Building adaptive system based on swarm intelligence

Abstract

The present invention provides a building adaptation system based on swarm intelligence, including: an environmental decision-making agent, which is used to analyze and fuse environmental parameters, user characteristics and behavior patterns, and system equipment operation status data to obtain fused data; a decision model is constructed using a machine learning algorithm to automatically adjust the environmental control strategy according to different input parameters; a third environmental control strategy is formulated according to the results output by the decision model; and the third environmental control strategy is executed by the environmental control agent. The present invention optimizes energy use and reduces unnecessary energy consumption through intelligent adjustment equipment; automatically adjusts environmental parameters according to real-time environmental data and user needs to provide a comfortable living environment; through the collaborative work of intelligent agents, comprehensive monitoring and management of the building environment is achieved to improve the operating efficiency and stability of the system; and potential system problems and faults are timely identified through real-time monitoring and machine learning algorithms.

Description

Technical Field

[0001] The invention relates to a building adaptation system based on swarm intelligence. Background Art

[0002] Existing building environmental control systems usually rely on preset rules, which is difficult to adapt to the complex and changing needs and high comfort requirements of users in public buildings. Especially in public buildings such as libraries and museums, the personalized needs of users and the real-time nature of environmental control have become key challenges for system optimization. Summary of the invention

[0003] The object of the present invention is to provide a building adaptation system based on swarm intelligence.

[0004] To solve the above problems, the present invention provides a building adaptation system based on swarm intelligence, comprising:

[0005] Environmental control agent, used to collect environmental parameters;

[0006] Image recognition agents, used to collect user characteristics and behavior patterns;

[0007] System operation and maintenance agent, used to collect system equipment operation status data;

[0008] The environmental decision-making agent DA is used to analyze and fuse the environmental parameters obtained from the environmental control agent, the user characteristics and behavior patterns obtained from the image recognition agent, and the system equipment operation status data obtained from the system operation and maintenance agent to obtain fused data; based on the fused data, a decision model is constructed using a machine learning algorithm to automatically adjust the environmental control strategy according to different input parameters to achieve optimal user comfort and energy efficiency; a third environmental control strategy is formulated according to the results output by the decision model; and the third environmental control strategy is executed through the environmental control agent.

[0009] Furthermore, in the above method, the environmental decision-making agent is also used to collect system operation data and user feedback through the system operation and maintenance agent after executing the third environmental control strategy, and evaluate the effectiveness of the current environmental control strategy; dynamically adjust the decision model and the third environmental control strategy according to the evaluation results; during the execution of the third environmental strategy, if conflicts between multiple agents arise, resolve the conflicts based on user roles and authority levels, as well as a multi-objective optimization algorithm.

[0010] Furthermore, in the above method, the environmental control agent is used to obtain the data feature vector F of the environmental data; based on the data feature vector F, a spatial object model of the environmental control agent is constructed; based on the spatial object model, the environmental control agent CA comprehensively evaluates the current status of the building environment and formulates a first environmental control strategy; based on the first environmental control strategy, the control behavior interface of the spatial object model is called to output a corresponding control instruction sequence to achieve adjustment of the building environment.

[0011] Furthermore, in the above method, the environmental control agent is used to deploy a variety of sensors in various spaces of the building, including sensors for temperature, humidity, light, air quality, sound and pedestrian flow; ensure that the sensors collect environmental data in real time and transmit the environmental data to the data processing center through the MQTT protocol; the data processing center uses Kafka as a message queue system to receive data streams from sensors; uses Flink to perform real-time data processing on the data streams of the sensors, performs data cleaning operations, and obtains cleaned environmental data; structures the cleaned environmental data and converts them into a unified format to obtain structured environmental data for subsequent analysis and processing; uses Flink's window analysis function to group and aggregate the structured environmental data according to time windows to extract preset features; extracts data feature vectors F from the preset features, including feature vectors of temperature, humidity, light intensity, air quality index, sound intensity and pedestrian flow; uses Redis as a cache database to store real-time data and data feature vectors F; and provides an API interface for the environmental control agent to access structured environmental data and data feature vectors F.

[0012] Furthermore, in the above method, the environmental control agent is used to extract various static information and internal component information of the space from the BIM model and the data collected on site; based on the various static information and internal component information of the space, summarize and construct the attributes of the space object model, including the area, floor height, orientation and internal equipment information of the space; use the data feature vector F to update the state of the space object model in real time; define the ability of the space object model to control various equipment therein, that is, the control behavior interface of the action, including: the control behavior interface of the equipment such as air conditioners, lights, curtains and fresh air fans; set the action of the environmental control agent to call the space object model, including calling the control behavior interface to execute the corresponding control instructions.

[0013] Furthermore, in the above method, the environmental control intelligent agent is used to output a control instruction sequence corresponding to the first environmental control strategy by calling the control behavior interface of the spatial object model; based on the control instruction sequence, fine-tuning of the building environment is achieved, including: adjusting the air-conditioning temperature, light brightness, curtain opening and closing, and fresh air fan operation.

[0014] Furthermore, in the above method, the image recognition agent is used to deploy cameras in key areas of the building to ensure that all important areas are covered; configure the camera to connect with the image recognition agent, collect video data in real time, and transmit it to the data processing center through the network; perform image enhancement processing on the collected video data to obtain enhanced video data; perform denoising processing on the enhanced video data to remove noise in the video to obtain denoised video data; use the image recognition large model to perform target detection on the denoised video data, and extract user characteristics, behavior patterns and body language; use machine learning algorithms to identify users' activity areas and behavior habits, and establish user behavior models by analyzing users' walking paths, residence times and activity frequency data; through the user behavior model, provide data support for environmental adaptation, so that the environmental control strategy is dynamically adjusted according to the user behavior pattern to obtain analysis results; transmit the user behavior analysis results to the environmental decision-making agent.

