An exhibition hall environment control system based on multi-exhibitor user behavior analysis

Abstract

This invention discloses an exhibition hall environment control system based on multi-exhibitor user behavior analysis, addressing the problems of insufficient control precision, poor accommodating of multiple user needs, and lagging linkage in existing systems. The system includes modules for behavioral data collection, multi-user behavior collaborative analysis, environmental demand matching, dynamic control decision-making, execution, and data feedback, forming a closed-loop chain. The collaborative analysis and control decision-making modules are the core innovative modules. Behavioral data collection employs a fusion of visual and millimeter-wave radar methods; the collaborative analysis module outputs dominant behaviors and demand weights through clustering-ranking-fusion logic; the control decision-making module generates optimal instructions through dynamic threshold iteration, cost optimization, and real-time correction; and the data feedback module ensures closed-loop optimization. This invention achieves precise matching between behavioral needs and environmental control, improves the visitor experience, reduces energy consumption, and has a wide range of applications.

Description

Technical Field

[0001] This invention relates to the field of intelligent control of exhibition hall environment, specifically an exhibition hall environment control system based on the analysis of the behavior of multiple exhibitors. Background Technology

[0002] With the increasing popularity of cultural and technological exhibition halls, the demand for comfortable and personalized visitor experiences is rising. Environmental control (such as temperature, humidity, light intensity, air volume, and sound levels) is a core factor influencing the visitor experience, making its precision and adaptability crucial. Existing exhibition hall environmental control systems generally suffer from the following specific problems: The control logic is too simplistic, relying primarily on preset thresholds or average environmental parameters of fixed areas, without considering the actual behavioral characteristics and real-time status of exhibitors. For example, when a large number of users are lingering and staring at exhibits in a certain exhibition area (high-focus behavior), lower environmental noise and suitable light intensity are needed. However, the existing system cannot identify this type of behavior and adjust accordingly, maintaining general environmental parameters, resulting in a decline in user focus. Conversely, when users are quickly moving through the exhibition area (flowing behavior), their sensitivity to temperature and humidity decreases, yet the existing system still maintains high-energy-consuming precise control, leading to energy waste.

[0003] The lack of collaborative analysis of multi-user behavior means that existing technologies often make simple judgments based on the isolated behavior of a single user, without considering the cumulative effect and interactive impact of multiple user behaviors within the same exhibition area. For example, when some users stop to watch, some whisper to each other, and some pass through quickly in the same exhibition area, the environmental parameter requirements for different behaviors vary. Existing systems cannot prioritize and integrate the needs of multiple user behaviors, resulting in control results that fail to take into account the experience of most users, leading to over-control in some areas and under-control in others.

[0004] The linkage between behavioral analysis and environmental control is lagging. In the existing system, there is a significant time lag between the collection and analysis of behavioral data and the issuance of environmental control commands, making it impossible to achieve a closed loop of "behavioral perception - real-time analysis - immediate control". For example, when a large number of users suddenly flood into the exhibition area (gathering behavior), the concentration of carbon dioxide produced by users' breathing will rise rapidly. The existing system needs to go through multiple stages of data collection, transmission, and analysis before activating the fresh air system, causing users to feel stuffy during the control lag phase, which affects the visitor experience.

[0005] To address the aforementioned issues, current technologies lack a system capable of accurately identifying the behavioral characteristics of multiple exhibitors, collaboratively analyzing their behavioral needs, and achieving real-time linkage between behavioral analysis and environmental control. Therefore, designing an exhibition hall environmental control system based on multi-exhibitor behavior analysis to resolve the problems of insufficient control precision, poor accommodating of multiple user needs, and lagging linkage in existing technologies has become a pressing technological bottleneck that needs to be overcome. Summary of the Invention

[0006] The purpose of this invention is to provide an exhibition hall environment control system based on the analysis of the behavior of multiple exhibitors, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an exhibition hall environment control system based on multi-exhibitor user behavior analysis, comprising: a behavior data acquisition module, a multi-user behavior collaborative analysis module, an environmental demand matching module, a dynamic control decision module, an execution module, and a data feedback module; the output end of the behavior data acquisition module is connected to the input end of the multi-user behavior collaborative analysis module, the output end of the multi-user behavior collaborative analysis module is connected to the input end of the environmental demand matching module, the output end of the environmental demand matching module is connected to the input end of the dynamic control decision module, the output end of the dynamic control decision module is connected to the input end of the execution module, the output end of the execution module is connected to the input end of the data feedback module, and the output end of the data feedback module is connected to the input ends of both the multi-user behavior collaborative analysis module and the dynamic control decision module, forming a closed-loop control link; The behavior data acquisition module is used to collect behavior data of exhibitors in various exhibition areas within the exhibition hall, including user location coordinates, movement speed, posture characteristics, dwell time, and distance data between users; it adopts a fusion acquisition method of distributed visual sensors and millimeter-wave radar to ensure the accuracy and real-time nature of data acquisition, with a sampling frequency set to 10 frames / second; The multi-user behavior collaborative analysis module receives raw multi-user behavior data transmitted from the behavior data acquisition module, and completes collaborative analysis of multi-user behavior through behavior clustering, priority ranking and demand fusion algorithms, outputting the dominant behavior type and behavior demand weight of each exhibition area. The environment requirement matching module: pre-stores a matching relationship library of different behavior types and environment parameters, receives the dominant behavior type and behavior requirement weight output by the multi-user behavior collaborative analysis module, retrieves the corresponding benchmark environment parameters from the matching relationship library, and performs preliminary correction on the benchmark environment parameters according to the behavior requirement weight. The dynamic control decision module receives the corrected environmental parameters output by the environmental demand matching module, combines the real-time environmental parameters and equipment operating status transmitted by the data feedback module, and generates the optimal control command through dynamic threshold iteration and control cost optimization algorithm. The execution module includes, but is not limited to, environmental control execution units for air conditioning, fresh air systems, lighting equipment, and audio equipment, which receive control instructions output by the dynamic control decision module and execute corresponding environmental parameter adjustment operations. The data feedback module uses distributed environmental sensors to collect real-time environmental parameters (temperature, humidity, light intensity, carbon dioxide concentration, noise decibels) from each exhibition area. It also collects equipment operating status data (operating power, adjustment range) from the execution module. The collected data is transmitted to the multi-user behavior collaborative analysis module and the dynamic control decision module to achieve dynamic correction of collaborative analysis and closed-loop optimization of control decisions.

