Business operation simulation system based on human-intelligent fusion

Abstract

The invention provides a business operation simulation system based on human-intelligent fusion, and the system is characterized in that the system comprises a human-computer interaction module which receives behavior data of a user; the object number linkage module is used for carrying out linkage on a real object and a real environment through the Internet of Things to obtain object number linkage data; the anthropomorphic agent module comprises an anthropomorphic agent simulating a real user and generates anthropomorphic agent interaction data; the object simulating agent module comprises an object simulating agent for simulating a real implement and a real environment and is used for generating object simulating agent interaction data; the market simulation agent module comprises a market simulation agent of the market corresponding to the simulation business and generates an operation plot and task data; and the business operation game engine module is used for carrying out operation simulation and deduction on the business according to the behavior data, the object number linkage data, the anthropomorphic agent interaction data, the object-imitated agent interaction data and the operation plot and task data. In a word, the method can simulate and deduce the business close to reality.

Description

Technical Field

[0001] The present invention belongs to the field of artificial intelligence, and specifically relates to a business operation simulation system based on human-intelligence fusion. Background Art

[0002] Existing role-playing games have rich storylines and character development, as well as complex equipment and skill systems, which can provide immersive gaming experiences and cultivate players' decision-making and teamwork abilities. At present, electronic games are an effective carrier for integrating and disseminating culture / tourism / sports. At the same time, e-sports has entered the fast lane of development relative to other cultural and sports activities. E-sports has become a sunrise industry, which has been recognized internationally and has also received a lot of policy support in China.

[0003] However, the existing game world is limited to vision or touch, and there are problems such as disconnection between digital and physical space, illogical plots, superficial content, and easy addiction due to ease of use, which seriously hinder the development of video games and the e-sports industry. In addition, most role-playing games are limited to entertainment value and have no practical value in cultivating knowledge, skills and cognitive abilities. Therefore, the mainstream game development direction recognized by the industry should be green, healthy, anti-addiction, and increase game benefits and social value.

[0004] In addition, the existing teaching and training operation simulation sandbox system adopts the concept of learning through doing, teaching through playing and working through practicing. It consolidates theoretical knowledge through simulation training while strengthening the ability to apply knowledge, enhances learning interest through multi-group / multi-role / multi-round operation games and / or cultivates collaboration skills, abilities and qualities by simulating real business scenarios.

[0005] In this type of simulation sandbox system, the decisions involved are all centered around relevant knowledge points, which rarely reflect the comprehensive qualities and skills involved in sudden tasks, and lack stimulation and fun. At the same time, this type of simulation sandbox system is based on classroom teaching, and the cooperating roles and opponents are all classmates. In essence, it is an in-class homework / task, and does not support spontaneous and casual development after class. There are problems such as fixed time and poor flexibility. Not only that, the simulation sandbox system focuses on practical training and light decision-making. The application scenarios are limited to teaching and practical training. It is difficult to adapt to the decision-making deduction content and rules involved in the military, government, and enterprises, and its results often cannot be directly used for decision support.

[0006] The concepts involved in this application are partially involved in some related works. For example, the prior art (Liu Xinyao. Research on simulation technology of flash flood disaster emergency drill based on serious games [D]. Fujian Normal University, 2015) discloses a new method of combining GIS software and game engine to generate simulation scenes with real geographical significance based on digital elevation model data, constructing a three-dimensional geographical scene and flood evolution model, and setting different flood scenarios for simulation. The technical principle of this method is to use the agent-based modeling and simulation method to simulate the behavior of the disaster victim agent, analyze the impact of individual psychological, physiological and environmental factors on the disaster victim evacuation process, establish a disaster victim agent evacuation model considering building information, simulate the evacuation behavior of the disaster victims when flash flood disasters strike, and use role-playing to simulate the roles of decision makers, disaster victims, and rescue personnel. The flash flood disaster emergency drill simulation system based on serious games is constructed using the Unity3D game engine, but the simulation drill system only provides technical solutions for flash flood disaster scenarios and application scenarios. The technology used is traditional agents rather than the large model agents involved in this patent, and does not involve self-deductive game engines.

[0007] Large language models are one of the hottest research areas in the field of computer technology in recent years. They usually use complex neural network structures such as Transformer to achieve powerful language understanding and multi-modal generation capabilities, can capture deep features in the data, and use large amounts of data and computing resources for fine training and optimization to automatically complete various complex tasks. Knowledge-enhanced large language models improve the performance of large models by integrating domain knowledge through technologies such as retrieval enhancement generation (RAG), enabling the model to have a deeper understanding of data semantics and improve the accuracy of reasoning and prediction. It usually involves technologies such as knowledge graphs and entity recognition, and specific application scenarios include intelligent customer service and intelligent recommendations.