[0015] Furthermore, in the above method, the system operation and maintenance agent is used to monitor the operating status of the building environment equipment system in real time, including: the working status of the equipment, the agent operation log, energy consumption, server resource usage and command response delay time; use machine learning algorithms and preset rules to identify potential problems and faults of equipment in the operating status of the building environment system; when abnormal operation or failure of the equipment is detected, the equipment parameters are automatically adjusted to ensure the stable operation of the system; and the adjusted equipment parameters are sent to the environmental decision-making agent.

[0016] Furthermore, in the above method, the environmental decision-making agent is used to define user roles at different levels; each user role is assigned corresponding attributes; the allocation of operating permissions is reviewed regularly, and operating permissions that are no longer needed are revoked in a timely manner; user requirements for environmental parameters are collected in real time through an interactive interface, and the semantic understanding ability of a general large model is used to analyze user requirements and organize them into structured demand parameters, and user characteristics and behavior analysis data provided by an image recognition agent and environmental control status data provided by a control agent are simultaneously collected, and the collected data is analyzed and preprocessed to obtain the current environmental status evaluation result; the environmental decision-making agent DA formulates an environmental control policy based on the current environmental status evaluation result and using a Pareto optimization algorithm.

[0017] Furthermore, in the above method, the environment control agent, image recognition agent, system operation and maintenance agent, and environment decision agent are used to realize information sharing and synchronization through a shared database or message queue to ensure that all agents can obtain the latest environmental data and control instructions; realize real-time data transmission between agents through a high-speed network to ensure the timeliness and accuracy of information; use encryption technology in the data transmission process to ensure data security; in the communication process, if an abnormal situation occurs, the system automatically detects and takes corresponding recovery measures; formulate a unified communication protocol to ensure that the agents can exchange information smoothly. The communication protocol includes data format, transmission method, and error detection and correction mechanism.

[0018] The present invention aims to solve the problem of complex and changeable user needs and high comfort requirements in public buildings such as libraries and museums. The system allocates an independently operated intelligent agent to each space, receives user needs, and dynamically adjusts equipment operating parameters and control strategies, thereby achieving refined control of the building environment and improving the response speed of intelligent operation and maintenance and user experience.

[0019] Compared with the prior art, the present invention aims to solve the following technical problems:

[0020] 1. Improve energy efficiency: optimize energy use and reduce unnecessary energy consumption through intelligent adjustment equipment.

[0021] 2. Improve living comfort: Automatically adjust environmental parameters based on real-time environmental data and user needs to provide a comfortable living environment.

[0022] 3. Realize intelligent management: Through the collaborative work of intelligent bodies, comprehensive monitoring and management of the building environment can be achieved, improving the operating efficiency and stability of the system.

[0023] 4. Fault warning and maintenance: Through real-time monitoring and machine learning algorithms, potential system problems and faults can be identified in a timely manner to ensure the stable operation of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 Schematic diagram of a building adaptation system based on swarm intelligence according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0026] In a typical configuration of the present application, the terminal, the device of the service network and the trusted party all include one or more processors (CPU), input / output interfaces, network interfaces and memories.

[0027] Memory may include non-permanent storage in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0028] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include non-transitory media such as modulated data signals and carrier waves.

[0029] like Figure 1 As shown in the figure, the building adaptation system based on swarm intelligence includes:

[0030] Step 1: Establish a building environment control agent (CA), obtain a data feature vector F of environmental data through the environment control agent; construct a spatial object model of the environment control agent based on the data feature vector F; based on the spatial object model, the environment control agent CA comprehensively evaluates the current status of the building environment and formulates a first environment control strategy; based on the first environment control strategy, call the control behavior interface of the spatial object model, output a corresponding control instruction sequence, and realize the adjustment of the building environment;

[0031] The system will set up an environmental control agent CA for each space in the building. The environmental control agent CA is responsible for collecting and processing environmental data in the area and evaluating the environmental status, formulating and implementing control strategies, automatically adjusting equipment, optimizing energy use and living comfort, and is the executor in the building intelligent debugging system.

[0032] Step 1.1, set up the data processing center of the environmental control agent to perform data preprocessing of the environmental data; the data preprocessing process includes: data cleaning, data structuring, data window analysis and feature extraction to obtain the data feature vector F; the data processing center provides a real-time data interface for the environmental control agent CA to access the environmental data and data feature vector F of sensors (such as temperature, humidity, light, air quality, sound, human flow, etc.).

[0033] Step 1.1.1, deploy a variety of sensors in various spaces of the building, including temperature, humidity, light, air quality, sound, and pedestrian flow sensors.

[0034] Step 1.1.2: Ensure that the sensor can collect environmental data in real time and transmit the environmental data to the data processing center through the MQTT protocol.

[0035] Step 1.1.3, the data processing center uses Kafka as a message queue system to receive data streams from sensors;

[0036] Step 1.1.4, use Flink to perform real-time data processing on the data stream of the sensor, perform data cleaning operations, and obtain cleaned environmental data; preferably, the data cleaning operations include: removing noise data, filling missing values, and removing duplicate data.

[0037] Step 1.1.5, structure the cleaned environmental data and convert it into a unified format to obtain structured environmental data for subsequent analysis and processing;

[0038] Step 1.1.6: Use Flink's window analysis function to group and aggregate structured environment data according to time windows and extract preset features.

