ChatGPT-Based housing performance simulation system for analyzing family and social bonding behaviors

Abstract

The present invention relates to a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors, and more specifically, to a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors that enables the implementation of simulations that automatically generate and control various family and social behaviors of home users through the generation of ChatGPT-based human object behavior schedules, thereby supporting the evaluation of the performance of housing design proposals according to the living space by reflecting various characteristics of home users, including social interactions and family bonding behaviors, during the design phase and by calculating the physical building form and semantic attributes.

Description

Technology Field

[0001] The present invention relates to a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors, and more specifically, to a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors that enables the implementation of simulations that automatically generate and control various family and social behaviors of home users through the generation of ChatGPT-based human object behavior schedules, thereby supporting the evaluation of the performance of housing design proposals according to the living space by reflecting various characteristics of home users, including social interactions and family bonding behaviors, during the design phase and by calculating the physical building form and semantic attributes. Background Technology

[0003] Generally, housing design is a process of efficiently and functionally organizing living spaces, and can include all activities involved in planning and designing the residential environment.

[0004] Traditionally, housing design prioritized simply optimizing spatial or energy efficiency; however, recently, there is a need for complex considerations that include not only the universality of residential forms—such as family composition, occupation, social role, and age—but also diversity and specificity, encompassing residents' lifestyle patterns, cultural backgrounds, and personal preferences.

[0005] However, there are many difficulties in predicting potential user behaviors in a space and analyzing spatial performance regarding human factors during the early design phase, prior to the building's use.

[0006] To address these issues, agent-based human behavior simulations using virtual users have recently been developed. These simulations compute complex physical and social interactions occurring between an independent human-like agent and the physical environment, enabling the analysis and exploration of the overall performance of the design, such as movement patterns and spatial usage density.

[0007] However, such agent-based human behavior simulations not only require a significant amount of time for long-term behavioral data collection and technical implementation, but also have limitations in implementing collective and interactive human behavior, particularly social bonding behaviors.

[0008] Furthermore, since expertise such as relevant programming knowledge is essential for its application in the design of buildings like housing, a simulation system that can be utilized in architectural design practice—specifically, a simulation system capable of simulating social activities and relationships among residents—has not yet been developed. Prior art literature

[0010] 1. Korean Patent Publication No. 10-24344489 (Registered Aug. 16, 2022) 2. Korean Patent Publication No. 10-1835738 (Registered Feb. 28, 2018) The problem to be solved

[0011] The present invention was devised to solve the problems of the prior art as described above. The objective of the present invention is to provide a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors, which enables the automatic generation and control of various family and social behaviors of home users through the creation of ChatGPT-based human object behavior schedules, thereby supporting the evaluation of the performance of housing design plans according to the living space by reflecting various characteristics of home users, including social interactions and family bonding behaviors, during the design phase and by calculating the physical building form and semantic attributes. means of solving the problem

[0013] The present invention for achieving the above objectives is,

[0014] It may include: a data input unit for inputting agent and building data required for a simulation for evaluating a residential building; a data generation unit linked with ChatGPT for generating action schedules of agents that can be used in the simulation from the data input unit; a data integration unit for integrating the action schedules of agents generated by the data generation unit with the building data; and a simulation unit for executing a simulation in which agents perform action schedules in a residential building using the data generated and integrated through the data generation unit and the data integration unit.

[0015] At this time, the data input unit may include a first data input unit that inputs attribute data of agents including one or more of age, gender, social status, occupation, country, household composition, position in family, and personality and transmits it to ChatGPT, and a second data input unit that inputs 3D modeling information of a residential building to be used in the simulation.

[0016] Additionally, the data generation unit may include a first data generation unit that automatically generates behavior schedule data containing behavior types defining behavior attributes of agents and behavior generation data, which is information necessary for agents to perform actions, from attribute data of agents input through a first data input unit linked with ChatGPT, and a second data generation unit that generates metadata information for residential building attributes including architectural elements, furniture elements, room tags, and functions of each element in 3D modeling information of a residential building input through a second data input unit.

