Intelligent exhibition hall control method and system

Abstract

The invention discloses an intelligent exhibition hall control method and system, relates to the technical field of intelligent management, and solves the problem that it is difficult to analyze monitoring data in an exhibition hall and comprehensively evaluate the exhibition hall. A neural network prediction model is difficult to establish for analysis and prediction according to the comprehensive evaluation result of the exhibition hall, so that intelligent scheduling control is performed on the exhibition hall; the thermal image data is difficult to be combined with a video monitoring image to carry out anomaly detection; and an optimal visiting route is difficult to generate according to a comprehensive evaluation result and an abnormal detection result of the exhibition hall. According to the invention, comprehensive evaluation is carried out on the exhibition hall by analyzing monitoring data; using a neural network prediction model to analyze and predict audience flow and energy consumption demands so as to carry out intelligent scheduling control; abnormality detection is carried out by fusing a thermal image and video monitoring, and a visiting route is optimized according to a comprehensive evaluation result.

Description

Technical Field

[0001] The invention belongs to the technical field of intelligent management, and in particular relates to an intelligent exhibition hall control method and system. Background Art

[0002] With the popularization of the Internet and mobile Internet, the development of information technology has provided exhibition halls with more display means and interactive methods, allowing exhibition halls to achieve richer content and forms. Big data analysis technology and artificial intelligence technology analyze and process various data in the exhibition hall to achieve intelligent control of the exhibition hall. By integrating and applying machine learning algorithms, it provides strong support for the intelligent management of the exhibition hall and realizes the efficient operation and high-quality service of the exhibition hall.

[0003] The following problems exist in the existing technology, including: it is difficult to analyze the monitoring data in the exhibition hall and conduct a comprehensive evaluation of the exhibition hall; it is difficult to establish a neural network prediction model for analysis and prediction based on the comprehensive evaluation results of the exhibition hall, so as to perform intelligent scheduling and control of the exhibition hall; it is difficult to combine thermal image data with video surveillance images for anomaly detection; it is difficult to generate the best visiting route based on the comprehensive evaluation results and anomaly detection results of the exhibition hall. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an intelligent exhibition hall control method for solving the above-mentioned problem. To this end, the first aspect of the present invention provides an intelligent exhibition hall control method, comprising the following steps:

[0005] S1: Real-time collection of monitoring data in the exhibition hall, including: environmental data, video surveillance images, audience behavior data and equipment operation data; use thermal imaging cameras to collect thermal image data in the exhibition hall in real time;

[0006] S2: preprocessing the collected monitoring data including: denoising, data cleaning and standardization; preprocessing the collected thermal image data and video surveillance images including: denoising, calibration and image enhancement;

[0007] S3: Analyze and calculate the environmental comfort index, the heat index of the i-th exhibit and the equipment operation efficiency index based on the collected and pre-processed monitoring data; conduct a comprehensive evaluation of the exhibition hall based on the calculation results of the analysis of the monitoring data; establish a neural network prediction model based on the comprehensive evaluation results of the exhibition hall to conduct analysis and prediction, and perform intelligent scheduling and control of the exhibition hall based on the prediction results;

[0008] S4: Perform anomaly detection based on the pre-processed thermal image data combined with the video surveillance image; monitor and warn the exhibition hall based on the anomaly detection results;

[0009] S5: Generate the best tour route based on the comprehensive evaluation results and anomaly detection results of the exhibition hall.

[0010] Preferably, the monitoring data in the exhibition hall is collected in real time in step S1, including: environmental data, video surveillance images, audience behavior data and equipment operation data, including the following steps: using sensors and surveillance cameras to collect monitoring data in the exhibition hall in real time, the environmental data including: temperature, humidity, light intensity and air quality; the audience behavior data including: audience flow, visiting path, length of stay and the number of times the exhibits are viewed or interacted with by the audience; the equipment operation data including: equipment startup time, number of startups, usage frequency and energy consumption data.