[0039] Step 1.1.7, extracting a data feature vector F from the preset features, including feature vectors of temperature, humidity, light intensity, air quality index, sound intensity and pedestrian flow;

[0040] Step 1.1.8, use Redis as a cache database to store real-time data and data feature vector F;

[0041] Step 1.1.9: Provide an API interface for the environment control agent CA to access the structured environment data and data feature vector F.

[0042] Step 1.2: The environmental control agent uses data fusion technology to build a unified spatial object model of the environmental control agent. The spatial object model consists of three parts: attributes, states and actions. The attributes are summarized and constructed from the BIM model and the data collected on site, including various static information of the space and internal component information; the state is updated in real time based on the data feature vector F extracted in step 1.1; the action is the ability of the spatial object model to control various devices within the spatial object model; the spatial object model can provide the agent with spatial attributes, real-time status and equipment control interface.

[0043] Step 1.2.1: The environmental control agent extracts various static information and internal component information of the space from the BIM model and the data collected on site;

[0044] Step 1.2.2: The environment control agent summarizes and constructs the attributes of the space model based on various static information and internal component information of the space, including the area, floor height, orientation and internal equipment information of the space;

[0045] Step 1.2.3: The environment control agent uses the data feature vector F extracted in step 1.1 to update the state of the space object model in real time;

[0046] Preferably, the status part includes real-time data of temperature, humidity, light intensity, air quality index, sound intensity and human flow.

[0047] Step 1.2.4: The environment control agent defines the ability of the space object model to control various devices within it, that is, the control behavior interface of the action, including: the control behavior interface of the air conditioner, light, curtain and fresh air fan;

[0048] Step 1.2.5: Set the action of the environment control agent CA to call the space object model, including calling the control behavior interface to execute the corresponding control instructions.

[0049] Step 1.3: Based on the spatial object model, the environmental control agent CA comprehensively evaluates the current status of the building environment and formulates a first environmental control strategy: combining the attributes provided by the spatial object model established in step 1.2, the real-time data in the state and the environmental parameters, comprehensively evaluate the current status of the building environment. This evaluation process incorporates multi-dimensional information of historical data trends, user preference settings and external environmental conditions to form a first environmental control strategy; then, based on the first environmental control strategy, the control behavior interface of the spatial object model is called to output the corresponding control instruction sequence to achieve refined adjustment of the building environment.

[0050] Step 1.3.1: The environment control agent CA outputs a control instruction sequence corresponding to the first environment control strategy by calling the control behavior interface of the space object model;

[0051] Step 1.3.2: The environmental control agent CA realizes fine-tuning of the building environment based on the control instruction sequence, including: adjusting the air-conditioning temperature, light brightness, curtain opening and closing, and fresh air fan operation.

[0052] Step 2: Establish an image recognition agent (IRA). The image recognition agent IRA analyzes the image data in the real-time video stream, identifies the user's characteristics and behavior patterns, and obtains the recognition results. The image recognition agent IRA transmits its recognition results directly to the environmental decision-making agent DA, which integrates this information to optimize the environmental control strategy.

[0053] Step 2.1: Deploy cameras and image recognition agents in key areas of the building to collect video data in real time.

[0054] Step 2.1.1: Deploy high-resolution cameras in key areas of the building (such as entrances, corridors, public areas, etc.) to ensure that all important areas are covered;

[0055] Step 2.1.2: Configure the camera to connect to the image recognition agent (IRA), collect video data in real time, and transmit it to the data processing center through the network.

[0056] Step 2.2: The image recognition agent (IRA) preprocesses the collected video data, including image enhancement, denoising and object detection; and performs object detection on the preprocessed images to extract user features, behavior patterns and body language.

[0057] Step 2.2.1: Perform image enhancement processing on the collected video data to obtain enhanced video data and improve image quality;

[0058] Step 2.2.2: De-noising the enhanced video data to remove the noise in the video to obtain denoised video data and improve recognition accuracy;

[0059] Step 2.2.3: Use the large image recognition model to perform target detection on the denoised video data, extract user characteristics (such as gender, age, body shape, spatial position), behavior patterns (such as dwell time, hot and cold reactions) and body language (hot and cold reactions).

[0060] Step 2.3: The image recognition agent (IRA) uses a multimodal large model to analyze user characteristics (age, gender, body shape), behavior patterns (activity area, behavior habits), and body language (hot and cold reactions) to obtain analysis results and provide data support for environmental adaptation;

[0061] Step 2.3.1: The image recognition agent (IRA) uses machine learning algorithms (such as cluster analysis and time series analysis) to identify the user's activity area and behavior habits, and establishes a user behavior model by analyzing the user's walking path, dwell time, and activity frequency data;

[0062] Step 2.3.3: The image recognition agent (IRA) provides data support for environmental adaptation through the user behavior model, so that the environmental control strategy is dynamically adjusted according to the user behavior pattern to obtain the analysis result;

[0063] Step 2.4: The image recognition agent (IRA) transmits the analysis results to the environmental decision-making agent DA, which integrates the analysis results to optimize the environmental control strategy.

[0064] Step 2.4.1: Transmit the user behavior analysis results to the environmental decision-making agent DA;

[0065] Step 2.4.2: The environmental decision agent DA integrates the user behavior information and environmental data in the analysis results to optimize the environmental control strategy and obtain the third environmental control strategy;

[0066] Step 2.4.3: The environmental decision agent DA dynamically adjusts the operating parameters of equipment such as air conditioners, lights, and curtains in the building according to the user behavior patterns and environmental conditions in the third environmental control strategy, and sends them to the environmental control agent CA to improve energy efficiency and user comfort.