[0008] Preferably, the behavioral data acquisition module is implemented using the following steps: The first step is data collection and deployment planning: Distributed visual sensors and millimeter-wave radars are deployed in each exhibition area of ​​the exhibition hall according to the principle of uniform coverage plus densification in key exhibit areas. Four visual sensors and two millimeter-wave radars are deployed in every 100 square meters of exhibition area to form a complementary data collection network. The second step is multi-source data synchronous acquisition: the visual sensor acquires the user's position coordinates and posture feature image data in real time, and the millimeter-wave radar synchronously acquires the user's movement speed and distance between users. The sampling frequency of both is uniformly set to 10 frames / second to ensure data timestamp synchronization. The third step is raw data preprocessing and fusion: The raw data collected by the two types of sensors are time-aligned (based on timestamp matching), and a weighted fusion algorithm is used to eliminate data redundancy and errors. The fusion formula is as follows: ,in For the final behavioral data after fusion, The visual sensor collects data, which includes the user's location coordinates. Posture characteristics ; Data is collected for millimeter-wave radar, and the collected data includes user movement speed. Distance between users ; The weight for visual sensor data is 0.6 (because visual sensors have higher accuracy in position and posture recognition). The data weight for millimeter-wave radar is set to 0.4 (because radar has stronger anti-interference capabilities for dynamic velocity and distance measurements), and satisfies the following conditions: ; The fourth step is data transmission. The fused behavioral data is categorized and packaged according to exhibition area to obtain raw data, which is then transmitted in real time to the multi-user behavior collaborative analysis module. This module adopts a "vision + radar" fusion acquisition method, which can effectively make up for the shortcomings of single sensors (such as visual sensors being easily affected by lighting and radar being insufficient for posture recognition). Combined with a weighted fusion algorithm, it further improves the accuracy of the data, providing a high-quality data foundation for subsequent cluster analysis and priority ranking.

[0009] Preferably, the specific implementation logic of the multi-user behavior collaborative analysis module is as follows: Step 1: Behavioral Data Preprocessing: Receive raw data transmitted from the behavioral data acquisition module, including the location coordinates of each user. Movement speed Posture characteristics When standing and gazing When walking During communication Duration of stay and the distance between users The original data was denoised using [method / process]. The criteria were used to remove outlier data (such as outlier data with a movement speed exceeding 5 m / s or erroneous data with negative dwell time), and the processed data was divided by exhibition area to obtain multi-user behavior datasets for each exhibition area. ,in For the first A user's behavioral feature vector , The number of users in this exhibition area; Step 2: Behavioral clustering based on density peaks: An improved density peak clustering algorithm is used to cluster the multi-user behavior datasets of each exhibition area. Clustering is performed to identify different types of behavioral clusters; the specific process is as follows: Define core clustering parameters: set local density For users Center, radius ,in The number of users within an area of ​​0.5-1m is determined based on the exhibition area size. ,in For households With users Euclidean distance of behavioral feature vectors; setting relative distance For users With all local densities greater than The minimum Euclidean distance between users, i.e. ; Determine cluster centers: Set threshold and threshold , will simultaneously satisfy and The user behavior feature vectors are used as cluster centers; where The limit is set at 30% of the maximum capacity of the exhibition area. Set according to 1 / 2 of the width of the exhibition area aisle; Behavioral clustering: The behavioral feature vectors of each user are assigned to the nearest cluster center to form multiple behavioral clusters. ,in The number of clusters is used to determine the behavior type of each cluster by using the mean of the behavioral feature vectors of each cluster. Behavior types include: gaze behavior. ; communication behavior ; mobility resting behavior ; Step 3, Prioritizing Behavioral Clusters: Based on the weights of user experience impact and energy consumption, a priority evaluation model is constructed to prioritize each behavioral cluster; the specific process is as follows: Determine the evaluation metrics: Select the percentage of users. This refers to the proportion of users in a certain behavior cluster to the total number of users in the exhibition area; experience sensitivity. This refers to the sensitivity of different behaviors to environmental parameters, which is pre-defined through user research: gaze behavior Communication behavior rest and stay behavior , mobility Energy consumption coefficient That is, the energy consumption required to meet the needs of this behavior, which is preset: staring behavior Communication behavior rest and stay behavior , mobility ; Constructing a priority calculation function: Priority ,in It is an inverse indicator of energy consumption weighting; that is, the lower the energy consumption, the higher the weight. Sorting results: Calculate the priority of each row in the cluster. ,according to Sort the behaviors from high to low to obtain a priority sequence. The behavior cluster with the highest priority is the dominant behavior cluster of the exhibition area, and its corresponding behavior type is the dominant behavior type of the exhibition area. Step 4: Multi-behavioral requirement fusion: Based on the priority sequence, a weighted fusion algorithm is used to fuse the environmental requirements of each behavioral cluster to obtain the comprehensive behavioral requirement weight of the exhibition area; the specific process is as follows: Extracting environmental requirement vectors for each behavior cluster: Pre-store the environmental requirement vectors for each behavior type. These vectors include, but are not limited to, the required ranges for temperature, humidity, light intensity, noise levels (decibels), and fresh air volume. For example, the requirement vector for the staring behavior is... The demand vector for communication behavior is wait; Calculate demand weights: based on priority Set the demand weights for each behavior cluster ,in The sum of the cluster priorities of all behaviors; Comprehensive Demand Output: The demand vectors of each behavior cluster are weighted... The weighted fusion is performed to obtain the comprehensive environmental demand scope and demand weight of the exhibition area, and the dominant behavior type and comprehensive demand weight are output to the environmental demand matching module.