[0008] The knowledge emergence capability of large language models and their capabilities in multimodality, embodied intelligence / intelligent bodies / humanoid robots / spatial intelligence continue to evolve, showing more and more characteristics of human-like intelligence and general intelligence. Empowering various scenarios based on human-like intelligence or general intelligence has become a hot area of ​​research and innovation. At present, the application depth and cutting-edge progress of large language models in general intelligence and human-like intelligence are still very limited, mainly limited to the application of basic text and multimedia knowledge generation levels, and there are limitations in the innovation of commercial scenarios with market benefits.

[0009] At present, there are three main application modes of large language model agents. The first is a single-agent system in which the agent provides decision support or companionship to real people. It is the most popular in all walks of life, such as the Rabbit r1 AI voice assistant, MetaAI character chat robot, Humane's wearable product Ai Pin, and the multi-functional chat robot AI.Character; the second is a single-agent system that performs task scheduling, because it has high requirements for the accuracy of the large language model output and is still in the development stage; the third is a multi-agent system formed entirely by agents. Research on this is very active, but it is mainly limited to the existing application modes of traditional agents, or the simple combination of the aforementioned single agents. For the multi-agent system of large model agents, no system involving a mixture of human, object, and system agents has been retrieved. Specifically in the game field, the application of large language model intelligent agents has only been found in the "AI Endgame Challenge" game released by Giant Network on November 22, 2024. It claims to be "the industry's first AI native game play built on a multi-agent large model." In this gameplay, the speeches and decisions of each AI player are generated in real time by the large model, without relying on human preset content. AI players can be guided and infected through language to achieve their own game goals. This scenario mainly utilizes the basic ability of the large model to generate text. The speeches and decisions involved are not the core elements of the game. The fundamental advantages of the intelligent agent in embodiment and human-like intelligence are not seen, and the user interest, market benefits and social value it brings are insufficient and cannot be demonstrated.

[0010] In fact, in the current market at home and abroad, large-model intelligent agents still have acceptance issues. Except as the terminal application software carrier of the generative large language model, even the first type of intelligent agent with a large number of products, among the four intelligent agent applications mentioned above, the first three have encountered market difficulties, and the second and third intelligent agents are generally still in the model driven by the previous generation of artificial intelligence technology. The large language model has not driven fundamental innovation in application and design models. The design concepts and application scenarios involved are full of challenges and have a lot of room for innovation. Summary of the invention

[0011] Based on the development of the prior art, the inventors of the present invention believe that:

[0012] The combination of games and business operations is a hot development area in the current game industry, especially in the field of sports, where game applications are becoming increasingly widespread. Small-scale group competition games can simulate competition scenarios and improve athletes' tactical coordination and adaptability. There are also many cases of games in business operations. For example, through role-playing games with sales scenarios, students can play the roles of salespersons and customers respectively, and exercise sales skills and the ability to deal with customer objections through interaction. Another example is designing virtual first aid scenario games in medical business training to strengthen students' practical skills. At the same time, the application of large language models in games is also a hot development area, but currently it is mostly limited to the dialogue and decision-making single-agent application of large language models.

[0013] However, whether it is the game that enables business operations or the large language model that is applied to games, the common problem and challenge is that the game design is not scientific enough and the degree of fit with actual needs needs to be improved, because business operations require the simulation of real enterprises, real businesses, real people, real objects, real environments, and real markets. However, although large language model agents, role-playing games, and business operation simulations each have different shortcomings and development trends, since large language model agents, as enabling technologies and software entities, can act on business operations and role-playing at the same time, and both business operations and role-playing games involve the collaboration of different roles, there is a logical opportunity for integration and innovation.

[0014] Based on the above judgment, the present invention proposes a business operation serious game system and implementation technology based on human-intelligence fusion, which takes business operation simulation as the goal, combines the design concept of large language model intelligent agent, and proposes a mimetic intelligent agent to simulate business entities such as enterprises, people, objects, environment, market and their nature (personality / humanity / objectivity), so as to present the deviations, errors and uncertainties caused by the nature in the process of environmental evolution, making the game scene more challenging and attractive, and making the game process and results more practical reference value; at the same time, a business operation serious game engine based on mimetic intelligent agent is adopted to dynamically generate and adjust the game plot and tasks, forming a game system composed of real entities and mimetic agents. The human-intelligence fusion model composed of intelligent agents has innovated the application model of large language model agents. In addition, the realization of business operation serious games can be achieved by developing the proposed business operation serious games from scratch, determining the roles, facilities, environment and interaction rules involved in the operation scenarios, and on this basis developing pseudo-market agents, anthropomorphic agents, and pseudo-object agents and obtaining intrinsic knowledge reinforcement data to enhance the large language model on which the agents rely, customizing the object-number linkage mechanism, and finally developing related software; or role-playing games can be used as the base, and game elements can be commercialized based on pseudo-intelligent agents to achieve element alignment and integrated development, thereby strengthening the inheritance of game elements and advantages.