[0011] Preferably, the step S3 includes the following steps:

[0012] Calculate the environmental data scoring function formula:

[0013] Among them, x represents the environmental data in the exhibition hall collected in real time, x c represents the optimal value of the environmental data in the exhibition hall, and K represents the steepness parameter of the environmental data scoring function curve;

[0014] By substituting the real-time collected and pre-processed temperature, humidity, light intensity and air quality in the exhibition hall into the environmental data scoring function formula, the temperature score, humidity score, light intensity score and air quality score are obtained respectively, and the temperature score, humidity score, light intensity score and air quality score are weighted and summed to obtain the environmental comfort index E;

[0015] The formula for calculating the heat index of the i-th exhibit is:

[0016] Get the heat index P of the i-th exhibit i ; Among them, S i represents the number of times the i-th exhibit is viewed or interacted with by visitors, F i represents the visitor flow in the i-th exhibit area, F T represents the total visitor flow in the exhibition hall, T i represents the length of time visitors stay in the i-th exhibit area, T a represents the average length of stay of visitors in all exhibition areas; α, β and γ represent weight coefficients, and the sum of α, β and γ is 1;

[0017] By calculating the equipment operation efficiency index formula:

[0018] The equipment operation efficiency index η is obtained; where f represents the frequency of equipment use, t represents the total start-up time of all equipment in the exhibition hall, M represents the total number of startups of all equipment, N represents the actual energy consumption, and N std Indicates standard energy consumption.

[0019] Preferably, the step S3 of comprehensively evaluating the exhibition hall according to the calculation results of analyzing the monitoring data includes the following steps:

[0020] According to the calculation results of the analysis of monitoring data, the environmental comfort index, the heat index of the i-th exhibit and the equipment operation efficiency index are obtained, and the comprehensive evaluation index formula of the exhibition hall is calculated:

[0021] The comprehensive evaluation index Z of the exhibition hall is obtained, where W1, W2 and W3 represent weight coefficients respectively; E represents the environmental comfort index, P i represents the heat index of the i-th exhibit, n represents the number of exhibits in the exhibition hall, i∈{1, 2, ..., n}, and η represents the equipment operation efficiency index.

[0022] Preferably, in step S3, a neural network prediction model is established according to the comprehensive evaluation results of the exhibition hall to perform analysis and prediction, and intelligent scheduling and control of the exhibition hall is performed according to the prediction results, including the following steps:

[0023] Based on the comprehensive evaluation results of the exhibition hall, the neural network prediction model in the machine learning algorithm is used to predict the visitor flow and equipment energy consumption requirements;

[0024] Collect the historical monitoring data of the exhibition hall, and pre-process the historical monitoring data including: data cleaning and normalization; according to the processed historical monitoring data, extract the characteristic parameters including: environmental comfort index, the heat index of the i-th exhibit, the equipment operation efficiency index and the comprehensive evaluation index of the exhibition hall;

[0025] Input the characteristic parameters into the neural network prediction model for training; according to the trained neural network prediction model, by inputting the monitoring data collected and processed in real time, the neural network prediction model predicts the audience flow and equipment energy consumption requirements;

[0026] According to the prediction results, the exhibition hall is intelligently dispatched and controlled, and the dispatching includes: environmental data dispatching, exhibit display dispatching and equipment energy consumption dispatching.

[0027] Preferably, the step S4 performs abnormality detection based on the preprocessed thermal image data combined with the video surveillance image, comprising the following steps:

[0028] Analyze the pre-processed thermal image data through image processing algorithms to identify hot spots in the thermal image data; convert the temperature information of the hot spot area into numerical form through color mapping or gray value conversion methods; smooth the temperature distribution map using interpolation algorithms or spatial filtering technology; extract temperature features based on the smoothed temperature distribution map, including: maximum temperature, average temperature, temperature gradient, number of hot spots and hot spot area;

[0029] Extract visual features from the preprocessed video surveillance images using image processing algorithms, including geometric moments, color moments, and motion trajectories; perform multimodal feature fusion on visual features and temperature features to generate multimodal feature vectors;

[0030] By utilizing the neural network model in the machine learning model, anomaly detection is performed on the extracted multimodal feature vectors;

[0031] Collect historical thermal image data and video surveillance images, including samples of normal and abnormal conditions, and perform temperature feature extraction and visual feature extraction on each sample to generate a multimodal feature vector dataset;

[0032] The multimodal feature vector data set is annotated to indicate whether each sample is normal or abnormal; the abnormal type and abnormal location are annotated for the abnormal samples, and the abnormal types include: abnormal temperature, abnormal flow of people, and equipment failure;

[0033] Divide the annotated multimodal feature vector dataset into a training set and a test set;

[0034] The neural network model is trained using the training set. During the training process, the generalization ability of the neural network model is evaluated by using cross-validation technology. The trained neural network model is evaluated using the test set to test whether the anomaly detection ability is normal. The real-time multimodal feature vector is input into the trained neural network model, and the neural network model performs anomaly detection and outputs real-time thermal image data and video surveillance images to see whether there are any abnormalities.