[0010] Preferably, the specific implementation logic of the environmental requirement matching module is as follows: The matching relation database is constructed and stored in advance. It includes a pre-built database of "behavior type - environmental parameter" matching relations. This database contains the optimal environmental parameter range and baseline values ​​for each behavior type. Behavior types include gazing behavior, communication behavior, movement behavior, and resting behavior. For example, the parameter corresponding to gazing behavior is temperature. (benchmark value) ),humidity (benchmark value) ), light intensity x (baseline value) ),noise (benchmark value) ), fresh air volume (benchmark value) Each benchmark value is taken as the median value of the corresponding parameter range; Baseline parameters are retrieved, and the dominant behavior types and demand weights of each behavior cluster in the exhibition area are received from the multi-user behavior collaborative analysis module. ( ), and retrieve the baseline environment parameter vector corresponding to the dominant behavior type from the matching relation database first. As an initial reference; The baseline parameters are weighted and adjusted by incorporating the demand weights of non-dominant behavior clusters to ensure that the needs of users with multiple behaviors are taken into account. The adjustment formula is as follows: ,in This is the corrected final baseline environment parameter vector. The demand weight for the dominant behavior cluster, For the first The baseline environment parameter vector corresponding to each non-dominant behavior cluster; The parameter output will be the corrected baseline environment parameter vector. It is transmitted in real time to the dynamic control and decision-making module, serving as the core basis for control and decision-making.

[0011] Preferably, the specific implementation logic of the dynamic control decision module is as follows: Step S1: Receive input data: Receive the corrected environmental parameters, i.e., the baseline parameters, output by the environmental requirements matching module. ,in Reference temperature, For reference humidity, As a reference light intensity, The benchmark noise level is decibels. The baseline fresh air volume is used as the reference, and real-time environmental parameters and operating status data of the execution equipment are received from the data feedback module. The real-time environmental parameters are denoted as follows: The execution equipment operating status data is recorded as ,in This represents the current operating power of the equipment. This is the current adjustment range; Step S2, Dynamic threshold iterative calculation: Based on benchmark parameters With real-time environmental parameters A dynamic threshold iteration model is constructed to calculate the dynamic control thresholds (upper and lower limits) for each environmental parameter, replacing the fixed thresholds in existing technologies to achieve dynamic adaptation of control precision. The specific process is as follows: Initial threshold setting: based on baseline parameters Set an initial threshold range centered on [the target]. ,in This is the initial threshold deviation, a preset value (set according to the type of environmental parameters; the initial deviation for temperature is 0.5℃, the initial deviation for humidity is 5%, etc.). Iterative correction: Calculating real-time environmental parameters With reference parameters deviation ,like Then, an iterative formula is used to correct the temperature threshold deviation: ,in Let be the number of iterations; if If the current threshold deviation remains unchanged, then the iteration termination condition is: Ultimately, the temperature dynamic control threshold range is obtained, namely the upper threshold and the lower threshold. Similarly, the remaining dynamic control threshold ranges can be obtained. ; Threshold validity determination: Based on historical control data transmitted from the data feedback module, the adaptability of the dynamic threshold is calculated. ,like If so, the current dynamic threshold is deemed valid; if If the dynamic threshold is not valid, the iterative calculation will be re-executed until the dynamic threshold is valid. Step S3: Calculation of control cost optimization: A regulation cost optimization model is constructed to minimize regulation energy consumption while meeting dynamic threshold ranges, and preliminary regulation commands are generated. The specific process is as follows: Determine the optimization objective: The optimization objective is to control total energy consumption. Minimum, ,in For the first Energy consumption per unit time of such execution devices For the first Control duration for this type of device; Set constraints: Real-time environmental parameters must be within the dynamic threshold range, i.e. ,in For the equipment adjustment range; The data feedback module collects real-time parameters of the exhibition area; at the same time, the equipment adjustment range must be within a safe range. Optimization results: The particle swarm optimization algorithm is used to solve the control cost optimization model to obtain the optimal adjustment range and adjustment duration of each execution device, and generate preliminary control instructions, including but not limited to air conditioning temperature adjustment, fresh air system air volume adjustment, and lighting equipment light intensity adjustment. Step S4, Real-time Correction and Command Issuance: Based on real-time data from the data feedback module, the initial control commands are dynamically revised to ensure the real-time performance and accuracy of the control. The specific process is as follows: Real-time deviation monitoring: Calculates predicted environmental parameters after the execution of initial control commands. With real-time environmental parameters deviation ; Instruction correction: If If so, the initial control order will remain unchanged; if Then, adjust the adjustment range according to the direction of the deviation, and the correction formula is as follows: ,in For the initial adjustment range, This is the maximum permissible deviation; Command issuance: The revised control command is issued to the execution module in real time, and the command issuance time and control parameters are recorded and transmitted to the data feedback module for subsequent closed-loop optimization.