[0017] In addition, the simulation unit can be configured to perform pathfinding within a designed residential space by providing visual functions to the agents, determine whether furniture elements are required for the execution of the action schedule and whether furniture elements exist, and sequentially execute the action schedule while moving to unit spaces.

[0018] In addition, the simulation unit can switch the agent to a state capable of social bonding behavior if the agent's behavior schedule corresponds to social behavior involving interaction with other agents.

[0019] In addition, the simulation unit may include a human scheduler for generating or modifying the agent's behavior schedule during the simulation process.

[0020] In addition, it may further include a performance analysis unit that analyzes the performance of a residential building design plan and generates analysis data using the simulation results from the simulation unit mentioned above. Effects of the invention

[0022] According to the present invention, by generating a schedule based on ChatGPT, scenarios for the use of a building can be identified by reflecting characteristics such as the daily behaviors, personal preferences, and family structure of various residents; thus, it has the excellent effect of enabling the examination of universal usability of residential spaces and the derivation of design proposals that reflect the characteristics of various household members.

[0023] Furthermore, according to the present invention, it is possible to support the analysis of design performance in terms of human factors through the visualization of dynamic situations between human objects and the physical environment via simulation, and by utilizing a user interface (UI) linked with ChatGPT, it additionally has the effect of enabling easy use not only by simulation experts but also by general users and in design practice. Brief explanation of the drawing

[0025] FIG. 1 is a diagram exemplarily showing a simulation using a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors according to the present invention. FIG. 2 is a diagram conceptually illustrating the configuration of a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors according to the present invention. FIGS. 3 and FIGS. 4 are diagrams showing the simulation process and system operation sequence using the present invention shown in FIGS. 2. FIG. 5 is a diagram exemplarily illustrating the generation of an agent's action schedule using the data generation unit of the present invention shown in FIG. 2. FIG. 6 is a diagram showing the process of calculating the action schedule of an agent in the simulation unit of the present invention shown in FIG. 2. FIG. 7 is a conceptual diagram showing the interaction between the simulation unit and ChatGPT of the present invention shown in FIG. 2. FIG. 8 is a diagram exemplarily showing the appearance displayed by the camera function during a simulation process using the present invention. FIG. 9 is a diagram exemplarily illustrating a method of operating a camera function during a simulation process using the present invention. FIG. 10 is a diagram exemplarily showing an interface displayed on a monitoring means when executing a simulation using the present invention. Figures 11 (a) to (d) are diagrams illustrating exemplary simulation results that can be verified through the performance analysis unit of the present invention shown in Figure 2. Specific details for implementing the invention

[0026] The embodiments of the present disclosure are illustrative for the purpose of explaining the technical concept of the present disclosure. The scope of rights according to the present disclosure is not limited to the embodiments presented below or the specific description thereof.

[0027] All technical and scientific terms used in this disclosure, unless otherwise defined, have the meaning generally understood by those skilled in the art to which this disclosure pertains. All terms used in this disclosure are selected for the purpose of further clarifying this disclosure and are not selected to limit the scope of the rights under this disclosure.

[0028] Expressions such as “comprising,” “comprising,” “having,” etc. used in this disclosure should be understood as open-ended terms implying the possibility of including other embodiments, unless otherwise stated in the phrase or sentence containing such expressions.

[0029] Unless otherwise stated, singular expressions described in this disclosure may include a plural meaning, and this applies likewise to singular expressions described in the claims.

[0031] Hereinafter, preferred embodiments of a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors according to the present invention will be described in detail with reference to the attached drawings.

