliance of regulatory entities in commodity trading, the system comprising a task input module, an alliance formation module, and an alliance evolution module. The task input module forms a regulatory task based on the basic information of a single transaction in commodity trading and the type of risk to be regulated. The regulatory task can be regarded as an event of supervising the risk in a commodity transaction, and sets the attribute items of the regulatory task; the alliance formation module is used to execute an alliance formation algorithm. The alliance formation algorithm is executed according to the regulatory task entered by the task input module, and the execution result is an initial regulatory alliance that can complete the regulatory task; the alliance evolution module is used to execute the alliance evolution algorithm. The alliance evolution algorithm is executed according to the output result of the alliance formation module, and the execution result is the optimal alliance in the system.

[0007] A method and system for forming a regulatory entity alliance in commodity trading, the method comprising the following steps:

[0008] (1) First, the task input module is called to form a supervision task based on the basic information of a single transaction in the commodity transaction and the type of risk to be supervised. The supervision task can be regarded as an event of supervising the risk in a transaction, and the supervision task attribute items are set;

[0009] (2) Calling the alliance formation module to execute the alliance formation algorithm based on the user input obtained;

[0010] (3) Calling the alliance evolution module. After the system receives the initial alliance obtained in the alliance formation module, the alliance evolution module provides an alliance evolution method for updating the alliance in the current system.

[0011] As a preferred technical solution of the present invention, in step (2), calling the alliance formation module includes the following steps:

[0012] (2-1) Complete the attributes and establish the supervision task. The system stores the basic transaction information, including the transaction party, transaction amount, etc., which can be retrieved according to the transaction number;

[0013] (2-2) Establish a regulatory subject model. By learning commodity trading knowledge, establish a regulatory subject model. The system stores a list of regulatory subjects that can perform regulatory tasks. Regulatory subjects are departments and institutions that can provide regulatory resources. Different regulatory subjects can provide different regulatory resources.

[0014] (2-3) Matching regulatory tasks with regulatory entities that meet resource needs to form an initial alliance;

[0015] (2-4) The concept of alliance value is used to represent the value of the regulatory alliance. The alliance value depends entirely on the value of the regulatory entities it includes;

[0016] (2-5) performing alliance compression on the formed initial alliance;

[0017] (2-6) performing local optimization on the compressed initial coalition;

[0018] As a preferred technical solution of the present invention, in the step (2-1), the system stores basic transaction information, including the transaction parties, transaction amount, etc. The basic information of the transaction can be retrieved according to the transaction number, and the transaction amount is used as the transaction income. The regulatory task attribute is extended to add the transaction income attribute; the system stores the platform address list of the trading platform, the platform address is the geographical location of the platform (city level), the regulatory task attribute list is extended according to the platform address list, the trading platform address is obtained according to the trading platform name, and the regulatory task attribute is extended to add the platform address attribute;

[0019] As a preferred technical solution of the present invention, in the step (2-2), the system stores a list of regulatory resources corresponding to each type of regulatory service. Regulatory resources can be regarded as the capabilities required to perform regulatory tasks, and different types of regulatory services require different regulatory resources. The regulatory resource requirement list is matched according to the regulatory service type attributes in the regulatory task, and the regulatory task attributes are extended to add regulatory resource attributes.

[0020] As a preferred technical solution of the present invention, in step (2-3), the system stores a list of regulatory entities that can perform regulatory tasks. Regulatory entities are departments or institutions that can provide regulatory resources. Different regulatory entities can provide different regulatory resources. According to the attributes of the regulatory tasks, regulatory entities that meet their resource requirements are matched to form an initial alliance. When matching entities, four constraints must be met: regulatory resource type, regulatory resource quantity, regulatory resource level, and regulatory area scope:

[0021] (1) The types of regulatory resources provided by the regulatory body are the same as the types of regulatory resources required for the regulatory task;

[0022] (2) The quantity of regulatory resources of this type provided by the regulatory body is not less than the quantity of regulatory resources of this type required for the regulatory task;

[0023] (3) The level of the regulatory resources of this type provided by the regulatory body is not lower than the level of the regulatory resources of this type required for the regulatory task;

[0024] (4) The regulatory scope of the regulatory body can cover the address of the trading platform;

[0025] The initial alliance is empty, and all regulatory entities in the system are divided according to the first dimension component of their respective resource vectors. , sort them from large to small, get the sequence of regulatory entities, and find the The minimum value n1 of (the first dimension resource requirement of the supervision task).