[0067] Step 3: Establish a System Operations and Maintenance Agent (SOMA). The System Operations and Maintenance Agent SOMA continuously monitors the operating status of the building environment system, uses preset rules and machine learning algorithms to identify potential problems and failures of equipment in the building environment system, and automatically adjusts equipment parameters when abnormal equipment operation or failure is detected, and sends the adjusted equipment parameters to the environmental decision-making agent.

[0068] Step 3.1: The system operation and maintenance agent SOMA monitors the operating status of the building environment equipment system in real time, including the working status of the equipment, agent operation log, energy consumption, server resource usage and command response delay;

[0069] Step 3.2: The system operation and maintenance agent uses machine learning algorithms and preset rules to identify potential equipment problems and failures in the operating status of the building environment system;

[0070] Step 3.2.1: The system operation and maintenance agent trains a machine learning model to identify abnormal patterns and fault characteristics in the operation of equipment in the building environment system;

[0071] Step 3.2.2: The system operation and maintenance agent defines preset rules to monitor the operating parameters of the equipment in the operating state of the building environment system and identify situations where the equipment exceeds the normal range;

[0072] Step 3.2.3: The system operation and maintenance agent analyzes the operating parameters of the equipment in the operating state of the building environment system in real time based on abnormal modes, fault characteristics and situations where the equipment exceeds the normal range in equipment operation, and combines machine learning algorithms and preset rules to identify potential problems and faults of the equipment.

[0073] Step 3.3: After the system operation and maintenance agent identifies potential problems and failures of the equipment, it quickly takes stabilization measures, such as adjusting equipment parameters or initiating maintenance requests, and reports abnormal situations to the system administrator to ensure the stability and robustness of the system.

[0074] Step 3.3.1: When the system operation and maintenance agent detects abnormal operation or failure of the equipment, SOMA automatically adjusts the equipment parameters, such as adjusting the air conditioning temperature, light brightness, curtain opening and closing, etc., to ensure the stable operation of the system; and sends the adjusted equipment parameters to the environmental decision-making agent.

[0075] Step 3.3.2: If adjusting the equipment parameters cannot solve the problem, the system operation and maintenance intelligent agent SOMA automatically initiates a maintenance request and notifies the maintenance personnel to handle it.

[0076] Step 3.3.3: The system operation and maintenance agent reports abnormal situations and handling measures to the system administrator to ensure that the administrator can understand the system's operating status and problems in a timely manner.

[0077] Step 3.4: The system operation and maintenance agent analyzes the system operation data and generates regular reports to provide to system administrators and users.

[0078] Step 3.4.1: The system operation and maintenance agent stores and analyzes the building environment system operation data for a long time, identifies the operation trend and energy consumption of the equipment, and obtains the analysis results;

[0079] Step 3.4.2 The system operation and maintenance agent generates regular reports based on the building environment system operation data, including the operation status, energy consumption, fault records, etc. of the equipment, and provides them to system administrators and users;

[0080] Step 3.4.3: Based on the analysis results, the system operation and maintenance agent makes optimization suggestions to help system administrators and users optimize the operating parameters of the equipment and improve energy efficiency and system stability.

[0081] Step 4: Establish the Building Environment Decision Agent (DA). The DA plays a key role in coordinating users, environmental control agents CA, image recognition agents IRA and system operation and maintenance agents SMOA in the building environment to ensure the efficiency of the environmental control system and user satisfaction. Its main responsibilities include collecting and analyzing user needs, guiding decision logic, monitoring system status and maintaining system stability.

[0082] Step 4.1: The environmental decision agent develops the user role and permission allocation backend page, and runs the system administrator and building manager to assign permissions and roles to building users:

[0083] Step 4.1.1: The environment decision agent defines different levels of user roles, such as administrator, ordinary user, visitor, etc.

[0084] Step 4.1.2: The environmental decision agent assigns corresponding attributes to each user role, including access level, operation authority and control scope;

[0085] Step 4.1.3: The environmental decision-making agent regularly reviews the allocation of operation permissions and promptly revokes operation permissions that are no longer needed;

[0086] Step 4.2: The environmental decision-making agent collects user demands for environmental parameters in real time through the interactive interface, uses the semantic understanding ability of the general large model to analyze user demands and organize them into structured demand parameters, and simultaneously collects user characteristics and behavior analysis data provided by the image recognition agent and environmental control status data provided by the control agent CA, and analyzes and preprocesses the collected data to understand and make decisions accordingly.

[0087] Here, the environmental decision agent formulates appropriate environmental control strategies based on the input data. These environmental control strategies should comprehensively consider user comfort, equipment operation efficiency and energy consumption. The decision logic of the environmental decision agent DA includes but is not limited to the following aspects:

[0088] Permission management: Each user has a corresponding role and permission in the system, which determines the access rights to device resources and control policies.

[0089] Conflict resolution: When multiple user demands conflict, DA can intelligently weigh the demands of all parties and develop the best solution.

[0090] Multi-objective optimization: Use multi-objective optimization algorithms to find the optimal balance between user needs, energy consumption and efficiency.

[0091] Adaptive learning A continuously learns user behaviors and preferences and dynamically adjusts its decision logic to improve user satisfaction.

[0092] Step 4.3: Build the coordination capability of the environmental decision-making agent DA system: The environmental decision-making agent DA acts as a coordinator between different agents, ensuring that the agents work together, avoid conflicts, optimize resource allocation, and improve system efficiency and performance.