[0032] FIG. 1 is a diagram exemplarily illustrating a simulation using a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors according to the present invention; FIG. 2 is a diagram conceptually illustrating the configuration of a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors according to the present invention; FIG. 3 and FIG. 4 are diagrams illustrating the simulation process and system operation sequence using the present invention shown in FIG. 2; FIG. 5 is a diagram exemplarily illustrating the generation of an agent's behavior schedule using the data generation unit of the present invention shown in FIG. 2; FIG. 6 is a diagram illustrating the process of calculating the agent's behavior schedule in the simulation unit of the present invention shown in FIG. 2; FIG. 7 is a diagram conceptually illustrating the interaction between the simulation unit and ChatGPT of the present invention shown in FIG. 2; FIG. 8 is a diagram exemplarily illustrating a display by the camera function during the simulation process using the present invention; FIG. 9 is a diagram exemplarily illustrating a method of operating the camera function during the simulation process using the present invention; FIG. 10 is a diagram exemplarily illustrating an interface displayed on a monitoring means when executing a simulation using the present invention; and FIG. 11 (a) to (d) are diagrams illustrating exemplary simulation results that can be verified through the performance analysis unit of the present invention shown in FIG. 2.

[0034] The present invention relates to a ChatGPT-based residential performance simulation system (100) (hereinafter referred to as the 'simulation system (100)') for analyzing family and social bonding behaviors, which enables the implementation of simulations that automatically generate and control various family and social behaviors of a home user through the generation of a ChatGPT-based human object behavior schedule, thereby supporting the evaluation of the performance of a housing design plan according to the living space by reflecting various characteristics of the home user, including social interaction and family bonding behaviors, during the design stage and by calculating the form of the physical building and semantic attributes. As shown in FIG. 2, the system may largely include a data input unit (110), a data generation unit (120), a data integration unit (130), a simulation unit (140), and a monitoring means (150).

[0035] First, the data input unit (110) is configured to input data necessary for residential performance simulation for evaluating a residential building (20) represented by a house, and a conventional input device such as a keyboard or mouse can be used as a data input means.

[0036] To explain in more detail, the data input unit (110) may include a first data input unit (112) for inputting attribute data of a human object to be simulated and a second data input unit (114) for inputting data related to a residential building (20). First, the first data input unit (112) may be provided in the form of a user interface (hereinafter referred to as 'UI') linked with ChatGPT.

[0037] That is, when a user selects a UI displayed on the monitoring means (150) to be described later, a window for inputting attribute data of virtual human objects (hereinafter referred to as 'agents (10)') to be included in the simulation is displayed on the monitoring means (150), and the user can input attribute data of each agent (10) through the first data input unit (112).

[0038] At this time, the attribute data may include age, gender, social status, occupation, country, household composition, position in the family, personality, etc., and the input attribute data can be transmitted to ChatGPT and used to generate an agent (10) action schedule in the data generation unit (120) to be described later.

[0039] Therefore, the more specific the attribute data input through the first data input unit (112) is, the more specifically the action schedule of the agents (10) to perform the simulation can be generated, so the first data input unit (112) can provide a guideline for the attribute data of the agents (10) that the user must input.

[0040] In addition, the second data input unit (114) is configured for inputting residential building (20) information to be used in the simulation, and a 3D modeling program such as Rhino can be used.

[0041] That is, the second data input unit (114) can load a modeling file of a residential building (20) from an external source or directly perform 3D modeling to generate a modeling file of a residential building (20) to be used in the simulation.

[0042] Next, the data generation unit (120) plays the role of generating data that can be used in a simulation using data input through the data input unit (110), and likewise may include a first data generation unit (122) for generating agent data and a second data generation unit (124) for generating building data.

[0043] First, the first data generation unit (122) is configured to generate a behavior schedule for agents (10) to perform a simulation using attribute data of agents (10) input through the first data input unit (112), and ChatGPT can be used to generate the behavior schedule of the agents (10).

[0044] That is, attribute data input through the first data input unit (112) is input into ChatGPT along with a specific instruction included therein, and ChatGPT can generate an action schedule for agents (10) corresponding to the input attribute data according to the instruction included in the instruction and transmit it to the first data generation unit (122).