[0035] Preferably, the step S4 includes monitoring and warning the exhibition hall based on the abnormality detection results, including the following steps: when an abnormality is identified in the real-time thermal image data and video surveillance images, the type and location of the abnormality are output to generate warning information for warning, otherwise the exhibition hall continues to be monitored and detected in real time.

[0036] Preferably, the step S5 comprises the following steps:

[0037] According to the comprehensive evaluation results and anomaly detection results of the exhibition hall, the hot spots and abnormal areas in the exhibition hall are obtained;

[0038] Obtain the exhibition hall layout plan and abstract the exhibition hall layout into a graph model, in which the exhibition areas are nodes and the passages between the exhibition areas are edges; by using the shortest path algorithm Dijkstra algorithm on the graph model, calculate the shortest path from the starting point to the hot spot area and generate the best visiting route;

[0039] Based on the exhibition hall layout plan, a digital map is created; the best tour route generated by the Dijkstra algorithm is displayed on the digital map, and hot spots and abnormal areas are marked.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] The present invention can comprehensively and accurately reflect the real-time status of the exhibition hall by collecting environmental data, video surveillance images, audience behavior data and equipment operation data in the exhibition hall in real time, as well as thermal image data obtained by using a thermal imaging camera;

[0042] The present invention conducts a comprehensive evaluation of the exhibition hall by analyzing the environmental comfort index, the exhibit heat index and the equipment operation efficiency index; by using a neural network prediction model to analyze and predict the visitor flow and energy consumption demand, and performing intelligent scheduling control based on the prediction results, the operation efficiency and intelligent control of the exhibition hall can be improved;

[0043] The present invention combines thermal image data and video surveillance images for anomaly detection, and can promptly discover abnormal situations in the exhibition hall; through monitoring and early warning, early warning information can be sent in real time to improve the emergency response speed; by generating the best visiting route based on the comprehensive evaluation results and anomaly detection results of the exhibition hall, a more convenient and comfortable visiting experience can be provided for the audience. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 is a flow chart of the method of the present invention;

[0046] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0047] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] See also Figure 1 As shown, the first embodiment of the present invention provides an intelligent exhibition hall control method, comprising the following steps:

[0049] S1: Real-time collection of monitoring data in the exhibition hall, including: environmental data, video surveillance images, audience behavior data and equipment operation data; use thermal imaging cameras to collect thermal image data in the exhibition hall in real time;

[0050] S2: preprocessing the collected monitoring data including: denoising, data cleaning and standardization; preprocessing the collected thermal image data and video surveillance images including: denoising, calibration and image enhancement;

[0051] S3: Analyze and calculate the environmental comfort index, the heat index of the i-th exhibit and the equipment operation efficiency index based on the collected and pre-processed monitoring data; conduct a comprehensive evaluation of the exhibition hall based on the calculation results of the analysis of the monitoring data; establish a neural network prediction model based on the comprehensive evaluation results of the exhibition hall to conduct analysis and prediction, and perform intelligent scheduling and control of the exhibition hall based on the prediction results;

[0052] S4: Perform anomaly detection based on the pre-processed thermal image data combined with the video surveillance image; monitor and warn the exhibition hall based on the anomaly detection results;

[0053] S5: Generate the best tour route based on the comprehensive evaluation results and anomaly detection results of the exhibition hall.

[0054] Specifically, comprehensive information is ensured through the collection of multiple data such as environmental data, video surveillance, audience behavior, equipment operation and thermal imaging; pre-processing is carried out through denoising, cleaning, standardization and image denoising, contrast enhancement, edge information, etc.; comprehensive evaluation is carried out based on the environmental comfort index, exhibit heat index and equipment operation efficiency index, assisted by a neural network prediction model to achieve intelligent scheduling and control; anomaly detection and early warning mechanisms are implemented by combining thermal imaging and video surveillance; and finally, the best visiting route is dynamically planned based on the evaluation results and anomaly detection, to improve the visiting experience and the level of exhibit protection, and to achieve optimal resource allocation and intelligent management.

[0055] In this embodiment, the monitoring data in the exhibition hall is collected in real time in step S1, including: environmental data, video surveillance images, audience behavior data and equipment operation data, and includes the following steps: using sensors and surveillance cameras to collect monitoring data in the exhibition hall in real time, the environmental data includes: temperature, humidity, light intensity and air quality; the audience behavior data includes: audience flow, visiting path, length of stay and the number of times the exhibits are viewed or interacted with by the audience; the equipment operation data includes: equipment startup time, number of startups, usage frequency and energy consumption data.