[0026] The minimum value n1 specifies the minimum number of supervisory entities required for the initial alliance calculated based on the first dimension resource requirements of the supervisory task. Similarly, the sequence of the minimum number of supervisory entities calculated based on the resource requirements of other dimensions is obtained, and the minimum value n of the sequence is found. k , n k That is, the minimum number of regulatory entities that the initial alliance must include to complete the regulatory task. At this time, the resource demand of the kth dimension is met. The difference between the resource demand quantity and the resource quantity vector of the regulatory entities in the alliance at this time is calculated, and the entity with the largest k-dimensional component among the remaining regulatory entities is added to the current alliance. The above process is repeated until all resources required for the regulatory task are met, and the initial alliance is obtained.

[0027] As a preferred technical solution of the present invention, in the step (2-4), the alliance formation is studied in the characteristic function game, and the concept of alliance value is used to represent the value of the regulatory alliance. The alliance value depends entirely on the value of the regulatory subjects it includes. The alliance value calculation method of the regulatory alliance C is as follows:

[0028]

[0029] P(t)=benefit∈t

[0030]

[0031] The alliance value V(C) is the alliance benefit value of completing supervision task t, P(t) is the reward obtained by completing supervision task t, benefit is the transaction income of the supervision task, is the cost function, Its value decreases monotonically with the increase of the independent variable. S is the sum of all costs spent on forming the alliance. is the normalized value of each influencing factor. Whenever a new alliance is generated, The value of will be recalculated. F(C) is the resource cost paid by the regulatory body in the alliance to perform tasks, E(C) is the cross-regional cost of collaboration among entities in the alliance, and M(C) is the cross-level cost of collaboration among entities in the alliance.

[0032] Resource costs paid by regulatory entities within the alliance to perform tasks:

[0033]

[0034] Where resource is the resource requirement list of the supervision task, and rn is the number of resources required for the supervision task.

[0035] Cross-regional costs of collaboration among entities within the alliance:

[0036]

[0037] Among them A i , A j is the supervisory body i and supervisory body j in the alliance, position i is the geographical location of the regulatory entity i, position j is the geographical location of regulatory entity j.

[0038] Cross-level costs of collaboration among entities within the alliance:

[0039]

[0040] Where M(C) is the cross-level cost of collaboration among entities within the alliance, A i is the regulatory entity i within the alliance, ram is the resource level provided by the regulatory entity, and rm is the resource level required for the regulatory task.

[0041] As a preferred technical solution of the present invention, in the step (2-5), for the existing initial alliance, the regulatory entities therein are sorted from small to large according to the number of resource vectors u that are finally satisfied, and the remaining regulatory entities that have not joined the initial alliance are also sorted from small to large according to the number of resource vectors u that they can provide, and the initial alliance is traversed to check whether the remaining regulatory entities can be used to replace its members; if the constraints can still be met after the replacement, the new alliance formed after the replacement and the updated remaining regulatory entities are still sorted according to the above method until they cannot be replaced, and the compressed initial alliance is obtained.

[0042] As a preferred technical solution of the present invention, in the step (2-6), for each supervisory subject in the compressed alliance, check whether there is a supervisory subject among all the remaining subjects that can replace the supervisory subject. The replacement condition is that the new alliance formed by the replaced subject still meets the resource requirements of the task, and the alliance value of the new alliance is greater than the alliance value of the original alliance. If so, the supervisory subject is replaced to update the supervisory alliance, otherwise the alliance remains unchanged. After performing local optimization, the optimal initial alliance that can complete the current supervision task is obtained.

[0043] As a preferred technical solution of the present invention, in step (3), after the system receives the initial alliance obtained in the alliance formation module, the alliance evolution module provides an alliance evolution method for updating the alliance in the current system. When a new optimal initial alliance arrives, it is compared with the existing regulatory alliance in the system, including the following steps:

[0044] A. If there is an existing regulatory alliance that can cover all alliance entities within the initial alliance, it can be considered that the existing regulatory alliance can perform the regulatory task corresponding to the initial alliance, and this regulatory task is added to the task set of the corresponding existing regulatory alliance, and no new regulatory alliance is added to the system.