[0012] Compared with the prior art, the beneficial effects of the present invention are: The multi-user behavior collaborative analysis module of this invention achieves accurate identification of multi-user behavior and determination of demand priority through the innovative logic of clustering-ranking-fusion. It solves the problems of conflicting multi-user behavior demands and poor balancing in the prior art, enabling environmental control to accurately match the dominant behavior demand of the exhibition area and improve the visiting experience of most users. The dynamic control decision module of this invention replaces the fixed threshold control mode in the existing technology with an innovative algorithm of dynamic threshold iteration and control cost optimization, realizing a balance between environmental control accuracy and energy consumption. At the same time, it solves the problem of control linkage lag through a real-time correction mechanism, thereby improving control accuracy while reducing the energy consumption of exhibition hall operation. The entire system of this invention achieves full-process linkage of behavioral data collection, collaborative analysis, demand matching, control decision-making, and execution feedback through the design of a closed-loop control link. This ensures the real-time performance and dynamic adaptability of environmental control, making it suitable for exhibition hall scenarios of different scales and types, and has broad application prospects. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the system structure of the present invention; Figure 2 This is a schematic diagram of the workflow of the multi-user behavior collaborative analysis module of the present invention; Figure 3 This is a schematic diagram of the workflow of the dynamic control decision module of the present invention. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] Please see Figure 1-3This invention provides a technical solution: an exhibition hall environment control system based on multi-exhibitor user behavior analysis, comprising: a behavior data acquisition module, a multi-user behavior collaborative analysis module, an environmental demand matching module, a dynamic control decision module, an execution module, and a data feedback module; the output end of the behavior data acquisition module is connected to the input end of the multi-user behavior collaborative analysis module, the output end of the multi-user behavior collaborative analysis module is connected to the input end of the environmental demand matching module, the output end of the environmental demand matching module is connected to the input end of the dynamic control decision module, the output end of the dynamic control decision module is connected to the input end of the execution module, the output end of the execution module is connected to the input end of the data feedback module, and the output end of the data feedback module is connected to the input ends of both the multi-user behavior collaborative analysis module and the dynamic control decision module, forming a closed-loop control link; Behavioral data acquisition module: Used to collect behavioral data of exhibitors in various exhibition areas within the exhibition hall, including user location coordinates, movement speed, posture characteristics, dwell time, and distance data between users; employing a fusion acquisition method of distributed visual sensors and millimeter-wave radar to ensure the accuracy and real-time nature of data acquisition, with a sampling frequency set to 10 frames / second; the specific implementation steps are as follows: The first step is data collection and deployment planning: Distributed visual sensors and millimeter-wave radars are deployed in each exhibition area of ​​the exhibition hall according to the principle of uniform coverage plus densification in key exhibit areas. Four visual sensors and two millimeter-wave radars are deployed in every 100 square meters of exhibition area to form a complementary data collection network. The second step is multi-source data synchronous acquisition: the visual sensor acquires the user's position coordinates and posture feature image data in real time, and the millimeter-wave radar synchronously acquires the user's movement speed and distance between users. The sampling frequency of both is uniformly set to 10 frames / second to ensure data timestamp synchronization. The third step is raw data preprocessing and fusion: The raw data collected by the two types of sensors are time-aligned (based on timestamp matching), and a weighted fusion algorithm is used to eliminate data redundancy and errors. The fusion formula is as follows: ,in For the final behavioral data after fusion, The visual sensor collects data, which includes the user's location coordinates. Posture characteristics ; Data is collected for millimeter-wave radar, and the collected data includes user movement speed. Distance between users ; The weight for visual sensor data is 0.6 (because visual sensors have higher accuracy in position and posture recognition). The data weight for millimeter-wave radar is set to 0.4 (because radar has stronger anti-interference capabilities for dynamic velocity and distance measurements), and satisfies the following conditions: ; The fourth step is data transmission. The fused behavioral data is categorized and packaged according to exhibition area to obtain raw data, which is then transmitted in real time to the multi-user behavior collaborative analysis module. This module adopts a "vision + radar" fusion acquisition method, which can effectively make up for the shortcomings of single sensors (such as visual sensors being easily affected by lighting and radar being insufficient for posture recognition). Combined with a weighted fusion algorithm, it further improves the accuracy of the data, providing a high-quality data foundation for subsequent cluster analysis and priority ranking.

[0016] Multi-user behavior collaborative analysis module: Receives raw multi-user behavior data transmitted from the behavior data acquisition module, and completes collaborative analysis of multi-user behavior through behavior clustering, priority ranking, and demand fusion algorithms, outputting the dominant behavior types and behavior demand weights for each exhibition area; the specific implementation logic of the multi-user behavior collaborative analysis module is as follows: Step 1: Behavioral Data Preprocessing: Receive raw data transmitted from the behavioral data acquisition module, including the location coordinates of each user. Movement speed Posture characteristics When standing and gazing When walking During communication Duration of stay and the distance between users The original data was denoised using [method / process]. The criteria were used to remove outlier data (such as outlier data with a movement speed exceeding 5 m / s or erroneous data with negative dwell time), and the processed data was divided by exhibition area to obtain multi-user behavior datasets for each exhibition area. ,in For the first A user's behavioral feature vector , The number of users in this exhibition area; Step 2: Behavioral clustering based on density peaks: An improved density peak clustering algorithm is used to cluster the multi-user behavior datasets of each exhibition area. Clustering is performed to identify different types of behavioral clusters; the specific process is as follows: Define core clustering parameters: set local density For users Center, radius ,in The number of users within an area of ​​0.5-1m is determined based on the exhibition area size. ,in For households With users Euclidean distance of behavioral feature vectors; setting relative distance For users With all local densities greater than The minimum Euclidean distance between users, i.e. ; Determine cluster centers: Set threshold and threshold , will simultaneously satisfy and The user behavior feature vectors are used as cluster centers; where The limit is set at 30% of the maximum capacity of the exhibition area. Set according to 1 / 2 of the width of the exhibition area aisle; Behavioral clustering: The behavioral feature vectors of each user are assigned to the nearest cluster center to form multiple behavioral clusters. ,in The number of clusters is used to determine the behavior type of each cluster by using the mean of the behavioral feature vectors of each cluster. Behavior types include: gaze behavior. ; communication behavior ; mobility resting behavior ; Step 3, Prioritizing Behavioral Clusters: Based on the weights of user experience impact and energy consumption, a priority evaluation model is constructed to prioritize each behavioral cluster; the specific process is as follows: Determine the evaluation metrics: Select the percentage of users. This refers to the proportion of users in a certain behavior cluster to the total number of users in the exhibition area; experience sensitivity. This refers to the sensitivity of different behaviors to environmental parameters, which is pre-defined through user research: gaze behavior Communication behavior rest and stay behavior , mobility Energy consumption coefficient That is, the energy consumption required to meet the needs of this behavior, which is preset: staring behavior Communication behavior rest and stay behavior , mobility ; Constructing a priority calculation function: Priority ,in It is an inverse indicator of energy consumption weighting; that is, the lower the energy consumption, the higher the weight. Sorting results: Calculate the priority of each row in the cluster. ,according to Sort the behaviors from high to low to obtain a priority sequence. The behavior cluster with the highest priority is the dominant behavior cluster of the exhibition area, and its corresponding behavior type is the dominant behavior type of the exhibition area. Step 4: Multi-behavioral requirement fusion: Based on the priority sequence, a weighted fusion algorithm is used to fuse the environmental requirements of each behavioral cluster to obtain the comprehensive behavioral requirement weight of the exhibition area; the specific process is as follows: Extracting environmental requirement vectors for each behavior cluster: Pre-store the environmental requirement vectors for each behavior type. These vectors include, but are not limited to, the required ranges for temperature, humidity, light intensity, noise levels (decibels), and fresh air volume. For example, the requirement vector for the staring behavior is... The demand vector for communication behavior is wait; Calculate demand weights: based on priority Set the demand weights for each behavior cluster ,in The sum of the cluster priorities of all behaviors; Comprehensive Demand Output: The demand vectors of each behavior cluster are weighted... The weighted fusion is performed to obtain the comprehensive environmental demand scope and demand weight of the exhibition area, and the dominant behavior type and comprehensive demand weight are output to the environmental demand matching module.