[0015] Specifically, the present invention provides a business operation simulation system based on the integration of human intelligence, including: a human-computer interaction module, which receives user behavior data; an object-number linkage module, which links real objects and real environments through the Internet of Things to obtain object-number linkage data; an anthropomorphic agent module, which includes an anthropomorphic agent that simulates real users, and generates real-time physical and mental state data corresponding to real users as anthropomorphic agent interaction data; a simulacrum agent module, which includes a simulacrum agent that simulates real objects and real environments, and generates real-time state data corresponding to real objects and real environments as simulacrum agent interaction data; a pseudo-market agent module, which includes a pseudo-market agent that simulates the market corresponding to the business, and dynamically generates operation scenarios and task data; a business operation game engine module, which performs operation simulation and deduction on the business according to the behavior data, object-number linkage data, anthropomorphic agent interaction data, simulacrum agent interaction data, and operation scenarios and task data, and generates feedback data for each module.

[0016] The business operation simulation system based on human-intelligence fusion provided by the present invention may also have the following features: wherein, the anthropomorphic agent, the pseudo-physical agent and the pseudo-market agent are all pseudo-agents based on a large language model enhanced by scene and nature knowledge, and are obtained by iteratively optimizing by aligning the element data of the role-playing game with the element data of the business operation.

[0017] In the business operation simulation system based on human-intelligence fusion provided by the present invention, it can also have the following characteristics: wherein, the anthropomorphic agent module includes: an anthropomorphic knowledge enhancement unit, including a large language model and anthropomorphic knowledge enhancement material, which is used to perform knowledge enhancement on the large language model through the anthropomorphic knowledge enhancement material to obtain an anthropomorphic agent; an anthropomorphic interaction unit, which is used to generate anthropomorphic agent interaction data corresponding to the feedback data through the anthropomorphic agent, the anthropomorphic knowledge enhancement material includes role portrait data, player nature data, operation history data, game knowledge data and business knowledge data, the role portrait data includes the professional values, professional qualities, professional business knowledge and job responsibilities of the corresponding role in the business, the player nature data includes the personality, body, knowledge, professional skills, qualities and qualities of the corresponding real person, the operation history data is the historical data of the role in the operation simulation and deduction, the game knowledge data is the relevant instructions for the operation simulation and deduction, and the business knowledge data is the business operation knowledge required for the role in the operation simulation and deduction.

[0018] The business operation simulation system based on human-intelligence fusion provided by the present invention may also have the following features: in a business operation simulation involving multiple roles, the role of a missing user is replaced by an anthropomorphic agent. When replacing the role, the anthropomorphic agent generates anthropomorphic agent interaction data based on the state synchronization and linkage data of the designated user as the behavior data of the role under the control of the designated user.

[0019] In the business operation simulation system based on human intelligence fusion provided by the present invention, it can also have the following characteristics: wherein, the simulacrum intelligent agent module includes: a simulacrum knowledge enhancement unit, including a large language model and simulacrum knowledge enhancement materials, which is used to perform knowledge enhancement on the large language model through the simulacrum knowledge enhancement materials to obtain a simulacrum intelligent agent; a simulacrum interaction unit, which is used to generate simulacrum intelligent agent interaction data corresponding to the feedback data through the simulacrum intelligent agent, and the simulacrum knowledge enhancement materials include instrument configuration data, game knowledge data, operation history data, instrument individual data and instrument portrait data, the instrument configuration data is the deployment and role of the corresponding real instrument and real environment in the operation simulation and deduction, the game knowledge data is the relevant description of the operation simulation and deduction, the operation history data is the historical data of the real instrument and real environment in the operation simulation and deduction, the instrument individual data is the description of the real instrument and the real environment, and the instrument portrait data is the characteristics of the real instrument and the real environment.

[0020] The business operation simulation system based on human-intelligence fusion provided by the present invention may also have the following features: in the operation simulation and deduction, the state of the simulated intelligent body and the corresponding real object are synchronized and linked.

[0021] In the business operation simulation system based on human intelligence integration provided by the present invention, it can also have the following characteristics: wherein, the pseudo-market intelligent agent module includes: a pseudo-market knowledge enhancement unit, including a large language model and pseudo-market knowledge enhancement materials, which is used to perform knowledge enhancement on the large language model through the pseudo-market knowledge enhancement materials to obtain a pseudo-market intelligent agent; a pseudo-market interaction unit, which is used to generate pseudo-market intelligent agent interaction data corresponding to feedback data through the pseudo-market intelligent agent, and the pseudo-market knowledge enhancement materials include domain market portrait data, current market portrait data, market history data, game knowledge data and market knowledge data, the market history data is the historical data of the market in operation simulation and deduction, and the market knowledge data is the existing knowledge of the market.

[0022] The business operation simulation system based on human-intelligence fusion provided by the present invention may also have the following features: in the operation simulation and deduction, the roles corresponding to the user and the anthropomorphic intelligent agent include business roles and evaluation roles, and the evaluation roles conduct performance evaluation, promotion and punishment on the business roles.