[0045] B. If the existing regulatory alliances cannot cover all alliance entities within the initial alliance that has been reached, then the existing regulatory alliances are traversed. It is assumed that new regulatory entities are added to the existing regulatory alliance to cover the reached alliance, and the updated alliance value is calculated. If the updated alliance value is greater than the sum of the alliance values ​​of the original regulatory alliance and the reached alliance, it is considered that the update is beneficial to the improvement of the alliance value within the system. The system updates the existing alliance accordingly and does not add new regulatory alliances. Otherwise, the initial alliance that has been reached is added to the system as a new regulatory alliance.

[0046] As a preferred technical solution of the present invention, the updated alliance value is replaced in step B, wherein the calculation formula is:

[0047] V(C_new)=[P(t)+P(t′)]-F(C_new)-E(C_new)-M(C_new)

[0048] Among them, V(C_new) is the alliance value after replacement and update, P(t′) is the reward obtained for reaching the alliance, F(C_new) is the resource cost paid by the replaced alliance, E(C_new) is the cross-regional cost of collaboration among entities within the updated alliance, and M(C_new) is the cross-level cost of collaboration among entities within the updated alliance.

[0049] Beneficial effects:

[0050] (1) Improve the efficiency of risk supervision in commodity trading. The traditional single vertical supervision model can only locate a single type of regulatory subject, with low efficiency in the allocation of regulatory subjects and lack of collaboration between regulatory subjects. This system abstractly models real problems in commodity trading, finds all regulatory subjects that can meet regulatory needs through resource matching, and forms an alliance to improve the efficiency of supervision of risk events.

[0051] (2) Strengthening the linkage and cooperation among regulatory bodies. The risks of commodity trading are complex and changeable. The supervision of risks requires the joint efforts of multiple regulatory resources. The system uses the alliance formation method to find regulatory bodies that can provide all the necessary regulatory resources to build an alliance. This alliance can achieve global benefit optimization and further strengthen the linkage and cooperation among alliance bodies.

[0052] (3) Dynamic update of regulatory alliances in the system In the alliance formation system in commodity trading, system users will input the tasks to be regulated in real time. The system can use the alliance evolution algorithm to adjust the existing alliance structure according to the newly added regulatory tasks, realize the dynamic allocation of regulatory resources and the dynamic update of the regulatory alliance, and ensure that the regulatory entities within the alliance can achieve long-term and stable cooperation. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is the overall schematic diagram of the system.

[0054] Figure 2 It is the overall logic diagram of this system.

[0055] Figure 3 It is a flow chart of the alliance formation method in this system. DETAILED DESCRIPTION

[0056] The present invention will be further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention.

[0057] A method and system for forming a regulatory alliance in commodity trading. The system is used in commodity e-commerce trading scenarios. Specific regulatory tasks are established for risks in commodity trading, and various types of regulatory resources are matched for each regulatory task. The regulatory resources are provided by regulatory entities. The regulatory entities are departments and institutions that provide regulatory services for risk events in the commodity trading process. The alliance formation method is used to find suitable regulatory entities to collaborate to complete the regulatory tasks. These regulatory entities that have achieved collaboration form a regulatory entity alliance.

[0058] A method and system for forming a regulatory alliance in commodity trading, the method comprising the following steps:

[0059] (1) A method and system for forming a regulatory alliance in commodity trading includes a task input module, an alliance formation module, and an alliance evolution module. The three modules are called in sequence in the system function;

[0060] (2) First, call the task input module to form a supervision task based on the basic information of a single transaction in commodity trading and the type of risk to be regulated. The supervision task can be regarded as an event of supervising the risk in a transaction. Set the supervision task attribute items:

[0061] The transaction number (id) is a required item, and the transaction number is the unique identifier of the transaction to be regulated;

[0062] The transaction commodity type is a mandatory option, and the system user selects the commodity name used for the transaction;

[0063] The trading platform name (flat) is a mandatory option, and the system user selects the name of the commodity e-commerce trading platform for the transaction;

[0064] The regulatory service type (taskType) is a mandatory option. The regulatory service type can be regarded as the regulatory service that needs to be performed for the risks in the commodity transaction. The system sets a list of regulatory service types according to the common risk types in commodity electronic transactions and provides it to system users for selection;

[0065] (3) Calling the alliance formation module and executing the alliance formation algorithm based on the obtained user input. The method steps are as follows:

[0066] ①Complete attributes and establish supervision tasks

[0067] The system stores basic transaction information, including transaction parties, transaction amounts, etc., which can be retrieved based on the transaction number.