[0093] Step 4.3.1: Real-time environmental data monitoring: monitor and collect real-time environmental data, such as temperature, humidity, light, air quality; equipment status data, such as operating status and energy consumption, etc. The real-time data and environmental parameters in the attributes and status provided by the spatial object model established in step 1.2 can be combined to comprehensively evaluate the current status of the building environment;

[0094] Step 4.3.2: Historical environmental data analysis: Analyze historical environmental data and equipment operation data to identify historical trends and patterns; consider the impact of external environmental factors such as weather and seasonal changes on environmental conditions.

[0095] Step 4.3.3: User characteristics and preferences integration: Combine user characteristic data, such as age, gender, activity habits, clothing insulation efficiency, human metabolic rate, etc. Integrate user preference settings, such as temperature preference, light preference, etc.

[0096] Step 4.3.4 Comprehensive assessment of environmental status: Comprehensively consider real-time data, historical trends, external environmental factors, user characteristics and preferences, evaluate the current environmental status, and obtain the current environmental status assessment result.

[0097] Step 4.4: The environmental decision-making agent DA formulates an environmental control strategy based on the current environmental status evaluation results and uses the Pareto optimization algorithm. The goal is to meet the user's comfort needs while requiring the optimal solution for equipment operation efficiency and energy consumption.

[0098] The environmental decision-making agent DA is responsible for analyzing the system operation data and generating regular reports to provide to system administrators and users. Specifically, it includes:

[0099] Application effect evaluation: Evaluate the operating effect of the system, including environmental comfort, energy consumption, equipment operation status, etc., in order to identify potential problems and room for improvement.

[0100] Trend analysis: By analyzing historical data, we can identify trends in environmental changes and user needs, and provide a basis for future system optimization.

[0101] User feedback: Collect user feedback on system operation and adjust control strategies and optimize system performance accordingly.

[0102] Step 4.4.1: The environmental decision agent DA calculates the user comfort PMV index. The PMV index is calculated based on the previous environmental status assessment results in step 4.3;

[0103] Step 4.4.2: The environmental decision-making agent DA uses the Pareto optimization algorithm to calculate the optimal solution of user comfort and energy consumption as the third environmental control strategy; the constraints for calculating the optimal solution of user comfort and energy consumption are user demand, energy consumption segment billing, equipment operation effective power and upper and lower limits.

[0104] Assume that the energy consumption model is a function E(T,H,V,T a ), the user comfort model is PMV(T,H,V,M,W,P,T a， T r， F cl， T cl ), then the objective function is:

[0105] min[E(T,H,V,T a ),PMV(T,H,V,M,W,P, T a， T r， F cl， T cl )]

[0106] T : room temperature (℃);

[0107] H: indoor humidity (%);

[0108] V : wind speed (m / s);

[0109] M : metabolic rate (W / m²);

[0110] W : external power (W / m²);

[0111] P: water vapor partial pressure (Pa);

[0112] T a : air temperature (℃);

[0113] T r : mean radiant temperature (℃);

[0114] F cl : The ratio of clothing surface to skin;

[0115] T cl : Clothing insulation efficiency.

[0116] In order to achieve the optimal solution for energy consumption resources while satisfying user comfort, iterative calculations are required to randomly generate a set of environmental parameters (temperature, humidity, wind speed); calculate the PMV value and energy consumption value of each set of environmental parameters; select the optimal solution based on the PMV value and energy consumption value, and generate a new combination of environmental parameters; repeat the evaluation and selection steps until convergence or the preset number of iterations is reached.

[0117] Step 4.4.3: The environmental decision-making agent DA develops a conflict resolution mechanism to deal with the contradictions between multiple user requirements. This can be modeled through a multi-objective optimization problem; weights are assigned to each objective function in combination with the user's authority level, and the multi-objective problem is converted into a single-objective problem; then a Pareto optimal solution set is found, which cannot be dominated by any other solution (that is, better or equal in all objectives); when a solution does not exist or is not effective, the environmental decision-making agent DA can combine the user portrait and make reasonable suggestions, such as adding or removing clothes, changing the spatial position, etc.

[0118] Step 4.5: The environmental decision agent DA is responsible for analyzing the system operation data and generating regular reports to provide to system administrators and users. The specific execution steps are as follows:

[0119] Step 4.5.1: The environmental decision agent DA evaluates the system's operating performance, including environmental comfort, energy consumption, etc., in order to identify potential problems and room for improvement. The specific steps are as follows:

[0120] Collect system operation data, including environmental parameters, equipment status, and energy consumption;

[0121] Use statistical analysis and machine learning techniques to analyze system operation data and evaluate system operation results;

[0122] Generate performance evaluation reports, identify system strengths and weaknesses, and make recommendations for improvements.

[0123] Step 4.5.2: By analyzing historical data, identify trends in environmental changes and user needs, and provide a basis for future system optimization. The specific steps are as follows:

[0124] Collect historical data, including environmental parameters, user requirements, and system operation data;

[0125] Use data mining and time series analysis techniques to analyze historical data and identify trends in environmental changes and user needs;

[0126] Generate trend analysis reports to provide a basis for future system optimization;

[0127] Step 4.5.3 Collect user feedback on system operation and adjust environmental control strategies and optimize system performance accordingly. The specific steps are as follows:

[0128] Collect user feedback on system operation through interactive interfaces and questionnaires;

[0129] Analyze user feedback data to identify user needs and problems in system operation;

[0130] Adjust control strategies and optimize system performance based on user feedback to improve user satisfaction.

[0131] Step 5: Establish a swarm intelligence communication mechanism

[0132] In this step, a real-time, secure, stable and scalable communication mechanism is established to ensure information exchange and collaborative work among various intelligent agents.

[0133] Step 5.1: Each agent shares and synchronizes information through a shared database or message queue to ensure that all agents can obtain the latest environmental data and control instructions.