[0023] Functions and Effects of the Invention

[0024] According to the business operation simulation system based on human-intelligence integration involved in the present invention, on the one hand, it simulates real enterprises, businesses, people, objects, environment, and market through pseudo-intelligent bodies to realize close to real business activities, and combines with corresponding real objects to form a business operation scenario of human-intelligence integration; on the other hand, it realizes the linkage of digital and physical elements driven by business activities based on the Internet of Things, thereby increasing the authenticity, uncertainty and fun of operations. Therefore, the business operation simulation system based on human-intelligence integration of the present invention can simulate and deduce the business closely to reality.

[0025] Based on the above technical effects, the business operation simulation system based on human-intelligence integration of the present invention can reflect the real scenes, businesses, enterprises, tasks, environments, market orientations, etc. in the display world, and integrate the nature of corresponding entities (such as human nature, material nature, and laws) so as to present the uncertainty and deviation caused by different natures, making the business operation process more real, more attractive, and the results more valuable for reference.

[0026] In terms of practical application, the business operation simulation system based on human-intelligence fusion of the present invention can take role-playing games as the base, and develop new products of a series of serious games such as skill training, operation simulation, combat simulation, and policy simulation in combination with specific scenarios. It can produce serious games for various scenarios such as production and service enterprises, supply chain finance, cultural and tourism hotels and scenic spots, community management, emergency management, and combat simulation. It has huge market potential in the fields of e-sports, training, and decision-making simulation, and provides an effective way for brand building, humanities and history communication, and popularization of cutting-edge technologies. It has green, healthy, and far-reaching economic and social value. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is a schematic diagram of the principle of a business operation simulation system based on human-intelligence fusion in an embodiment of the present invention;

[0028] Figure 2 is a block diagram of a business operation simulation system based on human-intelligence fusion in an embodiment of the present invention;

[0029] Figure 3 is a schematic diagram of an anthropomorphic intelligent agent meta-model in an embodiment of the present invention;

[0030] Figure 4 is a schematic diagram of a pseudo-intelligent agent meta-model in an embodiment of the present invention;

[0031] Figure 5 It is a schematic diagram of the virtual city intelligent body meta-model in an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the following embodiments and the accompanying drawings specifically illustrate the business operation simulation system based on human-intelligence fusion of the present invention.

[0033] This embodiment provides a business operation simulation system based on human intelligence fusion, which is used for business operation simulation.

[0034] Figure 1 It is a schematic diagram of the principles of a business operation simulation system based on human-intelligence fusion in an embodiment of the present invention.

[0035] like Figure 1 As shown in the figure, existing role-playing games have problems such as unrealistic and illogical plots, disconnection between digital and physical space, shallow and addictive content, limited entertainment value, and no practical value for knowledge, skills and cognitive abilities. Existing business operation simulations lack challenge, fun, uncertainty and game-playing. Existing large language models are limited to the application of basic text and multimedia knowledge generation levels, and the effective scenarios they create are limited, and innovation in application scenarios and products is urgently needed.

[0036] In response to the problems and needs in the above three areas, this embodiment uses a business operation simulation system based on human-intelligence fusion, takes business operation simulation as the goal, introduces a mimetic intelligent agent based on scenario and nature knowledge to strengthen the large language model, designs a game engine and game entities based on the mimetic intelligent agent, realizes the alignment and integration of game and business elements, forms a new entity of business operation role-playing game, and realizes mutual empowerment of role-playing games, operation games and large language model agents.

[0037] For the application scenario of the business operation simulation system based on human-intelligence integration, based on the serious game system scenario of human-intelligence integration, on the basis of role-playing games, the form and elements of the game are combined with professional business operation scenarios, and a human-intelligence collaborative player team composed of the proposed pseudo-intelligent agent and real people is adopted. The business operation serious game engine based on the proposed pseudo-intelligent agent is used to dynamically adjust the business scenario. The integration of digital and physical elements involved in real business is reflected based on the Internet of Things, and a human-intelligence integration game scenario is constructed. The large language model intelligent agent organization and application mode are innovated to form a new serious game application scenario that combines game entertainment, business deduction and human-like intelligence characteristics.

[0038] For the participating entities of the business operation simulation system based on the integration of human intelligence, a simulated intelligent agent that integrates nature and role is introduced, and innovative large language model intelligent agent design patterns and methods are introduced to simulate real enterprises, real businesses, real people, real objects, real environments, real markets and their natures, so as to present the deviations, errors and uncertainties caused by nature in the operation process, make the game process and results more practical reference value, meet the needs of business operations and simulation deduction, and make the plot more challenging and attractive, so as to solve the respective application pain points and technical innovation needs of role-playing games, teaching training and large language model intelligent agents.