[0045] At this time, as shown in FIG. 4, the action schedule data may include a 'behavior type' that defines the behavioral attributes of the agents (10) and 'behavior generation data' which is information necessary for the agents (10) to perform the action.

[0046] The above 'behavior type' can define whether the generated schedule is an independent action performed by the agent (10) alone or a social action involving interaction with another agent (10), and can be generated in a bool format having only two values, True or False.

[0047] Additionally, the above 'behavior generation data' may include a duration, a location for execution including unit space or furniture elements, behavior motion data, and schedule information designating the behavior, which are information necessary for agents (10) to perform the behavior through simulation.

[0048] In the first data generation unit (122) above, the 'behavior generation data' generated by ChatGPT can be generated in a string format using a C# script, and each element can be configured to be separated by a semicolon (;) so as to be linked with other 'behavior types' or 'behavior generation data'.

[0049] The behavior schedule data generated in this way can be configured to be displayed on a monitoring means (150), as shown in FIG. 5, so that the user can check the generated behavior schedule data in advance.

[0050] If the generated behavior schedule data is unsuitable or does not match the user's intention, the attribute data of the agent (10) can be re-entered through the first data input unit (112) to generate new behavior schedule data in the first data generation unit (122).

[0051] Next, the second data generation unit (124) plays the role of generating metadata to be used in the simulation using residential building (20) related data, i.e., modeling data, input through the second data input unit (114), and the commercially available 3D simulation platform Unity 3D can be used for generating the metadata.

[0052] That is, the second data generation unit (124) can generate metadata information for the attributes of a residential building (20), including architectural elements, furniture elements, room tags, and the functions of each element, in the 3D modeling of the residential building (20), so that it can be implemented in the simulation unit (140).

[0053] For example, the second data generation unit (124) may define the meaning and location of unit spaces including living room, master bedroom, bedroom, bathroom, etc., as well as furniture elements included in each unit space and their roles.

[0054] Next, the data integration unit (130) can play a role in enabling the implementation of the actions of the agents (10) in the simulation by integrating the action schedule data of the agents (10) generated by the data generation unit (120) and metadata regarding the attributes of the building (20).

[0055] That is, the data integration unit (130) plays the role of integrating data generated by the data generation unit (120) so that agents (10) can recognize building (20) data during the simulation process, explore residential spaces, and perform independent and social actions, and such data integration can also be implemented by Unity 3D.

[0056] For example, the data integration unit (130) can be configured so that agents (10) can recognize building (20) data and determine what actions to perform on furniture elements within a unit space, and based on the action schedule data of agents (10) generated by the first data generation unit (122), individual action rules for agents (10), that is, individual action rules for acting alone, and social action rules for agents (10) interacting with each other can be set.

[0057] Next, the simulation unit (140) is configured to execute a simulation, that is, a simulation in which agents (10) perform residential life based on integrated building (20) data, using data generated and integrated through the data generation unit (120) and data integration unit (130) described above. As described above, Unity 3D programs and C# scripts may be used for the implementation of this simulation.

[0058] That is, the simulation unit (140) can retrieve the behavior schedule data of the agent (10) and the building (20) attribute data generated by the data generation unit (120), and the data integrated through the data integration unit (130), so that the agents (10) can calculate wayfinding and space usage within the designed residential space and perform the generated behavior schedule.

[0059] For example, the algorithm sequence in which agents (10) calculate an action schedule may include, as shown in FIG. 6, first randomly selecting a unit space, checking whether a furniture element is needed to execute the action schedule, and if no furniture element is needed, moving to that unit space to execute the action schedule.

[0060] Additionally, if a household element is required to execute an action schedule, the process may include checking whether the said household element exists; if the household element exists, moving toward the nearest household element to execute the action schedule, and if the household element does not exist, abandoning the execution of the said action schedule and executing the next action schedule.