[0068] Retrieve the transaction amount according to the transaction number (id), take the transaction amount as the transaction benefit, and expand the supervision task attribute to add the transaction benefit (benefit) attribute

[0069] The system stores the platform address list of the trading platform (flat). The platform address (location) is the geographical location (city level) of the platform. The regulatory task attribute list is expanded according to the platform address list, and the trading platform address is matched according to the trading platform name. The regulatory task attribute is expanded to add the platform address attribute.

[0070] The system stores a list of regulatory resources (resource) corresponding to each regulatory service type (taskType). Regulatory resources can be regarded as the capabilities required to perform regulatory tasks. Different types of regulatory services require different regulatory resources. The regulatory resource list is matched according to the regulatory service type attributes in the regulatory task, and the regulatory task attributes are extended to include regulatory resource attributes.

[0071] The final regulatory task model established is:

[0072] t={resource, location, taskType, benefit}

[0073] Among them, t represents the regulatory task, resource represents the regulatory resources required for the regulatory task, location represents the geographical location of the trading platform, taskType represents the regulatory service type, and benefit represents the regulatory benefit.

[0074] The regulatory resource requirement list (resource) can be expressed as:

[0075] resource={(rn1,rn2,…,rn j ),(rm1,rm2,…,rm j )}

[0076] Among them, rn j represents the number of j-th type of supervision resources required for the supervision task, rm j represents the jth level of supervision resources required for the supervision task;

[0077] ② Establish a regulatory subject model

[0078] By learning about commodity trading knowledge, a regulatory subject model is established. The system stores a list of regulatory subjects that can perform regulatory tasks. Regulatory subjects are departments and institutions that can provide regulatory resources. Different regulatory subjects can provide different regulatory resources. The model is as follows:

[0079] A i ={resourceAgent, position, locationLimits}

[0080] Among them A i represents the regulatory subject i, resourceAgent represents the regulatory resources owned by the regulatory subject, locationAgent represents the location of the regulatory subject, and locationLimits represents the regulatory scope of the regulatory subject;

[0081] The resources owned by the regulatory subject (resourceAgent) can be expressed as:

[0082] resourceAgent={(ran1,ran2,…,ran j ),(ram1,ram2,…,ram j )}

[0083] Among them j Indicates the number of j-th resources owned by the regulatory entity, ram j Indicates the level of the j-th resource owned by the regulatory entity.

[0084] ③ Match the supervision tasks with the supervision subjects that meet the resource requirements to form an initial alliance. When matching the subjects, four constraints must be met: the type of supervision resources, the amount of supervision resources, the level of supervision resources, and the scope of supervision areas:

[0085] Constraint 1: The type of regulatory resources provided by the regulatory subject is the same as the type of regulatory resources required for the regulatory task;

[0086] Constraint 2: The number of regulatory resources of this type owned by the regulatory subject (ran) is not less than the number required for the regulatory task (rn);

[0087] Constraint 3: The level of regulatory resources of this type provided by the regulatory subject (ram) shall not be lower than the level of regulatory resources of this type required for the regulatory task (rm);

[0088] ran i ≥rn i ,ram i ≥rm i

[0089] (ran i ,ram i ∈resourceAgent, rn i ,rm i ∈resource)

[0090] The scope of supervision (locationLimits) of the four regulatory bodies includes the address of the regulatory task trading platform (location);

[0091] location∈locationLimits

[0092] ④ The initial alliance is empty, and all regulatory entities in the system are divided according to the first dimension component of their respective resource vectors , sort them from large to small, get the sequence of regulatory entities, and find the The minimum value n1 of (the first dimension resource requirement of the supervision task).

[0093] The minimum value n1 specifies the minimum number of supervisory entities required for the initial alliance calculated based on the first dimension resource requirements of the supervisory task. Similarly, the sequence of the minimum number of supervisory entities required calculated based on the resource requirements of other dimensions is obtained, and the minimum value n of the sequence is found. k , n k That is, the minimum number of regulatory entities that the initial alliance must include to complete the regulatory task. At this time, the resource demand of the kth dimension is met. The difference between the resource demand quantity and the resource quantity vector of the regulatory entities in the alliance at this time is calculated, and the entity with the largest k-dimensional component among the remaining regulatory entities is added to the current alliance. The above process is repeated until all resources required for the regulatory task are met, and the initial alliance is obtained.