[0039] For the technical implementation of the business operation simulation system based on human-intelligence fusion, the commercialization and integration development of game elements based on mimetic agents is adopted, with business operation simulation as the goal. On the basis of the role-playing game engine, mimetic agents with software or hardware-software integrated devices as carriers improve the game entity elements, realize the alignment and integration of game and business elements, form a new entity of business operation role-playing game, and realize the mutual empowerment of role-playing games, operational games and large language model agents.

[0040] Facing business scenarios such as e-sports games, operational games, and policy deduction, relying on business operation plots, tasks, and physical-digital linkage mechanisms, professional knowledge, professional skills, and personal accomplishments are integrated into relevant elements to reflect the healthy and green characteristics of the game. At the same time, it integrates tourism, culture, history, laws and regulations, and humanistic spirit, providing an effective way for brand building, humanities and history dissemination, and cutting-edge technology popularization, which has far-reaching economic and social value.

[0041] The business operation serious game system based on human-intelligence integration meets the requirements of e-sports, teaching training, simulation and deduction, and has high practical reference value, as well as core elements such as uncertainty and attraction. Through business customization and parameter setting of business operation serious games, it can support a variety of gameplay. Combined with specific scenarios, new products in the serious game series such as skill training, operation game, combat deduction, and policy deduction can be developed.

[0042] Skill training products: Teaching training operation game system based on mimetic intelligent body: customized for subject majors, supporting independent / in-class / online and offline teaching training; users are positioned in schools, competitions, and science popularization, and used for teaching training, skill competitions, employee training, and science popularization.

[0043] Operational game products: Enterprise operation simulation system / sandbox based on pseudo-intelligent entities: customized for business scenarios, supporting enterprise business simulation and scientific decision-making; users are targeted at various types of enterprises, and used for decision support, internal training, and entrepreneurship training.

[0044] Combat simulation products: Comprehensive combat simulation system / sand table based on simulated intelligent agent: customized for combat scenarios, supporting combat simulation and scientific decision-making; users are positioned in various exercises and training bases for simulation training and virtual exercises.

[0045] Policy simulation products: Comprehensive policy simulation system / sand table based on simulated intelligent body: customized for government affairs scenarios, supporting scientific decision-making, comprehensive policy implementation, and precise policy implementation. Users include government departments at all levels, corporate emergency departments, and are used by policy implementation departments as well as urban operations, emergency drills and other departments.

[0046] Based on this, the business operation simulation system based on human intelligence fusion of the present invention can meet the needs of different objects such as production and service enterprises, supply chain finance, cultural and tourism hotels and scenic spots, community management, combat simulation, emergency management, etc. The following describes the application of the business operation simulation system based on human intelligence fusion in some objects:

[0047] In the operation scenarios of production enterprises, for the management of production, supply, sales, storage, personnel, finance, etc. of agricultural products, food and industrial products, serious game products are selected, combined with business operation scenarios that integrate human intelligence, and reference is made to classic cases of related scenarios to set up departments or roles as well as the tasks, processes, items, environment and collaborative relationships involved in the business, set performance appraisal goals and indicators, and configure a serious game engine for business scheduling to drive the complete enterprise operation scenario.

[0048] In the supply chain finance business scenario, for the core enterprises and related upstream and downstream involved in the supply chain finance of specific industries, serious game products are selected, combined with the business operation scenarios of human-intelligence integration, and departments / roles and tasks, processes, items, environments and collaborative relationships involved in the business are set up with reference to classic cases of related scenarios. Performance appraisal goals and indicators are set, and a serious game engine for business scheduling is configured to drive the complete operation scenario of supply chain finance.

[0049] In the operation scenarios of cultural and tourism scenic spots, serious game products are selected for the publicity, marketing, management and scenic spot services involving specific culture, history and natural landscapes. Combined with the business operation scenarios of human-intelligence integration, departments / roles and business-related tasks, processes, items, environment and collaborative relationships are set up with reference to classic cases of related scenarios. Performance appraisal goals and indicators are set, and a serious game engine for business scheduling is configured to realize the complete operation scenarios of cultural and tourism scenic spots.

[0050] In the community management and operation scenarios, serious game products are selected for the neighborhood committees, property management committees and property management involved in the community management scenarios. Combined with the business operation scenarios of human-intelligence integration, departments / roles and the tasks, processes, items, environments and collaborative relationships involved in the business are set up with reference to classic cases of related scenarios. Performance appraisal goals and indicators are set, and a serious game engine for business scheduling is configured to drive the complete operation scenarios of community management.

[0051] In the emergency management operation scenario, serious game products are selected for the departments, resources, events, laws and regulations involved in emergency management. Combined with the business operation scenarios of human-intelligence integration, departments / roles and the tasks, processes, items, environment and collaborative relationships involved in the business are set up with reference to classic cases of related scenarios. Performance appraisal goals and indicators are set, and a serious game engine for business scheduling is configured to drive application management of the complete operation scenario.