[0017] Environment requirement matching module: This module pre-stores a matching relationship database for different behavior types and environment parameters. It receives the dominant behavior type and behavior requirement weights output by the multi-user behavior collaborative analysis module, retrieves the corresponding baseline environment parameters from the matching relationship database, and performs preliminary corrections to the baseline environment parameters based on the behavior requirement weights. The specific implementation logic is as follows: The matching relation database is constructed and stored in advance. It includes a pre-built database of "behavior type - environmental parameter" matching relations. This database contains the optimal environmental parameter range and baseline values ​​for each behavior type. Behavior types include gazing behavior, communication behavior, movement behavior, and resting behavior. For example, the parameter corresponding to gazing behavior is temperature. (benchmark value) ),humidity (benchmark value) ), light intensity x (baseline value) ),noise (benchmark value) ), fresh air volume (benchmark value) Each benchmark value is taken as the median value of the corresponding parameter range; Baseline parameters are retrieved, and the dominant behavior types and demand weights of each behavior cluster in the exhibition area are received from the multi-user behavior collaborative analysis module. ( ), and retrieve the baseline environment parameter vector corresponding to the dominant behavior type from the matching relation database first. As an initial reference; The baseline parameters are weighted and adjusted by incorporating the demand weights of non-dominant behavior clusters to ensure that the needs of users with multiple behaviors are taken into account. The adjustment formula is as follows: ,in This is the corrected final baseline environment parameter vector. The demand weight for the dominant behavior cluster, For the first The baseline environment parameter vector corresponding to each non-dominant behavior cluster; The parameter output will be the corrected baseline environment parameter vector. It is transmitted in real time to the dynamic control and decision-making module, serving as the core basis for control and decision-making.

[0018] Dynamic control decision-making module: Receives corrected environmental parameters from the environmental demand matching module, combines them with real-time environmental parameters and equipment operating status transmitted by the data feedback module, and generates the optimal control command through dynamic threshold iteration and control cost optimization algorithms; the specific implementation logic is as follows: Step S1: Receive input data: Receive the corrected environmental parameters, i.e., the baseline parameters, output by the environmental requirements matching module. ,in Reference temperature, For reference humidity, As a reference light intensity, The benchmark noise level is decibels. The baseline fresh air volume is used as the reference, and real-time environmental parameters and operating status data of the execution equipment are received from the data feedback module. The real-time environmental parameters are denoted as follows: The execution equipment operating status data is recorded as ,in This represents the current operating power of the equipment. This is the current adjustment range; Step S2, Dynamic threshold iterative calculation: Based on benchmark parameters With real-time environmental parameters A dynamic threshold iteration model is constructed to calculate the dynamic control thresholds (upper and lower limits) for each environmental parameter, replacing the fixed thresholds in existing technologies to achieve dynamic adaptation of control precision. The specific process is as follows: Initial threshold setting: based on baseline parameters Set an initial threshold range centered on [the target]. ,in This is the initial threshold deviation, a preset value (set according to the type of environmental parameters; the initial deviation for temperature is 0.5℃, the initial deviation for humidity is 5%, etc.). Iterative correction: Calculating real-time environmental parameters With reference parameters deviation ,like Then, an iterative formula is used to correct the temperature threshold deviation: ,in Let be the number of iterations; if If the current threshold deviation remains unchanged, then the iteration termination condition is: Ultimately, the temperature dynamic control threshold range is obtained, namely the upper threshold and the lower threshold. Similarly, the remaining dynamic control threshold ranges can be obtained. ; Threshold validity determination: Based on historical control data transmitted from the data feedback module, the adaptability of the dynamic threshold is calculated. ,like If so, the current dynamic threshold is deemed valid; if If the dynamic threshold is not valid, the iterative calculation will be re-executed until the dynamic threshold is valid. Step S3: Calculation of control cost optimization: A regulation cost optimization model is constructed to minimize regulation energy consumption while meeting dynamic threshold ranges, and preliminary regulation commands are generated. The specific process is as follows: Determine the optimization objective: The optimization objective is to control total energy consumption. Minimum, ,in For the first Energy consumption per unit time of such execution devices For the first Control duration for this type of device; Set constraints: Real-time environmental parameters must be within the dynamic threshold range, i.e. ,in For the equipment adjustment range; The data feedback module collects real-time parameters of the exhibition area; at the same time, the equipment adjustment range must be within a safe range. Optimization results: The particle swarm optimization algorithm is used to solve the control cost optimization model to obtain the optimal adjustment range and adjustment duration of each execution device, and generate preliminary control instructions, including but not limited to air conditioning temperature adjustment, fresh air system air volume adjustment, and lighting equipment light intensity adjustment. The steps for generating preliminary control commands are explained below: With the core objective of minimizing total energy consumption, and based on the environmental parameter constraints and equipment safety constraints set earlier, a particle swarm optimization algorithm is used to solve the optimization model, obtaining the optimal adjustment range and duration for each actuator, and generating preliminary control commands. Specifically... Step 1: Initialize algorithm parameters and determine the core parameters of the particle swarm, where the particle dimension corresponds to the adjustment parameters of each execution device (such as the temperature adjustment range of an air conditioner). Fresh air system air volume adjustment range Lighting equipment illumination adjustment range (etc.), particle population size set to 30-50 (balancing solution efficiency and accuracy), maximum number of iterations set to 50 (avoiding excessive iteration leading to control lag), learning factor , All parameters are set to 2 (classic parameter configuration to ensure particle convergence speed), inertia weight. The initial value is set to 0.9 (to enhance global search capabilities in the early stages of iteration). The second step is to construct the particle fitness function to regulate the total energy consumption formula. Let be the fitness function, where Total energy consumption, For the first Energy consumption per unit time for this type of equipment For the first The control time for this type of equipment (calculated from the adjustment range and the equipment's safe adjustment rate, such as air conditioning) A lower fitness value indicates a better control scheme; The third step is population iterative optimization. Initialize the particle population (randomly generate feasible solutions for the adjustment range of each device), calculate the fitness value of each particle, and record the optimal fitness value and corresponding particle position (optimal control scheme). Then iteratively update the particle velocity and position. The velocity update formula is: , ( For particle velocity, For the number of iterations, For the optimal position of an individual, The optimal position for the population (The number is a random number between 0 and 1). The position update formula is: At the same time, ensure that the updated particle positions satisfy the constraints set above (such as...). , (etc.), the positions of particles that do not meet the constraints are truncated and corrected; The fourth step is to terminate the iteration and output the results. When the number of iterations reaches the maximum number of iterations, or when the change in the optimal fitness value of the population is ≤0.01 after 5 consecutive iterations, the iteration is terminated. The adjustment range and duration of each device corresponding to the optimal position of the final population are extracted, and preliminary control instructions are generated. The instructions include specific parameters such as adjusting the temperature of the air conditioner, adjusting the air volume of the fresh air system, adjusting the light intensity of the lighting equipment, and reducing the noise level of the audio equipment.