[0061] At this time, if the action schedule of the agent (10) corresponds to a social action involving interaction with other agents (10), the simulation unit (140) can convert the agent (10) into a state capable of social bonding action.

[0062] When the behavior of the above agent (10) is converted into social behavior, temporary social behavior may occur even during the process of moving to a unit space or furniture element to perform the behavior schedule.

[0063] For example, an agent (10) that performs a behavior schedule corresponding to a social behavior such as 'eating together' can be set to perform social behaviors for a relatively short period of time, such as waving hands or stopping briefly to talk, when certain conditions are satisfied while moving to a unit space corresponding to a restaurant.

[0064] In addition, when arriving at a restaurant and carrying out the behavior schedule, it can be set to perform actions included in social behaviors, such as eating while conversing.

[0065] The execution process of the simulation described above can all be displayed by a monitoring means (150), and accordingly, the user can observe the actions of the agents (10) that are visualized during the simulation and derive problems or improvement plans for the housing design.

[0066] Meanwhile, the simulation unit (140) may include a human scheduler (142), which is configured to generate or modify the action schedule of the agent (10) during the simulation process and may be displayed in the form of a UI on the monitoring means (150).

[0067] That is, the above human scheduler (142) acts as a first data generation unit (122) that generates an action schedule of an agent (10). During the simulation process, after selecting an agent (10) for whom a new action schedule is to be created or modified, and inputting attribute data of the agent (10) to be input into the human scheduler UI, the input content is transmitted to ChatGPT to generate an action schedule. The generated action schedule is displayed on a monitoring means (150), and after the user checks the generated action schedule data and clicks the send button, the generated action schedule data can be applied to the corresponding agent (10).

[0068] The configuration of such a human scheduler (142) can be implemented by linking with the Unity 3D program through the ChatGPT API, as shown in FIG. 7.

[0069] In addition, the simulation unit (140) can provide visual perception capabilities to each agent (10) conducting the simulation, and can be configured to function as a visual sensor by generating a virtual ray that mimics the agent's (10) field of vision using the ray casting algorithm of the Unity 3D program.

[0070] That is, the simulation unit (140) can create a visual sensor at a location corresponding to the eyes of the agents (10) and set the viewing angle and viewing distance of the visual sensor, wherein the viewing angle can be set in the range of about 60 to 90 degrees and the viewing range can be set in the range of about 5 to 10 m.

[0071] Accordingly, agents (10) with visual perception capabilities can move to a specific unit space or furniture element according to a behavior schedule during the simulation process and can perform social behavior with other agents (10).

[0072] Additionally, the simulation unit (140) may include a camera function that allows the simulation process to be viewed from various angles, and as shown in FIG. 8, the image captured by the camera function may be displayed in real time on a monitoring means (150).

[0073] To explain in more detail, the camera functions may include an isometric camera function that displays a 2D screen at an angle with a fixed tilt, a perspective camera function that displays a sense of perspective, and an orthographic camera function that displays the movement paths or usage density of agents (10), and such camera functions can be controlled by input means such as a keyboard and a mouse, as shown in FIG. 9.

[0074] That is, by selecting any one of the above camera functions or by operating each camera function, control such as moving, rotating, zooming in and out, or adjusting visibility can be performed, and through this, the user can observe in detail the behavior schedules, including the social behaviors of the agents (10), that are performed during the simulation process.

[0075] Meanwhile, the simulation system (100) according to the present invention may further include a database (160) and a performance analysis unit (170). First, the database (160) is configured to store data that is input, generated, and integrated in a data input unit (110), a data generation unit (120), and a data integration unit (130). The simulation unit (140) can retrieve data stored in the database (160) and execute a simulation.

[0076] At this time, the database (160) can store the data input, generated, and integrated from the data input unit (110), data generation unit (120), and data integration unit (130) separately, and as the amount of data accumulated in the database (160) increases, the time required for the simulation can be shortened and the accuracy of the simulation can be improved.