[0052] In modern ancient combat scenarios, for modern war simulations or historical war re-enactments, serious game products are selected, combined with business operation scenarios that integrate human intelligence, and teams / roles, tasks, rules, equipment (cultural relics), environment (battlefield) and collaborative relationships are set up with reference to classic cases of related scenarios. Combat goals and indicators are set, and a serious game engine based on self-interpreting mimetic intelligent agents is configured to achieve war scenes with mixed reality effects.

[0053] Self-interpretation scenes of film and television dramas: Based on the self-interpretation scenes of film and television drama plots based on role-playing, select serious game products, customize film and television drama scenes that integrate human intelligence, set styles, roles, and interpretation rules, add equipment, environment and collaborative relationships, and configure a serious game engine based on self-interpreting mimetic intelligent bodies to achieve self-interpretation scenes of film and television dramas with mixed reality effects.

[0054] The business operation simulation system based on human-intelligence fusion of the present invention can also be applied to the application needs of various other scenarios not mentioned above. In its specific application, the intelligent body can be used as an entity, or the intelligent body can be used as a part of an entity and included in the same simulation system together with various actual entities such as actual people, objects, and enterprises.

[0055] Figure 2 It is a block diagram of a business operation simulation system based on human intelligence integration in an embodiment of the present invention.

[0056] like Figure 2 As shown, the business operation simulation system 100 based on human-intelligence integration includes a human-computer interaction module 11, an object-digital linkage module 12, an anthropomorphic agent module 13, a quasi-physical agent module 14, a quasi-market agent module 15 and a business operation game engine module 16.

[0057] The human-computer interaction module 11 receives the user's behavior data.

[0058] The object-number linkage module 12 links the real objects and the real environment through the Internet of Things to obtain the object-number linkage data.

[0059] The anthropomorphic agent module 13 includes an anthropomorphic agent that simulates a real user and generates real-time physical and mental state data corresponding to the real user as anthropomorphic agent interaction data.

[0060] In the operation simulation and deduction of this embodiment, the roles corresponding to the user and the anthropomorphic intelligent agent include business roles and evaluation roles. The evaluation role performs performance evaluation, promotion and punishment on the business role.

[0061] Figure 3 Schematic diagram of an anthropomorphic intelligent agent meta-model in an embodiment of the present invention.

[0062] like Figure 3 As shown, the anthropomorphic intelligent agent meta-model includes any large language model enhanced with knowledge, and uses the large language model to generate personal real-time physical and mental state data that can represent real players, and interacts with the game engine, i.e., the business operation game engine module 16, through agent interaction. Therefore, in this embodiment, the anthropomorphic intelligent agent meta-model can reflect the characteristics of the represented subject, i.e., the real player, and its ability to interact with the environment. It can replace various role subjects to participate in the game as needed, and make autonomous decisions as a substitute for the represented subject, thereby strengthening the curiosity and attraction brought by the uncertainty of partners and opponents, and realizing subjective personalized interaction.

[0063] In this example, the anthropomorphic agent module 13 is used to implement the above-mentioned anthropomorphic agent meta-model, including an anthropomorphic agent that simulates a real user, and generates real-time physical and mental state data corresponding to the real user as anthropomorphic agent interaction data.

[0064] The anthropomorphic agent module 13 includes an anthropomorphic knowledge enhancement unit 131 and an anthropomorphic interaction unit 132 .

[0065] The anthropomorphic knowledge enhancement unit 131 includes a large language model and anthropomorphic knowledge enhancement materials, and is used to perform knowledge enhancement on the large language model through the anthropomorphic knowledge enhancement materials to obtain an anthropomorphic intelligent agent.

[0066] Among them, the anthropomorphic knowledge enhancement materials include character portrait data, player nature data, operation history data, game knowledge data and business knowledge data. Character portrait data includes the professional values, professional qualities, professional business knowledge and job responsibilities of the corresponding character in the business. Player nature data includes the personality, body, knowledge, professional skills, quality and quality of the corresponding real person. Operation history data is the historical data of the character in the operation simulation and deduction. Game knowledge data is the relevant instructions for operation simulation and deduction. Business knowledge data is the business operation knowledge required for the character in the operation simulation and deduction.

[0067] The anthropomorphic agent of this embodiment is a hardware-software integrated mimetic agent that has a large language model based on scene and nature knowledge reinforcement, and has cognitive ability based on scene and nature knowledge reinforcement of large language model. Furthermore, the anthropomorphic agent also has physical and digital integrated perception and processing capabilities, which are synchronized with the personality and status of the subject it represents.

[0068] The anthropomorphic interaction unit 132 is used to generate anthropomorphic agent interaction data corresponding to the feedback data through the anthropomorphic agent.

[0069] In this embodiment, in a business operation simulation including multiple roles, the role of the missing user is replaced by an anthropomorphic agent. When replacing the role, the anthropomorphic agent generates anthropomorphic agent interaction data according to the state synchronization and linkage data of the designated user as the behavior data of the role under the control of the designated user.