[0134] Each agent shares and synchronizes information by connecting to a shared database or message queue system. The shared database can use a distributed database system such as Apache Cassandra or MongoDB to support high availability and consistency of large-scale data. The message queue system can be selected such as Apache Kafka or RabbitMQ, which can handle high-throughput message delivery and guarantee the order and reliability of messages. Agents interact with these systems through standard APIs or SDKs to read and write data.

[0135] Step 5.2: Realize real-time data transmission between intelligent entities through high-speed networks to ensure the timeliness and accuracy of information; use encryption technology during data transmission to ensure data security;

[0136] Real-time data transmission between intelligent entities is achieved by using high-speed, low-latency wireless local area networks (WLAN). Data transmission can use the TCP / IP protocol stack and use encryption protocols such as SSL / TLS to encrypt data to ensure the security of data during transmission. If necessary, VPN technology can also be used to establish a secure communication tunnel.

[0137] Step 5.3 During the communication process, if any abnormal situation occurs (such as network interruption, data loss, etc.), the system can automatically detect and take corresponding recovery measures to ensure the stability and reliability of communication.

[0138] Integrate monitoring and fault detection mechanisms in the communication system, such as heartbeat detection, timeout retransmission, etc., to automatically detect abnormal situations such as network interruption or data loss. Once an abnormality is detected, the system will trigger a recovery process, such as trying to reestablish a connection, retransmitting lost data packets, or switching to an alternate network path. These recovery measures can be implemented based on software-defined networking (SDN) technology to quickly adapt to network changes.

[0139] Step 5.4: Develop a unified communication protocol to ensure smooth information exchange between agents. The communication protocol includes data format, transmission method, error detection and correction mechanism, etc.

[0140] Design and implement a unified communication protocol that defines the data format, transmission method, and error detection and correction mechanism for information exchange between agents. The data format can use JSON, XML, or Protobuf to support data structuring and serialization. The transmission method can be based on TCP or UDP protocol, and reliable transmission or low-latency transmission can be selected according to application requirements. Error detection and correction mechanisms can include checksums, redundancy checks, etc. to ensure data integrity and reliability. The design of the protocol should take into account future scalability to add new functions or support new agent types.

[0141] Step 6: Establish a swarm intelligent decision-making mechanism to formulate the optimal environmental control strategy by integrating the analysis results of each intelligent agent and user needs.

[0142] A close cooperation mechanism is established among the self-checking of the environment control agent CA, the environment decision agent DA, the image recognition agent IRA and the system operation and maintenance agent SOMA to realize the intelligent adjustment of the building environment; the environment control agent CA is responsible for executing specific environment control operations, the environment decision agent DA is responsible for formulating and optimizing environment control strategies, the image recognition agent IRA provides user characteristics and behavior data support, and the system operation and maintenance agent SOMA ensures the stable operation of the system. This cooperation model can ensure the efficient operation of the building environment control system while improving the comfort and satisfaction of users.

[0143] Step 6.1: The environmental decision-making agent DA collects the data transmitted by each agent, including user needs, equipment status, user identification tags, etc., performs data fusion and analysis, and extracts key features and trends.

[0144] Data fusion and analysis. First, the environmental decision agent DA analyzes and fuses the data collected from the environmental control agent CA, the image recognition agent IRA, and the system operation and maintenance agent SOMA to obtain fused data. These data include environmental parameters (such as temperature, humidity, and light), user characteristics (such as age, gender, and location), behavior patterns (such as activity areas and behavior habits), and system equipment operating status (such as equipment working status and energy consumption).

[0145] Step 6.2: The environmental decision agent DA builds a decision model based on the extracted key features and trends; the decision model uses a machine learning algorithm and can automatically adjust the control strategy according to different input parameters.

[0146] Decision model construction: Based on the fused data, the environmental decision agent DA uses machine learning algorithms to build a decision model to automatically adjust the environmental control strategy according to different input parameters to achieve the best user comfort and energy efficiency. The factors considered by the model include user needs, equipment status, external environmental conditions, etc.

[0147] Step 6.3: Introduce a feedback mechanism to adjust and optimize the decision model based on the actual execution results.

[0148] Control strategy formulation and execution: The environmental decision-making agent DA formulates detailed third-environment control strategies based on the results output by the decision model; these strategies include air-conditioning temperature adjustment, lighting brightness control, curtain opening and closing, etc.; after the environmental decision-making agent DA formulates the strategy, the environmental decision-making agent DA executes these third-environment control strategies through the environmental control agent CA, the image recognition agent IRA provides user behavior data support, and the system operation and maintenance agent SOMA monitors the system status during the execution of the environmental control strategy.

[0149] Step 6.4: Transmit the optimized environmental control strategy to each agent to perform corresponding control operations.

[0150] Feedback mechanism and strategy optimization: After executing the third environmental control strategy, the environmental decision agent DA collects system operation data and user feedback through the system operation and maintenance agent SOMA to evaluate the effect of the current environmental control strategy; based on the evaluation results, the environmental decision agent DA dynamically adjusts the decision model and the third environmental control strategy to improve user satisfaction and system efficiency. For example, if user feedback indicates that the temperature in a certain area is uncomfortable, the environmental decision agent DA will adjust the temperature setting of that area.

[0151] Step 6.5: Collaboration and conflict resolution: During the execution of the third environmental strategy, if conflicts arise between multiple agents, such as conflicts in temperature preferences among different users, the environmental decision-making agent DA will resolve these conflicts based on user roles and authority levels, as well as multi-objective optimization algorithms, to ensure the consistency and efficiency of the overall operation of the system.