[0019] Step S4, Real-time Correction and Command Issuance: Based on real-time data from the data feedback module, the initial control commands are dynamically revised to ensure the real-time performance and accuracy of the control. The specific process is as follows: Real-time deviation monitoring: Calculates predicted environmental parameters after the execution of initial control commands. With real-time environmental parameters deviation ; Instruction correction: If If so, the initial control order will remain unchanged; if Then, adjust the adjustment range according to the direction of the deviation, and the correction formula is as follows: ,in For the initial adjustment range, This is the maximum permissible deviation; Command issuance: The revised control command is issued to the execution module in real time, and the command issuance time and control parameters are recorded and transmitted to the data feedback module for subsequent closed-loop optimization. The execution module includes, but is not limited to, environmental control execution units for air conditioning, fresh air system, lighting equipment, and audio equipment. It receives the control command output by the dynamic control decision module and executes the corresponding environmental parameter adjustment operation. Data feedback module: It uses distributed environmental sensors to collect real-time environmental parameters (temperature, humidity, light intensity, carbon dioxide concentration, noise decibels) of each exhibition area, and at the same time collects equipment operation status data (operating power, adjustment range) of the execution module. The collected data is transmitted to the multi-user behavior collaborative analysis module and the dynamic control decision module to realize the dynamic correction of collaborative analysis and the closed-loop optimization of control decision.

[0020] The data feedback module is briefly described below: As a key supporting unit for the system's closed-loop control, its core function is to collect real-time environmental status and equipment operation data after control, providing accurate and synchronous data for front-end behavior analysis and correction, and back-end control decision optimization, thus ensuring the integrity and effectiveness of the closed-loop chain. The specific working steps are as follows: A. Data collection node deployment and planning: In accordance with the principle of "full coverage of the exhibition area + deployment of control equipment nearby", distributed environmental sensors are evenly deployed in each exhibition area (1 unit deployed every 50㎡). At the same time, status data collection sensors are installed at the end of each execution device such as air conditioning and fresh air system to form a two-dimensional data collection network of "environmental parameters + equipment status". B. Multi-dimensional data synchronous acquisition: Distributed environmental sensors collect core environmental parameters of each exhibition area in real time, including temperature, humidity, light intensity, carbon dioxide concentration, and noise decibels. The sampling frequency is consistent with that of the behavioral data acquisition module (10 frames / second) to ensure data timestamp synchronization. Equipment status sensors synchronously collect the operating data of the execution module, including key parameters such as the operating power of each device, actual adjustment range, and running time. The acquisition frequency is 5 frames / second to ensure the real-time monitoring of equipment status. C. Data preprocessing: The collected raw data is cleaned and standardized: the moving average filtering method (window size set to 5) is used to remove instantaneous fluctuation noise in environmental parameters, and the threshold judgment method is used to remove outliers in equipment status data (such as negative power or invalid data with adjustment range exceeding the safe range). Then all data is standardized and packaged in the format of "exhibition area-timestamp-parameter type" to form a structured data packet. D. Precise data distribution and transmission: Structured data packets are classified and transmitted to corresponding modules according to their purpose. Real-time environmental parameters are synchronously transmitted to the multi-user behavior collaborative analysis module and the dynamic control decision module to provide a basis for dynamic correction of behavior analysis (such as combining environmental comfort feedback when judging user behavior) and real-time correction of control decisions. Equipment operating status data is mainly transmitted to the dynamic control decision module and simultaneously synchronized to the multi-user behavior collaborative analysis module for energy consumption optimization model iteration and behavior-energy consumption correlation analysis. E. Data Storage and Traceability: All pre-processed data is simultaneously stored in the system's local database, retaining at least 72 hours of historical data to support dynamic threshold adaptation and long-term system optimization iterations. This module achieves a closed-loop linkage of "execution results - data feedback - decision optimization" through dual-dimensional data collection and precise distribution, making multi-user behavior collaborative analysis more aligned with real-world scenarios and dynamic control decisions more targeted and scientific.

[0021] This invention discloses an exhibition hall environment control system based on multi-exhibitor behavior analysis, aiming to solve the technical problems of insufficient control accuracy, poor accommodating of multiple user needs, and lagging linkage in existing exhibition hall environment control systems. The system includes a behavior data acquisition module, a multi-user behavior collaborative analysis module, an environmental demand matching module, a dynamic control decision-making module, an execution module, and a data feedback module. These modules are sequentially connected to form a closed-loop control chain, with the multi-user behavior collaborative analysis module and the dynamic control decision-making module being the core innovative modules.

[0022] The core of the technical solution lies in the following: the behavior data acquisition module collects multi-user behavior data through the fusion of visual sensors and millimeter-wave radar; the multi-user behavior collaborative analysis module adopts a three-level logic of "clustering-ranking-fusion" to achieve accurate identification of multi-user behavior and determination of demand priority, and outputs the dominant behavior type and demand weight; the environmental demand matching module transforms behavioral demands into corrected baseline environmental parameters; the dynamic control decision module generates the optimal control command through a three-level decision logic of "dynamic threshold iteration-control cost optimization-real-time correction"; and the data feedback module collects environmental and equipment status data in real time to achieve closed-loop optimization.