[0056] Specifically, temperature and humidity sensors, light intensity sensors, air quality monitoring sensors and high-definition surveillance cameras are deployed at key locations in the exhibition hall, including the entrance, exhibition area, rest area, etc., to collect monitoring data in the exhibition hall in real time. The equipment in the exhibition hall includes lighting equipment, air conditioning equipment, interactive equipment, etc.

[0057] In this embodiment, the step S3 analyzes and calculates the environmental comfort index, the heat index of the i-th exhibit and the equipment operation efficiency index according to the collected and pre-processed monitoring data, including the following steps:

[0058] Calculate the environmental data scoring function formula:

[0059] Among them, x represents the environmental data in the exhibition hall collected in real time, x c represents the optimal value of the environmental data in the exhibition hall, and K represents the steepness parameter of the environmental data scoring function curve;

[0060] By substituting the real-time collected and pre-processed temperature, humidity, light intensity and air quality in the exhibition hall into the environmental data scoring function formula, the temperature score, humidity score, light intensity score and air quality score are obtained respectively, and the temperature score, humidity score, light intensity score and air quality score are weighted and summed to obtain the environmental comfort index E;

[0061] The formula for calculating the heat index of the i-th exhibit is:

[0062] Get the heat index P of the i-th exhibit i ; Among them, S i represents the number of times the i-th exhibit is viewed or interacted with by visitors, F i represents the visitor flow in the i-th exhibit area, F T represents the total visitor flow in the exhibition hall, T i represents the length of time visitors stay in the i-th exhibit area, T a represents the average length of stay of visitors in all exhibition areas; α, β and γ represent weight coefficients, and the sum of α, β and γ is 1;

[0063] By calculating the equipment operation efficiency index formula:

[0064] The equipment operation efficiency index η is obtained; where f represents the frequency of equipment use, t represents the total start-up time of all equipment in the exhibition hall, M represents the total number of startups of all equipment, N represents the actual energy consumption, and N std Indicates standard energy consumption.

[0065] Specifically, the optimal values ​​of the environmental data in the exhibition hall are determined, including: the optimal value of temperature is set at 22°C, the optimal value of humidity is set at 50%, the optimal value of light intensity is set at 500lx lux, and the optimal value of PM2.5 in air quality is set at 35μg / m 3 , the optimal value of carbon dioxide concentration is set to 800ppm; the steepness parameter of the environmental data scoring function curve is set to 10; the real-time collected environmental data, the optimal value of the environmental data and the steepness parameter are substituted into the environmental data scoring function formula to obtain the temperature score, humidity score, light intensity score and air quality score respectively; the temperature score, humidity score, light intensity score and air quality score are weighted and summed to obtain the environmental comfort index, where the weights of the temperature score, humidity score, light intensity score and air quality score are 0.3, 0.3, 0.2 and 0.2 respectively; the optimal value and weight are dynamically adjusted according to the actual exhibition hall and historical data. The weight coefficients in the formula of the heat index of the i-th exhibit are set to α, β and γ as 0.4, 0.3 and 0.3 respectively, and the weight coefficients are dynamically adjusted according to the actual exhibition hall situation; the heat index of the i-th exhibit is obtained by substituting the real-time collected audience behavior data into the formula. Obtain the standard energy consumption of the equipment. The standard energy consumption can be calculated based on the rated power and normal operating time of the equipment, or refer to the energy consumption standards provided by the equipment manufacturer, and substitute the real-time collected equipment operation data into the formula to obtain the equipment operation efficiency index.

[0066] In this embodiment, the step S3 performs a comprehensive evaluation on the exhibition hall according to the calculation results of the analysis monitoring data, including the following steps:

[0067] According to the calculation results of the analysis of monitoring data, the environmental comfort index, the heat index of the i-th exhibit and the equipment operation efficiency index are obtained, and the comprehensive evaluation index formula of the exhibition hall is calculated:

[0068] The comprehensive evaluation index Z of the exhibition hall is obtained, where W1, W2 and W3 represent weight coefficients respectively; E represents the environmental comfort index, P irepresents the heat index of the i-th exhibit, n represents the number of exhibits in the exhibition hall, i∈{1, 2, ..., n}, and η represents the equipment operation efficiency index.