[0094] ⑤ The concept of alliance value is used to represent the value of the regulatory alliance. The alliance value depends entirely on the value of the regulatory entities it includes. The alliance value calculation method of the regulatory alliance C is as follows:

[0095]

[0096] P(t)=benefit∈t i

[0097]

[0098] The alliance value V(C) is the alliance benefit value of completing supervision task t; P(t) is the reward obtained by completing supervision task t, that is, the attribute value of transaction benefit; is the cost function, and its value decreases monotonically with the increase of the independent variable; S is the sum of all costs spent on forming the alliance, is the normalized value of each influencing factor. Whenever a new alliance is generated, The value of will be recalculated, and F(C) is the resource cost paid by the regulatory body in the alliance to perform the task:

[0099]

[0100] E(C) is the cross-regional cost of collaboration among entities within the alliance:

[0101]

[0102] Among them A i , A j is the supervisory body i and supervisory body j in the alliance, position i is the geographical location of the regulatory entity i, position j is the geographical location of regulatory entity j.

[0103] M(C) is the cross-level cost of collaboration among entities within the alliance:

[0104]

[0105] Where M(C) is the cross-level cost of collaboration among entities within the alliance, A i is the regulatory entity i within the alliance, ram is the resource level provided by the regulatory entity, and rm is the resource level required for the regulatory task.

[0106] ⑥Perform alliance compression on the initial alliance formed.

[0107] The regulatory entities are sorted in ascending order according to the number of resource vectors j that are finally satisfied. The remaining regulatory entities that have not joined the initial alliance are also sorted in ascending order according to the number of resource vectors j that they can provide. The initial alliance is traversed to check whether the remaining regulatory entities can be used to replace its members. If the constraints can still be met after the replacement, the new alliance formed after the replacement and the updated remaining regulatory entities are still sorted according to the above method until they cannot be replaced, and the compressed initial alliance is obtained.

[0108] ⑦Perform local optimization on the compressed initial alliance.

[0109] For each supervisory subject in the compressed alliance, check whether there is a supervisory subject among all the remaining subjects that can replace the supervisory subject. The replacement condition is that the new alliance formed by the replaced subject still meets the resource requirements of the task and the alliance value of the new alliance is greater than the alliance value of the original alliance. If so, replace the supervisory subject and update the supervisory alliance, otherwise the alliance remains unchanged. After optimization, the optimal initial alliance that can complete the current supervision task is obtained.

[0110] (4) Calling the alliance evolution module

[0111] After the system receives the initial alliance obtained in the alliance formation module, the alliance evolution module provides an alliance evolution method for updating the alliance in the current system. The steps are as follows: When the new optimal initial alliance arrives, it is compared with the existing regulatory alliance in the system:

[0112] A. If there is an existing regulatory alliance that can cover all alliance entities in the initial alliance, it can be considered that the existing regulatory alliance can perform the regulatory task corresponding to the initial alliance, and this regulatory task is added to the task set of the corresponding existing regulatory alliance, and no new regulatory alliance is added to the system;

[0113] B. If the existing regulatory alliances cannot cover all alliance entities in the initial alliance, then traverse the existing regulatory alliances, assume that new regulatory entities are added to the existing regulatory alliance to cover the alliance, and calculate the updated alliance value, where the calculation formula is as follows:

[0114] V(C_new)=[P(t)+P(t′)]-F(C_new)-E(C_new)-M(C_new)

[0115] Where V(C_new) is the alliance value after replacement and update, P(t′) is the reward for reaching the alliance, F(C_new) is the resource cost paid by the alliance after replacement, E(C_new) is the cross-regional cost of the main body collaboration within the updated alliance, and M(C_new) is the cross-level cost of the main body collaboration within the updated alliance;

[0116] If the updated alliance value is greater than the sum of the alliance values ​​of the original regulatory alliance and the reached alliance, it is considered that the update is beneficial to the improvement of the alliance value within the system. The system modifies the existing alliance according to this update and does not add a new regulatory alliance; otherwise, the initial alliance reached is added to the system as a new regulatory alliance.