[0070] Figure 4 Schematic diagram of the meta-model of the quasi-intelligent agent in the embodiment of the present invention.

[0071] like Figure 4 As shown, the quasi-intelligent agent meta-model includes an arbitrary large language model that has been enhanced with knowledge, and uses the large language model to generate real-time status data that can represent real objects or real environments, and interacts with the game engine, i.e., the business operation game engine module 16, through agent interaction.

[0072] In this example, the quasi-agent module 14 is used to implement the quasi-agent meta-model, including a quasi-agent that simulates real objects and real environments, and generates real-time status data corresponding to the real objects and real environments as quasi-agent interaction data.

[0073] The simulacrum intelligent agent module 14 includes a simulacrum knowledge enhancement unit 141 and a simulacrum interaction unit 142 .

[0074] The pseudo-knowledge enhancement unit 141 includes a large language model and pseudo-knowledge enhancement materials, and is used to perform knowledge enhancement on the large language model through the pseudo-knowledge enhancement materials to obtain a pseudo-intelligent agent.

[0075] Among them, the virtual object knowledge enhancement materials include artifact configuration data, game knowledge data, operation history data, artifact individual data and artifact portrait data. The artifact configuration data is the deployment and role of the corresponding real artifacts and real environment in the operation simulation and deduction. The game knowledge data is the relevant description of the operation simulation and deduction. The operation history data is the historical data of the real artifacts and real environment in the operation simulation and deduction. The artifact individual data is the description of the real artifacts and real environment. The artifact portrait data is the characteristics of the real artifacts and real environment.

[0076] The skeuomorphic intelligent agent of this embodiment is used to represent various items involved in business operations in business activities under specified circumstances, such as due to lack of equipment and environment, simulation and deduction needs, etc. It can respond to physical, mechanical, digital and ecological signals given like a real object.

[0077] The skeuomorphic agent can realistically simulate the represented object, including its inherent physical, biological, environmental and commodity characteristics, so it does not need to be linked with the represented object. However, in certain cases, in order to increase authenticity and enhance complexity, it can synchronize and link with the real object during the game.

[0078] The skeuomorphic interaction unit 142 is used to generate skeuomorphic agent interaction data corresponding to feedback data through the skeuomorphic agent.

[0079] Figure 5 It is a schematic diagram of the virtual city intelligent body meta-model in an embodiment of the present invention.

[0080] like Figure 5 As shown, the simulated market intelligent agent meta-model includes an arbitrary large language model that has been enhanced with knowledge, and uses the large language model to perceive the market status, dynamically generate plots and tasks, and generate adjustment plans for the real market involved in the current business, and interact with the game engine, i.e., the business operation game engine module 16, through intelligent agent interaction to exchange data.

[0081] In this example, the pseudo-market agent module 15 is used to implement the above-mentioned pseudo-market agent meta-model. The pseudo-market agent module 15 includes a pseudo-market agent of the market corresponding to the simulated business, and dynamically generates operation scenarios and task data.

[0082] The pseudo-market agent module 15 includes a pseudo-market knowledge enhancement unit 151 and a pseudo-market interaction unit 152 .

[0083] The pseudo-market knowledge enhancement unit 151 includes a large language model and pseudo-market knowledge enhancement materials, and is used to perform knowledge enhancement on the large language model through the pseudo-market knowledge enhancement materials to obtain a pseudo-market intelligent agent.

[0084] Among them, the proposed market knowledge enhancement materials include field market portrait data, current market portrait data, market history data, game knowledge data and market knowledge data. Market history data refers to the historical data of the market in operation simulation and deduction. Market knowledge data refers to the existing knowledge of the market.

[0085] The pseudo-market agent of this embodiment can simulate the external resources involved in the business operation process in the real environment, as well as the market dynamics caused by internal and external events that affect business operations. For example, external resources include orders or services, and internal and external events include unexpected events and policy interventions. Therefore, the pseudo-market agent can achieve self-adaptation and self-interpretation of the game based on dynamic plots and tasks, increase uncertainty, and enhance fun.

[0086] The pseudo-market interaction unit 152 is used to generate pseudo-market agent interaction data corresponding to feedback data through the pseudo-market agent.

[0087] The business operation game engine module 16 simulates and deduce the business operation according to the behavior data, object-number linkage data, anthropomorphic agent interaction data, quasi-physical agent interaction data, and operation plot and task data, and generates feedback data for each module.

[0088] In this embodiment, the business operation game engine module 16 is also used to define customizable parameters of the simulation system, such as the roles and functions involved in the plot and tasks, business and processes, participating objects and simulated intelligent entities, objects or environments or enterprises, performance and performance, as well as the professional knowledge, professional skills, personal qualities of the design, and tourism, culture, history, laws and regulations and humanistic elements. Further, the game parameters can be automatically defined by players, administrators or the system according to actual needs.