[0069] Specifically, W1, W2 and W3 represent weight coefficients of 0.3, 0.4 and 0.3 respectively, and the weight coefficients are dynamically adjusted according to the actual exhibition hall conditions. The environmental comfort index, the heat index of the i-th exhibit and the equipment operation efficiency index obtained by the calculation results of the analysis of the monitoring data are substituted into the formula to obtain the comprehensive evaluation index of the exhibition hall.

[0070] In this embodiment, the step S3 establishes a neural network prediction model according to the comprehensive evaluation results of the exhibition hall to perform analysis and prediction, and performs intelligent scheduling and control of the exhibition hall according to the prediction results, including the following steps:

[0071] Based on the comprehensive evaluation results of the exhibition hall, the neural network prediction model in the machine learning algorithm is used to predict the visitor flow and equipment energy consumption requirements;

[0072] Collect the historical monitoring data of the exhibition hall, and pre-process the historical monitoring data including: data cleaning and normalization; according to the processed historical monitoring data, extract the characteristic parameters including: environmental comfort index, the heat index of the i-th exhibit, the equipment operation efficiency index and the comprehensive evaluation index of the exhibition hall;

[0073] Input the characteristic parameters into the neural network prediction model for training; according to the trained neural network prediction model, by inputting the monitoring data collected and processed in real time, the neural network prediction model predicts the audience flow and equipment energy consumption requirements;

[0074] According to the prediction results, the exhibition hall is intelligently dispatched and controlled, and the dispatching includes: environmental data dispatching, exhibit display dispatching and equipment energy consumption dispatching.

[0075] Specifically, the audience flow, environmental parameters, and equipment operation data in the past period of time are collected from the monitoring system of the exhibition hall. Based on the collected and preprocessed historical data, the comprehensive evaluation index of the exhibition hall is obtained through the comprehensive evaluation index formula of the exhibition hall. Based on the preprocessed historical monitoring data, feature parameters are extracted. A neural network model suitable for time series prediction, such as a multi-layer perceptron or a long short-term memory network, is selected. The extracted feature parameters are used as input layer neurons. An appropriate number of hidden layers and neurons are set, and the weights and biases are adjusted through the back propagation algorithm so that the model can learn the complex relationships in the data. The output layer neurons correspond to the predicted audience flow and equipment energy consumption requirements. The model is trained using the processed historical monitoring data, and the model parameters are optimized through cross-validation and other methods to prevent overfitting. The audience flow, environmental parameters, equipment operation data, etc. of the exhibition hall are collected in real time through sensors and monitoring systems, and preprocessed. The real-time processed data is input into the trained neural network prediction model to predict future audience flow and equipment energy consumption requirements. According to the prediction results, the environmental parameters in the exhibition hall are adjusted, such as automatically adjusting the air conditioning temperature and humidity, and optimizing the lighting system to provide a comfortable visiting environment. According to the exhibition popularity index and predicted visitor flow, the exhibition display strategy is dynamically adjusted, such as adding interactive links for popular exhibits or adjusting the exhibition layout to guide the flow of people. Based on the predicted energy consumption demand of equipment, energy allocation is optimized, such as reducing the operating power of equipment during the period when the visitor flow is expected to be low, to achieve energy conservation and emission reduction.

[0076] In this embodiment, the abnormality detection in step S4 is performed based on the preprocessed thermal image data combined with the video surveillance image, including the following steps:

[0077] Analyze the pre-processed thermal image data through image processing algorithms to identify hot spots in the thermal image data; convert the temperature information of the hot spot area into numerical form through color mapping or gray value conversion methods; smooth the temperature distribution map using interpolation algorithms or spatial filtering technology; extract temperature features based on the smoothed temperature distribution map, including: maximum temperature, average temperature, temperature gradient, number of hot spots and hot spot area;

[0078] Extract visual features from the preprocessed video surveillance images using image processing algorithms, including geometric moments, color moments, and motion trajectories; perform multimodal feature fusion on visual features and temperature features to generate multimodal feature vectors;

[0079] By utilizing the neural network model in the machine learning model, anomaly detection is performed on the extracted multimodal feature vectors;

[0080] Collect historical thermal image data and video surveillance images, including samples of normal and abnormal conditions, and perform temperature feature extraction and visual feature extraction on each sample to generate a multimodal feature vector dataset;

[0081] The multimodal feature vector data set is annotated to indicate whether each sample is normal or abnormal; the abnormal type and abnormal location are annotated for the abnormal samples, and the abnormal types include: abnormal temperature, abnormal flow of people, and equipment failure;

[0082] Divide the annotated multimodal feature vector dataset into a training set and a test set;

[0083] The neural network model is trained using the training set. During the training process, the generalization ability of the neural network model is evaluated by using cross-validation technology. The trained neural network model is evaluated using the test set to test whether the anomaly detection ability is normal. The real-time multimodal feature vector is input into the trained neural network model, and the neural network model performs anomaly detection and outputs real-time thermal image data and video surveillance images to see whether there are any abnormalities.