Claims

1. A method for forming a regulatory entity alliance in commodity trading, characterized by: Based on a regulatory subject alliance system in commodity trading, the system includes a task input module, an alliance formation module, and an alliance evolution module. The task input module forms a regulatory task according to the basic information of a single transaction in the commodity trading and the type of risk to be regulated. The regulatory task is regarded as an event of performing supervision on the risk in a commodity transaction, and the regulatory task attribute items are set; the alliance formation module is used to execute the alliance formation algorithm; the alliance formation algorithm is executed according to the regulatory task entered by the task input module, and the execution result is the initial regulatory alliance that completes the regulatory task; the alliance evolution module is used to execute the alliance evolution algorithm; the alliance evolution algorithm is executed according to the output result of the alliance formation module, and the execution result is the optimal alliance in the system; The method comprises the following steps: (1) First, the task input module is called to form a supervision task based on the basic information of a single transaction in the commodity trading and the type of risk to be supervised. The supervision task is regarded as an event of supervising the risk in a transaction, and the supervision task attribute items are set; (2) calling the alliance formation module and executing the alliance formation algorithm according to the obtained user input; in step (2), calling the alliance formation module includes the following steps: (2-1) Complete the attributes and establish the supervision task. The system stores the basic transaction information, including the transaction party, transaction amount, and retrieval based on the transaction number; (2-2) Establish a regulatory subject model. By learning commodity trading knowledge, a regulatory subject model is established. The system stores a list of regulatory subjects that can perform regulatory tasks. Regulatory subjects are departments and institutions that can provide regulatory resources. Different regulatory subjects provide different regulatory resources. (2-3) Matching the supervision task with the supervision subject that meets the resource requirements to form an initial alliance; in the step (2-3), the system stores a list of supervision subjects that can perform the supervision task, and matches the supervision subject that meets its resource requirements according to the attributes of the supervision task to form an initial alliance; when matching the subjects, four constraints must be met: the type of supervision resource, the amount of supervision resources, the level of supervision resources, and the scope of supervision area: (2-3-1) The type of regulatory resources provided by the regulatory body is the same as the type of regulatory resources required for the regulatory task; (2-3-2) The number of types of regulatory resources provided by the regulatory body shall not be less than the number of types of regulatory resources required for the regulatory task; (2-3-3) The level of the type of regulatory resources provided by the regulatory body is not lower than the level of the type of regulatory resources required for the regulatory task; (2-3-4) The regulatory scope of the regulatory body can cover the address of the trading platform; The initial alliance is empty, and all regulatory entities in the system are divided according to the first dimension component of their respective resource vectors. 1≤i≤r, sort from large to small, get the sequence of regulatory entities, and find the one that satisfies The minimum value n1 of the first dimension resource requirement of the supervision task; The minimum value n1 specifies the minimum number of regulatory entities required for the initial alliance calculated based on the first dimension resource requirements of the regulatory task; similarly, the sequence of the minimum number of regulatory entities required calculated based on the resource requirements of other dimensions is obtained, and the minimum value n of the sequence is found. k , n k That is, the minimum number of regulatory entities that must be included in the initial alliance to complete the regulatory task. At this time, the k-th dimension resource demand is met. The difference between the resource demand quantity and the resource quantity vector of the regulatory entities in the alliance at this time is calculated, and the entity with the largest k-dimensional component among the remaining regulatory entities is added to the current alliance until all resources required for the regulatory task are met. At this time, the initial alliance is obtained; (2-4) The alliance value is used to represent the value of the regulatory alliance. The alliance value depends entirely on the value of the regulatory subjects it includes. In the step (2-4), the alliance formation is studied in the characteristic function game. The alliance value depends entirely on the value of the regulatory subjects it includes. The alliance value calculation method of the regulatory alliance C is as follows: P(t)=benefit∈t The alliance value V(C) is the alliance benefit value of completing supervision task t, P(t) is the reward obtained by completing supervision task t, benefit is the transaction income of the supervision task, is the cost function, Its value decreases monotonically with the increase of the independent variable. S is the sum of all costs spent on forming the alliance. is the normalized value of each influencing factor. Whenever a new alliance is generated, The value of will be recalculated; F(C) is the resource cost paid by the regulatory body within the alliance to perform tasks, E(C) is the cross-regional cost of collaboration within the alliance, and M(C) is the cross-level cost of collaboration within the alliance; Resource costs paid by regulatory entities within the alliance to perform tasks: Where resource is the resource requirement list of the supervision task, and rn is the number of resources required for the supervision task; Cross-regional costs of collaboration among entities within the alliance: Among them A i , A j is the supervisory body i and supervisory body j in the alliance, position i is the geographical location of the regulatory entity i, position j is the geographical location of the regulatory entity j; Cross-level costs of collaboration among entities within the alliance: Where M(C) is the cross-level cost of collaboration between entities within the alliance, ram is the resource level provided by the regulatory entity, and rm is the resource level required for the regulatory task; (2-5) performing alliance compression on the formed initial alliance; (2-6) performing local optimization on the compressed initial coalition; (3) Calling the alliance evolution module. After the system receives the initial alliance obtained in the alliance formation module, the alliance evolution module provides an alliance evolution method for updating the alliance in the current system. In step (3), after the system receives the initial alliance obtained in the alliance formation module, the alliance evolution module provides an alliance evolution method for updating the alliance in the current system. When a new optimal initial alliance arrives, it is compared with the existing regulatory alliance in the system, including the following steps: A. If there is an existing regulatory alliance that can cover all alliance entities in the initial alliance, it is considered that the existing regulatory alliance can perform the regulatory task corresponding to the initial alliance, and this regulatory task is added to the task set of the corresponding existing regulatory alliance, and no new regulatory alliance is added to the system; B. If the existing regulatory alliances cannot cover all alliance entities in the initial alliance that has been reached, then the existing regulatory alliances are traversed. It is assumed that a new regulatory entity is added to the existing regulatory alliance to cover the reached alliance, and the updated alliance value is calculated. If the updated alliance value is greater than the sum of the alliance values ​​of the original regulatory alliance and the reached alliance, it is considered that the update is conducive to the improvement of the alliance value in the system. The system updates the existing alliance accordingly and does not add a new regulatory alliance. Otherwise, the initial alliance that has been reached is added to the system as a new regulatory alliance; In step B, the updated alliance value is replaced, where the calculation formula is: V(C_new)=[P(t)+P(t′)]-F(C_new)-E(C_new)-M(C_new) Among them, V(C_new) is the alliance value after replacement and update, P(t′) is the reward obtained for reaching the alliance, F(C_new) is the resource cost paid by the replaced alliance, E(C_new) is the cross-regional cost of collaboration among entities within the updated alliance, and M(C_new) is the cross-level cost of collaboration among entities within the updated alliance.