[0152] Step 6.6 Regular reporting and system optimization. The environmental decision agent DA regularly generates system operation reports, including key indicators such as environmental comfort, energy consumption, and equipment status. These reports are provided for reference by system administrators and users to guide system optimization and improvement.

[0153] Step 7: Group intelligence supervision and optimization implementation plan

[0154] Step 7.1: To ensure continuous improvement and efficient operation of the building environment control system, each agent (CA, DA, IRA, and SOMA) is equipped with learning and evolution mechanisms. These mechanisms allow the agents to self-optimize based on experience, user feedback, and new data patterns.

[0155] Step 7.1.1: The environmental control agent CA learns to optimize its control algorithm by analyzing environmental data and the effects of control strategies. It can predict environmental changes based on historical data and adaptively adjust control actions. For example, if the temperature of a certain area always needs to be adjusted in the evening, CA will automatically learn this pattern and make corresponding temperature adjustments in advance.

[0156] Step 7.1.2: The environmental decision agent DA continuously optimizes its decision model by analyzing the long-term trends of user behavior and environmental data. It automatically learns the user's preferences and habits and adjusts the environmental control strategy based on this information. For example, if the user often adjusts the indoor temperature during a certain period of time, DA will learn this pattern and adjust the default temperature setting for that period of time.

[0157] Step 7.1.3: The image recognition agent IRA improves its user recognition and behavior analysis algorithms by continuously analyzing image data in the video stream. It is able to identify new user characteristics and behavior patterns, thereby providing a more accurate user profile. For example, IRA can learn to recognize users of different age groups and genders and adjust the environment control strategy based on this information.

[0158] Step 7.1.4: SOMA, the system operation and maintenance agent, improves its anomaly detection and fault prediction capabilities by monitoring the system's operating status and learning system behavior. It can predict potential problems and take preventive measures in advance based on the system's historical performance and current status. For example, if a device often fails under certain conditions, SOMA will learn this pattern and perform maintenance in advance when similar conditions occur.

[0159] Step 7.2 Each agent regularly evaluates its performance, including key indicators such as response speed, decision accuracy, and user satisfaction. These evaluation results are used to generate performance reports and provide them to system administrators and users. Based on the performance evaluation results, the agent makes optimization suggestions to help improve energy efficiency and system stability.

[0160] Step 7.3 To ensure that the system is always in the best condition, each agent participates in the upgrade and maintenance of the system. This includes updating algorithm models, upgrading software versions, improving control strategies, etc. At the same time, the agent pays attention to emerging technologies and methods, and constantly explores how to integrate new technologies into building environment control systems to improve the intelligence level of the system.

[0161] (1) Personalized control: Realize refined control of each spatial environment to meet the personalized needs of users.

[0162] (2) Improving user satisfaction: Improving user comfort and satisfaction by dynamically adjusting control strategies and quickly responding to user needs.

[0163] (3) Efficient operation and maintenance: Use intelligent agents to adjust and optimize equipment operating parameters in real time, improve operation and maintenance efficiency, and reduce energy consumption.

[0164] (4) Technological innovation: Promote the application of large models and intelligent agent technologies in the field of building environmental control and improve the level of building intelligence.

[0165] In summary, the present invention aims to solve the problem of complex and changeable user needs and high comfort requirements in public buildings such as libraries and museums. The system allocates an independently operated intelligent agent to each space, receives user needs, and dynamically adjusts equipment operating parameters and control strategies, thereby achieving refined control of the building environment and improving the response speed of intelligent operation and maintenance and user experience.

[0166] The present invention aims to solve the following technical problems:

[0167] 1. Improve energy efficiency: optimize energy use and reduce unnecessary energy consumption through intelligent adjustment equipment.

[0168] 2. Improve living comfort: Automatically adjust environmental parameters based on real-time environmental data and user needs to provide a comfortable living environment.

[0169] 3. Realize intelligent management: Through the collaborative work of intelligent bodies, comprehensive monitoring and management of the building environment can be achieved, improving the operating efficiency and stability of the system.

[0170] 4. Fault warning and maintenance: Through real-time monitoring and machine learning algorithms, potential system problems and faults can be identified in a timely manner to ensure the stable operation of the system.

[0171] The detailed contents of the device embodiments of the present invention can be found in the corresponding parts of the method embodiments, which will not be described in detail here.

[0172] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

[0173] It should be noted that the present invention can be implemented in software and / or a combination of software and hardware, for example, it can be implemented using an application-specific integrated circuit (ASIC), a general-purpose computer or any other similar hardware device. In one embodiment, the software program of the present invention can be executed by a processor to implement the steps or functions described above. Similarly, the software program of the present invention (including related data structures) can be stored in a computer-readable recording medium, for example, a RAM memory, a magnetic or optical drive or a floppy disk and the like. In addition, some steps or functions of the present invention can be implemented using hardware, for example, as a circuit that cooperates with a processor to perform various steps or functions.

[0174] In addition, a part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. The program instruction for calling the method of the present invention may be stored in a fixed or removable recording medium, and / or transmitted through a data stream in a broadcast or other signal-bearing medium, and / or stored in a working memory of a computer device that runs according to the program instruction. Here, according to an embodiment of the present invention, a device is included, the device including a memory for storing computer program instructions and a processor for executing program instructions, wherein, when the computer program instruction is executed by the processor, the device is triggered to run the method and / or technical solution based on the aforementioned multiple embodiments of the present invention.