[0089] Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A business operation simulation system based on human intelligence fusion, characterized in that: include: Human-computer interaction module, receiving user behavior data; The object-data linkage module links real objects and real environments through the Internet of Things to obtain object-data linkage data; An anthropomorphic agent module, comprising an anthropomorphic agent simulating a real user, generating real-time physical and mental state data corresponding to the real user as anthropomorphic agent interaction data; A pseudo-agent module, comprising a pseudo-agent that simulates the real object and the real environment, and generates real-time status data corresponding to the real object and the real environment as pseudo-agent interaction data; The pseudo-market agent module includes a pseudo-market agent that simulates the market corresponding to the business and dynamically generates operation scenarios and task data; The business operation game engine module performs operation simulation and deduction on the business according to the behavior data, the object-number linkage data, the anthropomorphic agent interaction data, the simulacrum agent interaction data and the operation plot and task data, and generates feedback data for each module.

2. The business operation simulation system based on human-intelligence fusion according to claim 1, characterized in that: in, The anthropomorphic agent, the pseudo-object agent and the pseudo-market agent are all pseudo-agents that strengthen the large language model based on scene and nature knowledge, and are obtained by iterative optimization by aligning the element data of role-playing games with the element data of business operations.

3. The business operation simulation system based on human-intelligence fusion according to claim 1, Features: Wherein, the anthropomorphic agent module includes: An anthropomorphic knowledge enhancement unit, comprising a large language model and anthropomorphic knowledge enhancement material, for performing knowledge enhancement on the large language model through the anthropomorphic knowledge enhancement material to obtain the anthropomorphic intelligent agent; an anthropomorphic interaction unit, configured to generate the anthropomorphic agent interaction data corresponding to the feedback data through the anthropomorphic agent, The anthropomorphic knowledge enhancement material includes character portrait data, player nature data, operation history data, game knowledge data and business knowledge data. The role portrait data includes the professional values, professional qualities, professional business knowledge and job responsibilities of the role corresponding to the business. The player's nature data includes the character, physique, knowledge, professional skills, quality and accomplishment of the corresponding real person. The operation history data is the historical data of the role in the operation simulation and deduction. The game knowledge data is the relevant description of the operation simulation and deduction. The business knowledge data is the business operation knowledge required by the role in the operation simulation and deduction.

4. The business operation simulation system based on human-intelligence fusion according to claim 1, characterized in that: in, In a business operation simulation involving multiple roles, the role in which the user is missing is replaced by the anthropomorphic agent. When replacing a role, the anthropomorphic agent generates the anthropomorphic agent interaction data according to the state synchronization and linkage data of the designated user as the behavior data of the role under the control of the designated user.

5. The business operation simulation system based on human-intelligence fusion according to claim 1, Features: Wherein, the quasi-physical agent module includes: A pseudo-knowledge enhancement unit, comprising a large language model and pseudo-knowledge enhancement material, for performing knowledge enhancement on the large language model through the pseudo-knowledge enhancement material to obtain the pseudo-intelligent agent; a simulacrum interaction unit, configured to generate the simulacrum agent interaction data corresponding to the feedback data through the simulacrum agent, The virtual object knowledge enhancement material includes object configuration data, game knowledge data, operation history data, object individual data and object portrait data. The artifact configuration data is the deployment and role of the corresponding real artifact and the real environment in the operation simulation and deduction, The game knowledge data is the relevant description of the operation simulation and deduction. The operation history data is the history data of the real objects and the real environment in the operation simulation and deduction. The object individual data is a description of the real object and the real environment. The object image data is the characteristics of the real object and the real environment.

6. The business operation simulation system based on human-intelligence fusion according to claim 1, characterized in that: in, In the operation simulation and deduction, the virtual intelligent entity and the corresponding real object are synchronized and linked in status.

7. The business operation simulation system based on human-intelligence fusion according to claim 1, Features: Wherein, the simulated market agent module includes: A pseudo-market knowledge enhancement unit, comprising a large language model and pseudo-market knowledge enhancement material, is used to perform knowledge enhancement on the large language model through the pseudo-market knowledge enhancement material to obtain the pseudo-market intelligent agent; a pseudo-market interaction unit, configured to generate the pseudo-market agent interaction data corresponding to the feedback data through the pseudo-market agent, The proposed market knowledge enhancement materials include domain market portrait data, current market portrait data, market history data, game knowledge data and market knowledge data. The market historical data is the historical data of the market in the operation simulation and deduction. The market knowledge data is the existing knowledge of the market.

8. The business operation simulation system based on human-intelligence fusion according to claim 1 is characterized by: in, In the operation simulation and deduction, the roles corresponding to the user and the anthropomorphic agent include business roles and evaluation roles. The evaluation role performs performance evaluation, promotion and punishment on the business role.

Images