[0023] This invention, through innovative design, achieves precise matching of multi-user behavioral needs and real-time dynamic control of environmental parameters. While improving the visitor experience for most users, it reduces the energy consumption of exhibition hall operation. It is applicable to exhibition hall scenarios of different sizes and types and has broad application prospects.

[0024] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A venue environment control system based on multi-exhibitor user behavior analysis, characterized in that, include: The system includes a behavior data acquisition module, a multi-user behavior collaborative analysis module, an environmental demand matching module, a dynamic control and decision-making module, an execution module, and a data feedback module. The output of the behavior data acquisition module is connected to the input of the multi-user behavior collaborative analysis module. The output of the multi-user behavior collaborative analysis module is connected to the input of the environmental demand matching module. The output of the environmental demand matching module is connected to the input of the dynamic control decision module. The output of the dynamic control decision module is connected to the input of the execution module. The output of the execution module is connected to the input of the data feedback module. The output of the data feedback module is connected to the inputs of the multi-user behavior collaborative analysis module and the dynamic control decision module, forming a closed-loop control link. The behavior data acquisition module is used to collect behavior data of exhibitors in various exhibition areas within the exhibition hall, including user location coordinates, movement speed, posture characteristics, dwell time, and distance data between users; it adopts a fusion acquisition method of distributed visual sensors and millimeter-wave radar to ensure the accuracy and real-time nature of data acquisition, with a sampling frequency set to 10 frames / second; The multi-user behavior collaborative analysis module receives raw multi-user behavior data transmitted from the behavior data acquisition module, and completes collaborative analysis of multi-user behavior through behavior clustering, priority ranking and demand fusion algorithms, outputting the dominant behavior type and behavior demand weight of each exhibition area. The environment requirement matching module: pre-stores a matching relationship library of different behavior types and environment parameters, receives the dominant behavior type and behavior requirement weight output by the multi-user behavior collaborative analysis module, retrieves the corresponding benchmark environment parameters from the matching relationship library, and performs preliminary correction on the benchmark environment parameters according to the behavior requirement weight. The dynamic control decision module receives the corrected environmental parameters output by the environmental demand matching module, combines the real-time environmental parameters and equipment operating status transmitted by the data feedback module, and generates the optimal control command through dynamic threshold iteration and control cost optimization algorithm. The execution module includes, but is not limited to, environmental control execution units for air conditioning, fresh air systems, lighting equipment, and audio equipment, which receive control instructions output by the dynamic control decision module and execute corresponding environmental parameter adjustment operations. The data feedback module uses distributed environmental sensors to collect real-time environmental parameters of each exhibition area, and simultaneously collects equipment operation status data from the execution module. The collected data is then transmitted to the multi-user behavior collaborative analysis module and the dynamic control decision module to achieve dynamic correction of collaborative analysis and closed-loop optimization of control decisions.

2. The exhibition hall environment control system based on multi-exhibitor user behavior analysis according to claim 1, characterized in that: The specific implementation steps of the behavior data acquisition module are as follows: The first step is data collection and deployment planning: Distributed visual sensors and millimeter-wave radars are deployed in each exhibition area of ​​the exhibition hall according to the principle of uniform coverage plus densification in key exhibit areas. Four visual sensors and two millimeter-wave radars are deployed in every 100 square meters of exhibition area to form a complementary data collection network. The second step is multi-source data synchronous acquisition: the visual sensor acquires the user's position coordinates and posture feature image data in real time, and the millimeter-wave radar synchronously acquires the user's movement speed and distance between users. The sampling frequency of both is uniformly set to 10 frames / second to ensure data timestamp synchronization. The third step is raw data preprocessing and fusion: the raw data collected by the two types of sensors are time-aligned, and a weighted fusion algorithm is used to eliminate data redundancy and errors. The fusion formula is as follows: ,in For the final behavioral data after fusion, The visual sensor collects data, which includes the user's location coordinates. Posture characteristics ; Data is collected for millimeter-wave radar, and the collected data includes user movement speed. Distance between users ; Weights for visual sensor data; For millimeter-wave radar data weights, and satisfying ; The fourth step is data transmission. The merged behavioral data is packaged according to the exhibition area to obtain the raw data, which is then transmitted in real time to the multi-user behavior collaborative analysis module.

3. The exhibition hall environment control system based on multi-exhibitor user behavior analysis according to claim 1, characterized in that: The specific implementation logic of the multi-user behavior collaborative analysis module is as follows: Step 1: Behavioral Data Preprocessing: Receive raw data transmitted from the behavioral data acquisition module, including the location coordinates of each user. Movement speed Posture characteristics When standing and gazing When walking During communication Duration of stay and the distance between users ; The original data was denoised using the following method: The criteria were used to remove outlier data, and the processed data was divided into exhibition areas to obtain multi-user behavior datasets for each exhibition area. ,in For the first A user's behavioral feature vector , The number of users in this exhibition area; Step 2: Behavioral clustering based on density peaks: An improved density peak clustering algorithm is used to cluster the multi-user behavior datasets of each exhibition area. Clustering is performed to identify different types of behavioral clusters; the specific process is as follows: Define core clustering parameters: set local density For users Center, radius ,in The number of users within an area of ​​0.5-1m is determined based on the exhibition area size. ,in For households With users Euclidean distance of behavioral feature vectors; setting relative distance For users With all local densities greater than The minimum Euclidean distance between users, i.e. ; Determine cluster centers: Set threshold and threshold , will simultaneously satisfy and The user behavior feature vectors are used as cluster centers; where The limit is set at 30% of the maximum capacity of the exhibition area. Set according to 1 / 2 of the width of the exhibition area aisle; Behavioral clustering: The behavioral feature vectors of each user are assigned to the nearest cluster center to form multiple behavioral clusters. ,in This represents the number of clusters. The behavior type of each cluster is determined by the mean of its behavioral feature vectors. Behavior types include: gaze behavior. ; communication behavior ; mobility resting behavior ; Step 3, Prioritizing Behavioral Clusters: Based on the weights of user experience impact and energy consumption, a priority evaluation model is constructed to prioritize each behavioral cluster; the specific process is as follows: Determine the evaluation metrics: Select the percentage of users. This refers to the proportion of users in a certain behavior cluster to the total number of users in the exhibition area; experience sensitivity. This refers to the sensitivity of different behaviors to environmental parameters, which is pre-defined through user research: gaze behavior Communication behavior rest and stay behavior , mobility Energy consumption coefficient That is, the energy consumption required to meet the needs of this behavior, which is preset: staring behavior Communication behavior rest and stay behavior , mobility ; Constructing a priority calculation function: Priority ,in It is an inverse indicator of energy consumption weighting; that is, the lower the energy consumption, the higher the weight. Sorting results: Calculate the priority of each row in the cluster. ,according to Sort the behaviors from high to low to obtain a priority sequence. The behavior cluster with the highest priority is the dominant behavior cluster of the exhibition area, and its corresponding behavior type is the dominant behavior type of the exhibition area. Step 4: Multi-behavioral requirement fusion: Based on the priority sequence, a weighted fusion algorithm is used to fuse the environmental requirements of each behavioral cluster to obtain the comprehensive behavioral requirement weight of the exhibition area; the specific process is as follows: Extract the environmental requirement vectors for each behavior cluster: Pre-store the environmental requirement vectors for each behavior type, including but not limited to the required ranges for temperature, humidity, light intensity, noise decibels, and fresh air volume; Calculate demand weights: based on priority Set the demand weights for each behavior cluster ,in The sum of the cluster priorities of all behaviors; Comprehensive Demand Output: The demand vectors of each behavior cluster are weighted... The weighted fusion is performed to obtain the comprehensive environmental demand scope and demand weight of the exhibition area, and the dominant behavior type and comprehensive demand weight are output to the environmental demand matching module.