2. The method for forming a regulatory entity alliance in bulk commodity trading according to claim 1, characterized in that: In the step (2-1), the system stores basic transaction information, including the transaction parties and the transaction amount. The basic information of the transaction is retrieved according to the transaction number, and the transaction amount is used as the transaction income. The regulatory task attribute is extended to include the transaction income attribute. The system stores a platform address list of the transaction platform, and the platform address is the geographical location of the platform. The regulatory task attribute list is extended according to the platform address list, and the transaction platform address is obtained according to the transaction platform name. The regulatory task attribute is extended to include the platform address attribute.

3. The method for forming a regulatory entity alliance in bulk commodity trading according to claim 1, characterized in that: In the step (2-2), the system stores a list of regulatory resources corresponding to each type of regulatory service. Regulatory resources are regarded as the capabilities required to perform regulatory tasks, and different types of regulatory services require different regulatory resources. The regulatory resource requirement list is matched according to the regulatory service type attributes in the regulatory task, and the regulatory task attributes are extended to include regulatory resource attributes.

4. The method for forming a regulatory entity alliance in bulk commodity trading according to claim 1, characterized in that: In the step (2-5), for the existing initial alliance, the regulatory entities therein are sorted from small to large according to the number of resource vectors u that are finally satisfied, and the remaining regulatory entities that have not joined the initial alliance are also sorted from small to large according to the number of resource vectors u they provide, and the initial alliance is traversed to check whether the remaining regulatory entities can be used to replace its members; If the constraints are still met after the replacement, the new alliance formed after the replacement and the remaining updated regulatory entities will still be sorted according to the above method until they are irreplaceable, and the compressed initial alliance is obtained.

5. The method for forming a regulatory entity alliance in bulk commodity trading according to claim 1, characterized in that: In the step (2-6), for each regulatory subject in the compressed alliance, check whether there is a replacement for the regulatory subject among all the remaining subjects, and the replacement condition is that the new alliance formed by the replaced subject still meets the resource requirements of the task, and the alliance value of the new alliance is greater than the alliance value of the original alliance; if so, the regulatory subject is replaced to update the regulatory alliance, otherwise the alliance remains unchanged; after performing local optimization, the optimal initial alliance that can complete the current regulatory task is obtained.

Images