[0084] Specifically, the pre-processed thermal image data is analyzed by image processing algorithms including edge detection, threshold segmentation, etc., to identify the hotspot area with higher temperature; the temperature information of the hotspot area is converted into numerical form by color mapping or gray value conversion method. Color mapping can convert the color of the hotspot area into the corresponding temperature value according to the preset color-temperature correspondence relationship; gray value conversion can be converted according to the correspondence between gray level and temperature. The temperature distribution map is smoothed by using interpolation algorithm or spatial filtering technology to reduce the influence of noise and outliers; temperature features are extracted from the smoothed temperature distribution map, including the highest temperature, average temperature, temperature gradient, number of hotspots and hotspot area. These features can fully reflect the temperature distribution and change of the hotspot area. According to the pre-processed video surveillance image, the image processing algorithms including shape analysis, color space transformation, motion estimation, etc. are used to extract visual features, including geometric moments, color moments and motion trajectories, etc. These features can describe the shape, color and motion state of objects in the monitoring scene. The extracted visual features and temperature features are multi-modal feature fusion to generate a multi-modal feature vector. Multimodal feature fusion can use feature-level fusion methods, including splicing or weighted averaging, to combine features of different modalities. Collect historical thermal image data and video surveillance images, including samples of normal and abnormal conditions. Perform temperature feature extraction and visual feature extraction on each sample, and generate a multimodal feature vector data set. Label the multimodal feature vector data set to indicate whether each sample is normal or abnormal. For samples of abnormal conditions, further label the abnormal type and abnormal location. Abnormal types include but are not limited to abnormal temperature, abnormal flow of people, and equipment failure. Divide the labeled multimodal feature vector data set into a training set and a test set. The training set is used to train the neural network model, and the test set is used to evaluate the model's anomaly detection capability. Use the training set to train the neural network model. During the training process, improve the generalization ability of the model by adjusting the model parameters and optimizer settings. At the same time, use cross-validation technology to evaluate the performance of the model to avoid overfitting and underfitting. Use the test set to evaluate the trained neural network model to test whether its anomaly detection capability is normal. The real-time multimodal feature vector is input into the trained neural network model, which performs anomaly detection and outputs real-time thermal image data and video surveillance images to determine whether there are any anomalies.

[0085] In this embodiment, the exhibition hall is monitored and warned according to the abnormality detection results in step S4, including the following steps: when an abnormality is identified in the real-time thermal image data and video surveillance images, the type and location of the abnormality are output to generate warning information for warning, otherwise the exhibition hall continues to be monitored and detected in real time.

[0086] Specifically, when the anomaly detection algorithm identifies an abnormality in the thermal image data or video surveillance image, the early warning mechanism is immediately triggered, and the early warning information is automatically generated based on the identified abnormality. The early warning information should include the type of abnormality, such as abnormal temperature, abnormal behavior, abnormal object, etc., and the location, such as which area of ​​the exhibition hall, and other key information. After outputting the early warning information, the exhibition hall continues to be monitored and tested in real time. The purpose of real-time monitoring and testing is to ensure that the safety status in the exhibition hall is continuously monitored so that new abnormalities can be discovered and handled in a timely manner.

[0087] In this embodiment, step S5 includes the following steps:

[0088] According to the comprehensive evaluation results and anomaly detection results of the exhibition hall, the hot spots and abnormal areas in the exhibition hall are obtained;

[0089] Obtain the exhibition hall layout plan and abstract the exhibition hall layout into a graph model, in which the exhibition areas are nodes and the passages between the exhibition areas are edges; by using the shortest path algorithm Dijkstra algorithm on the graph model, calculate the shortest path from the starting point to the hot spot area and generate the best visiting route;

[0090] Based on the exhibition hall layout plan, a digital map is created; the best tour route generated by the Dijkstra algorithm is displayed on the digital map, and hot spots and abnormal areas are marked.