[0077] In addition, simulation analysis results from the performance analysis unit (170), which will be described later, can be additionally stored in the database (160).

[0078] Next, the performance analysis unit (170) is responsible for analyzing the performance of a house, i.e., a building (20) design plan using the simulation results from the simulation unit (140) and generating analysis data. It can generate data analyzing the simulation results using an executable file (exe) extracted from Unity 3D.

[0079] To explain in more detail, the performance analysis unit (170) can check the real-time movement paths and occupancy density of the agents (10) in the building (20) design plan from the results of the execution of the agents' (10) behavior schedules through simulation, and can graph the number of visual interferences between the agents (10) through information obtained from the visual sensor.

[0080] In addition, as shown in (a) to (d) of FIG. 11, the analysis results by the performance analysis unit (170) can be displayed on the monitoring means (150), and through this, the user can check the real-time analyzed residential performance including the efficiency of the independent and social actions of each agent (10) during the simulation process, the length of the movement path when performing each action schedule, and the number of personnel suitable for the building (20) design plan.

[0081] Next, the monitoring means (150) serves to display the preparation process of the simulation by the simulation system (100) according to the present invention, the simulation execution process, and the residential performance analysis results, and a conventional display device such as a monitor may be used.

[0082] The above monitoring means (150) may display a user interface linked with ChatGPT and an interface (144) of a simulation unit (140) used in the simulation process, and the interface (144) of the simulation unit (140) may include, as shown in FIG. 10, simulation start / end buttons (144a, 144b), a visual mode button (144c), a human scheduler UI (144d), a behavior schedule information UI (144e), a performance analysis UI (144f), etc.

[0083] To explain in more detail, the simulation start / end buttons (144a, 144b) are buttons for performing the start and end of a simulation in the simulation unit (140), and the visual mode button (144c) may be configured to allow the user to check the simulation process through the monitoring means (150).

[0084] At this time, the visual mode button (144c) may include a visual sensor mode that allows checking the direction of gaze and area confirmed by the visual sensor assigned to the agents (10) in the simulation, as shown in FIG. 11 (a); a movement path mode that allows checking the movement path of the agents (10) in the simulation process, as shown in FIG. 11 (b); a space usage density mode that allows checking how many agents (10) stayed in a specific unit space in the simulation process, i.e., the space usage density, as shown in FIG. 11 (c); and a social network mode that allows checking cases where social behavior occurred between the agents (10) in the simulation process, as shown in FIG. 11 (d).

[0085] Additionally, as described above, the human scheduler UI (144d) is configured to allow the input of attribute data of the agent (10) for the purpose of creating or modifying the agent (10)'s behavior schedule, the behavior schedule information UI (144e) is configured to allow checking the behavior schedule information that each agent (10) is performing during the simulation process, and the performance analysis UI (144f) is configured to allow checking the simulation contents analyzed through the performance analysis unit (170), and allows checking the real-time analyzed residential performance including quantitative figures and graphs for each residential performance.

[0086] Accordingly, the simulation system (100) according to the present invention as described above has various advantages, such as the ability to verify scenarios for building usage by reflecting characteristics such as the daily behaviors, personal preferences, and family structure of various residents through schedule generation based on ChatGPT, thereby enabling the derivation of a design plan that reflects the characteristics of various household members and the review of universal usability of residential spaces, and the ability to support the analysis of design plan performance in terms of human factors through the visualization of dynamic situations between human objects and the physical environment via simulation, and the ability to easily use the system not only by simulation experts but also by general users and in design practice by utilizing a user interface (UI) linked with ChatGPT.