4. The exhibition hall environment control system based on multi-exhibitor user behavior analysis according to claim 1, characterized in that: The specific implementation logic of the environmental demand matching module is as follows: The matching relationship database is constructed and stored in advance. The matching relationship database of "behavior type-environmental parameter" is pre-built and stored. The database contains the optimal environmental parameter range and benchmark value corresponding to each behavior type. The behavior types include staring behavior, communication behavior, movement behavior, and resting behavior. Baseline parameters are retrieved, and the dominant behavior types and demand weights of each behavior cluster in the exhibition area are received from the multi-user behavior collaborative analysis module. First, retrieve the baseline environment parameter vector corresponding to the dominant behavior type from the matching relationship database. As an initial reference; The baseline parameters are weighted and adjusted by incorporating the demand weights of non-dominant behavior clusters to ensure that the needs of users with multiple behaviors are taken into account. The adjustment formula is as follows: ,in This is the corrected final baseline environment parameter vector. The demand weight for the dominant behavior cluster, For the first The baseline environment parameter vector corresponding to each non-dominant behavior cluster; The parameter output will be the corrected baseline environment parameter vector. It is transmitted in real time to the dynamic control and decision-making module, serving as the core basis for control and decision-making.

5. The exhibition hall environment control system based on multi-exhibitor user behavior analysis according to claim 1, characterized in that: The specific implementation logic of the dynamic control decision-making module is as follows: Step S1: Receive input data: Receive the corrected environmental parameters, i.e., the baseline parameters, output by the environmental requirements matching module. ,in Reference temperature, For reference humidity, As a reference light intensity, The benchmark noise level is decibels. The baseline fresh air volume is used as the reference, and real-time environmental parameters and operating status data of the execution equipment are received from the data feedback module. The real-time environmental parameters are denoted as follows: The execution equipment operating status data is recorded as ,in This represents the current operating power of the equipment. This is the current adjustment range; Step S2, Dynamic threshold iterative calculation: Based on benchmark parameters With real-time environmental parameters A dynamic threshold iteration model is constructed to calculate the dynamic adjustment thresholds for various environmental parameters, replacing the fixed thresholds in existing technologies and achieving dynamic adaptation of adjustment accuracy. The specific process is as follows: Initial threshold setting: based on baseline parameters Set an initial threshold range centered on [the target]. ,in This represents the initial threshold deviation, which is a pre-set value. Iterative correction: Calculating real-time environmental parameters With reference parameters deviation ,like Then, an iterative formula is used to correct the temperature threshold deviation: ,in Let be the number of iterations; if If the current threshold deviation remains unchanged, then the iteration termination condition is: Ultimately, the temperature dynamic control threshold range is obtained, namely the upper threshold and the lower threshold. Similarly, the remaining dynamic control threshold ranges can be obtained. ; Threshold validity determination: Based on historical control data transmitted from the data feedback module, the adaptability of the dynamic threshold is calculated. ,like If so, the current dynamic threshold is deemed valid; if If the dynamic threshold is not valid, the iterative calculation will be re-executed until the dynamic threshold is valid. Step S3: Calculation of control cost optimization: A regulation cost optimization model is constructed to minimize regulation energy consumption while meeting dynamic threshold ranges, and preliminary regulation commands are generated. The specific process is as follows: Determine the optimization objective: The optimization objective is to control total energy consumption. Minimum, ,in For the first Energy consumption per unit time of such execution devices For the first Control duration for this type of device; Set constraints: Real-time environmental parameters must be within the dynamic threshold range, i.e. ,in This refers to the adjustment range of the equipment. The data feedback module collects real-time parameters of the exhibition area; at the same time, the equipment adjustment range must be within a safe range. Optimization results: The particle swarm optimization algorithm is used to solve the control cost optimization model to obtain the optimal adjustment range and adjustment duration of each execution device, and generate preliminary control instructions, including but not limited to air conditioning temperature adjustment, fresh air system air volume adjustment, and lighting equipment light intensity adjustment. Step S4, Real-time Correction and Command Issuance: Based on real-time data from the data feedback module, the initial control commands are dynamically revised to ensure the real-time performance and accuracy of the control. The specific process is as follows: Real-time deviation monitoring: Calculates predicted environmental parameters after the execution of initial control commands. With real-time environmental parameters deviation ; Instruction correction: If If so, the initial control order will remain unchanged; if Then, adjust the adjustment range according to the direction of the deviation, and the correction formula is as follows: ,in For the initial adjustment range, This is the maximum permissible deviation; Command issuance: The revised control command is issued to the execution module in real time, and the command issuance time and control parameters are recorded and transmitted to the data feedback module for subsequent closed-loop optimization.

Images