[0091] Specifically, by collecting and analyzing multi-dimensional data such as visitor flow, exhibit attraction, and environmental comfort of each exhibition area in the exhibition hall, a comprehensive evaluation result of the exhibition hall is obtained, so as to identify the hot spots in the exhibition hall, that is, the exhibition areas with large visitor flow and strong exhibit attraction. By performing anomaly detection on the exhibition hall in real time, the abnormal area is determined. The latest exhibition hall layout plan is obtained from the exhibition hall management department to ensure the accuracy and timeliness of the drawing information. The exhibition hall layout plan is abstracted into a graph model, in which the exhibition area is used as a node and the passages between the exhibition areas are used as edges. According to the actual situation of the exhibition hall, one or more starting points are set, which are usually the entrances or main passages for visitors to enter the exhibition hall. On the graph model, a distance value is set for each node, that is, the exhibition area, indicating the current shortest path length from the starting point to the node. Initially, the distance value of the starting point is 0, and the distance values ​​of other nodes are infinite. At the same time, an empty set of visited nodes is created. The node closest to the starting point is selected from the unvisited nodes and added to the set of visited nodes. For each neighbor node of the visited node, check whether a shorter path can be obtained through the currently visited node. If possible, update the distance value of the neighbor node and record the predecessor node of the shortest path. Repeat the above steps until all nodes are visited or no shorter path can be found. According to the results calculated by the Dijkstra algorithm, generate the shortest path from the starting point to each hot spot area as the best tour route. Based on the exhibition hall layout plan, use the geographic information system GIS or related software tools to create a digital map. The digital map should be able to accurately reflect the layout of the exhibition hall, the location of the exhibition area and the direction of the channel. On the digital map, the best tour route generated by the Dijkstra algorithm is displayed in the form of lines or arrows. At the same time, mark the hot spots and abnormal areas so that the audience and managers can intuitively understand the situation in the exhibition hall.

[0092] See also Figure 2 As shown, the present invention is an intelligent exhibition hall control system, comprising the following modules:

[0093] Data collection module: real-time collection of monitoring data in the exhibition hall, including: environmental data, video surveillance images, audience behavior data and equipment operation data; use thermal imaging cameras to collect thermal image data in the exhibition hall in real time;

[0094] Data preprocessing module: preprocess the collected monitoring data including: denoising, data cleaning and standardization; preprocess the collected thermal image data and video surveillance images including: denoising, calibration and image enhancement;

[0095] Data analysis module: Analyze and calculate the environmental comfort index, the heat index of the i-th exhibit and the equipment operation efficiency index based on the collected and pre-processed monitoring data; conduct a comprehensive evaluation of the exhibition hall based on the calculation results of the analysis of the monitoring data; establish a neural network prediction model based on the comprehensive evaluation results of the exhibition hall to conduct analysis and prediction, and perform intelligent scheduling and control of the exhibition hall based on the prediction results;

[0096] Monitoring and early warning module: perform anomaly detection based on pre-processed thermal image data combined with video surveillance images; monitor and warn the exhibition hall based on the anomaly detection results;

[0097] Route planning module: Generates the best visiting route based on the comprehensive evaluation results and anomaly detection results of the exhibition hall.

[0098] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An intelligent exhibition hall control method, characterized in that: The following steps are involved: S1: Real-time collection of monitoring data in the exhibition hall, including: environmental data, video surveillance images, audience behavior data and equipment operation data; use thermal imaging cameras to collect thermal image data in the exhibition hall in real time; S2: preprocessing the collected monitoring data including: denoising, data cleaning and standardization; preprocessing the collected thermal image data and video surveillance images including: denoising, calibration and image enhancement; S3: Analyze and calculate the environmental comfort index, the heat index of the i-th exhibit and the equipment operation efficiency index based on the collected and pre-processed monitoring data; conduct a comprehensive evaluation of the exhibition hall based on the calculation results of the analysis of the monitoring data; establish a neural network prediction model based on the comprehensive evaluation results of the exhibition hall to conduct analysis and prediction, and perform intelligent scheduling and control of the exhibition hall based on the prediction results; S4: Perform anomaly detection based on the pre-processed thermal image data combined with the video surveillance image; monitor and warn the exhibition hall based on the anomaly detection results; S5: Generate the best tour route based on the comprehensive evaluation results and anomaly detection results of the exhibition hall.