[0088] The system described above may be implemented as a hardware component, a software component, and / or a combination of a hardware component and a software component. For example, the components described in the embodiments may be implemented using one or more general-purpose or special-purpose computers, such as a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing unit may execute an operating system (OS) and one or more software applications executed on said operating system. Additionally, the processing unit may access, store, manipulate, process, and generate data in response to the execution of the software. The processing unit may include multiple processors or one processor and one controller. Additionally, other processing configurations, such as a parallel processor, are also possible.

[0089] Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure a processing unit to operate as desired or command the processing unit independently or collectively. Software and / or data may be permanently or temporarily embodied in any type of machine, component, physical device, virtual equipment, computer storage medium or device, or transmitted signal wave in order to be interpreted by the processing unit or to provide instructions or data to the processing unit. Software may be distributed over networked computer systems and may be stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

[0090] The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and may be recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. The program instructions recorded on the medium may be those specifically designed and configured for the embodiment, or they may be those known and available to a person with ordinary knowledge in the field of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory.

[0091] Although the embodiments have been described above with reference to limited examples and drawings, those skilled in the art can make various modifications and variations from the description above. For example, suitable results can be achieved even if the described techniques are performed in a different order than described, and / or the components of the described system, structure, device, circuit, etc. are combined or assembled in a form different from described, or replaced or substituted by other components or equivalents. Explanation of the symbols

[0093] 10 : Agent 20 : Building 100: Simulation System 110 : Data input section 112 : 1st data input section 114 : Second data input section 120 : Data generation section 122 : First data generation unit 124 : Second data generation unit 130 : Data Integration Department 140 : Simulation section 142 : Human Scheduler 144 : Interface 144a : Start button 144b : Exit button 144c : Visual mode button 144d : Human Scheduler UI 144e : Action Schedule Information UI 144f : Performance Analysis UI 150 : Monitoring means 160 : Database 170 : Performance Analysis Department

Claims

Claim 1 A ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors, comprising: a data input unit for inputting agent and building data required for a simulation for evaluating a residential building; a data generation unit linked with ChatGPT for generating behavioral schedules of agents that can be used in the simulation from the data input through the data input unit; a data integration unit for integrating the behavioral schedules of agents generated by the data generation unit with the building data; and a simulation unit for executing a simulation in which agents perform behavioral schedules in a residential building using the data generated and integrated through the data generation unit and the data integration unit. Claim 2 In claim 1, the data input unit comprises a first data input unit that inputs attribute data of agents including one or more of age, gender, social status, occupation, country, household composition, position in the family, and personality and transmits it to ChatGPT, and a second data input unit that inputs 3D modeling information of a residential building to be used in the simulation, thereby forming a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors. Claim 3 In claim 2, the data generation unit comprises: a first data generation unit that automatically generates behavior schedule data including behavior types defining behavior attributes of agents and behavior generation data, which is information necessary for agents to perform actions, from attribute data of agents input through a first data input unit linked with ChatGPT; and a second data generation unit that generates metadata information for residential building attributes including architectural elements, furniture elements, room tags, and functions of each element in 3D modeling information of a residential building input through a second data input unit, for a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors. Claim 4 A ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors, characterized in that, in claim 1, the simulation unit grants visual functions to agents to perform wayfinding within a designed residential space, determines whether furniture elements are required and whether furniture elements exist for the execution of a behavior schedule, and sets the agents to sequentially execute a behavior schedule while moving to a unit space. Claim 5 A ChatGPT-based residential performance simulation system for family and social bonding behavior analysis, wherein, in claim 4, the simulation unit switches the agent to a state capable of social bonding behavior when the agent's behavior schedule corresponds to social behavior involving interaction with other agents. Claim 6 In claim 1, the simulation unit comprises a human scheduler for generating or modifying an agent's behavior schedule during the simulation process, a ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors. Claim 7 A ChatGPT-based residential performance simulation system for analyzing family and social bonding behaviors according to claim 1, further comprising a performance analysis unit that analyzes the performance of a residential building design plan using the simulation results from the simulation unit and generates analysis data.

Images