[0175] It is obvious to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved. In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the device claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

Claims

1. A building adaptation system based on swarm intelligence, characterized in that: include: Environmental control agent, used to collect environmental parameters; Image recognition agents, used to collect user characteristics and behavior patterns; System operation and maintenance agent, used to collect system equipment operation status data; The environment decision agent is used to analyze and fuse the environment parameters obtained from the environment control agent, the user characteristics and behavior patterns obtained from the image recognition agent, and the system equipment operation status data obtained from the system operation and maintenance agent to obtain fused data; Based on the fused data, a decision model is constructed using a machine learning algorithm to automatically adjust the environmental control strategy according to different input parameters to achieve optimal user comfort and energy efficiency; a third environmental control strategy is formulated according to the results output by the decision model; and the third environmental control strategy is executed by the environmental control agent; the environmental decision agent is also used to collect system operation data and user feedback through the system operation and maintenance agent after executing the third environmental control strategy, and evaluate the effect of the current environmental control strategy; according to the evaluation effect, the decision model and the third environmental control strategy are dynamically adjusted; during the execution of the third environmental strategy, if conflicts between multiple agents occur, the conflicts are resolved according to user roles and authority levels, as well as a multi-objective optimization algorithm; the environmental control agent is used to obtain a data feature vector F of the environmental data; Based on the data feature vector F, a spatial object model of the environment control agent is constructed; based on the spatial object model, the environment control agent CA comprehensively evaluates the current status of the building environment and formulates a first environment control strategy; based on the first environment control strategy, the control behavior interface of the spatial object model is called to output a corresponding control instruction sequence to achieve adjustment of the building environment; The environmental control agent is used to deploy a variety of sensors in various spaces of the building, including sensors for temperature, humidity, light, air quality, sound, and human flow; ensure that the sensors collect environmental data in real time, and transmit the environmental data to the data processing center through the MQTT protocol; the data processing center uses Kafka as a message queue system to receive data streams from sensors; uses Flink to perform real-time data processing on the data streams of the sensors, performs data cleaning operations, and obtains cleaned environmental data; performs structured processing on the cleaned environmental data and converts it into a unified format to obtain structured environmental data for subsequent analysis and processing; uses Flink's window analysis function to group and aggregate the structured environmental data according to time windows to extract preset features; extracts data feature vectors F from the preset features, including feature vectors of temperature, humidity, light intensity, air quality index, sound intensity, and human flow; uses Redis as a cache database to store real-time data and data feature vectors F; provides an API interface for the environmental control agent to access structured environmental data and data feature vectors F; The environment control agent is used to extract various static information and internal component information of the space from the BIM model and the data collected on site; based on the various static information and internal component information of the space, summarize and construct the attributes of the space object model, including the area, floor height, orientation and internal equipment information of the space; Use the data feature vector F to update the status of the spatial object model in real time; Define the ability of the spatial object model to control various devices within it, that is, the control behavior interface of the action, including: the control behavior interface of the air conditioner, light, curtain and fresh air fan; set the action of the environmental control agent to call the spatial object model, including calling the control behavior interface to execute the corresponding control instructions; The environment control agent is used to output a control instruction sequence corresponding to the first environment control strategy by calling the control behavior interface of the space object model; based on the control instruction sequence, fine adjustment of the building environment is achieved, including: adjusting the air conditioning temperature, light brightness, curtain opening and closing, and fresh air fan operation; The image recognition agent is used to deploy cameras in key areas of the building to ensure that all important areas are covered; configure the camera to connect with the image recognition agent, collect video data in real time, and transmit it to the data processing center through the network; perform image enhancement processing on the collected video data to obtain enhanced video data; perform denoising processing on the enhanced video data to remove noise in the video to obtain denoised video data; use the image recognition large model to perform target detection on the denoised video data, and extract user characteristics, behavior patterns and body language; use machine learning algorithms to identify users' activity areas and behavior habits, and establish user behavior models by analyzing users' walking paths, stay time and activity frequency data; provide data support for environmental adaptation through the user behavior model, so that the environmental control strategy is dynamically adjusted according to the user behavior pattern to obtain analysis results; transmit the user behavior analysis results to the environmental decision-making agent; The system operation and maintenance agent is used to monitor the operating status of the building environment equipment system in real time, including: the working status of the equipment, the agent operation log, the energy consumption, the server resource usage and the instruction response delay; using machine learning algorithms and preset rules, identifying potential equipment problems and faults in the operating status of the building environment system; automatically adjusting equipment parameters when abnormal equipment operation or failure is detected to ensure the stable operation of the system; and sending the adjusted equipment parameters to the environmental decision-making agent; The environmental decision-making agent is used to define user roles at different levels; each user role is assigned corresponding attributes; the allocation of operation permissions is regularly reviewed, and no longer required operation permissions are promptly revoked; the user's needs for environmental parameters are collected in real time through the interactive interface, the user needs are analyzed and organized into structured demand parameters using the semantic understanding ability of the general large model, and the user characteristics and behavior analysis data provided by the image recognition agent and the environmental control status data provided by the control agent are simultaneously collected, and the collected data are analyzed and preprocessed to obtain the current environmental status evaluation results; the environmental decision-making agent formulates an environmental control strategy based on the current environmental status evaluation results and using the Pareto optimization algorithm; Environmental control agents, image recognition agents, system operation and maintenance agents, and environmental decision-making agents are used to achieve information sharing and synchronization through shared databases or message queues to ensure that all agents can obtain the latest environmental data and control instructions; real-time data transmission between agents is achieved through high-speed networks to ensure the timeliness and accuracy of information; encryption technology is used in the data transmission process to ensure data security; during the communication process, if any abnormal situation occurs, the system automatically detects and takes corresponding recovery measures; a unified communication protocol is formulated to ensure smooth information exchange between agents; the communication protocol includes data format, transmission method, error detection and correction mechanism.

Images