2. The intelligent exhibition hall control method according to claim 1, characterized in that: The monitoring data in the exhibition hall in the step S1 is collected in real time, including: environmental data, video surveillance images, audience behavior data and equipment operation data, including the following steps: using sensors and surveillance cameras to collect the monitoring data in the exhibition hall in real time, the environmental data includes: temperature, humidity, light intensity and air quality; the audience behavior data includes: audience flow, visiting path, length of stay and the number of times the exhibits are viewed or interacted with by the audience; the equipment operation data includes: equipment startup time, number of startups, usage frequency and energy consumption data.

3. The intelligent exhibition hall control method according to claim 1, characterized in that: The step S3 includes the following steps: analyzing and calculating the environmental comfort index, the heat index of the i-th exhibit, and the equipment operation efficiency index based on the collected and pre-processed monitoring data: Calculate the environmental data scoring function formula: Among them, x represents the environmental data in the exhibition hall collected in real time, x c represents the optimal value of the environmental data in the exhibition hall, and K represents the steepness parameter of the environmental data scoring function curve; By substituting the real-time collected and pre-processed temperature, humidity, light intensity and air quality in the exhibition hall into the environmental data scoring function formula, the temperature score, humidity score, light intensity score and air quality score are obtained respectively, and the temperature score, humidity score, light intensity score and air quality score are weighted and summed to obtain the environmental comfort index E; The formula for calculating the heat index of the i-th exhibit is: Get the heat index P of the i-th exhibit i ; Among them, S i represents the number of times the i-th exhibit is viewed or interacted with by visitors, F i represents the visitor flow in the i-th exhibit area, F T represents the total visitor flow in the exhibition hall, T i represents the length of time visitors stay in the i-th exhibit area, T a represents the average length of stay of visitors in all exhibition areas; α, β and γ represent weight coefficients, and the sum of α, β and γ is 1; By calculating the equipment operation efficiency index formula: The equipment operation efficiency index η is obtained; where f represents the frequency of equipment use, t represents the total start-up time of all equipment in the exhibition hall, M represents the total number of startups of all equipment, N represents the actual energy consumption, and N std Indicates standard energy consumption.

4. The intelligent exhibition hall control method according to claim 3 is characterized in that: In step S3, a comprehensive evaluation of the exhibition hall is performed based on the calculation results of the analysis monitoring data, including the following steps: According to the calculation results of the analysis of monitoring data, the environmental comfort index, the heat index of the i-th exhibit and the equipment operation efficiency index are obtained, and the comprehensive evaluation index formula of the exhibition hall is calculated: The comprehensive evaluation index Z of the exhibition hall is obtained, where W1, W2 and W3 represent weight coefficients respectively; E represents the environmental comfort index, P i represents the heat index of the i-th exhibit, n represents the number of exhibits in the exhibition hall, i∈{1,2,...,n}, and η represents the equipment operation efficiency index.

5. The intelligent exhibition hall control method according to claim 4, characterized in that: In step S3, a neural network prediction model is established according to the comprehensive evaluation results of the exhibition hall to perform analysis and prediction, and intelligent scheduling and control of the exhibition hall is performed according to the prediction results, including the following steps: Based on the comprehensive evaluation results of the exhibition hall, the neural network prediction model in the machine learning algorithm is used to predict the visitor flow and equipment energy consumption requirements; Collect the historical monitoring data of the exhibition hall, and pre-process the historical monitoring data including: data cleaning and normalization; according to the processed historical monitoring data, extract the characteristic parameters including: environmental comfort index, the heat index of the i-th exhibit, the equipment operation efficiency index and the comprehensive evaluation index of the exhibition hall; Input the characteristic parameters into the neural network prediction model for training; according to the trained neural network prediction model, by inputting the monitoring data collected and processed in real time, the neural network prediction model predicts the audience flow and equipment energy consumption requirements; According to the prediction results, the exhibition hall is intelligently dispatched and controlled, and the dispatching includes: environmental data dispatching, exhibit display dispatching and equipment energy consumption dispatching.

6. The intelligent exhibition hall control method according to claim 1, characterized in that: In step S4, abnormality detection is performed based on the preprocessed thermal image data combined with the video surveillance image, including the following steps: Analyze the pre-processed thermal image data through image processing algorithms to identify hot spots in the thermal image data; convert the temperature information of the hot spot area into numerical form through color mapping or gray value conversion methods; smooth the temperature distribution map using interpolation algorithms or spatial filtering technology; extract temperature features based on the smoothed temperature distribution map, including: maximum temperature, average temperature, temperature gradient, number of hot spots and hot spot area; Extract visual features from the preprocessed video surveillance images using image processing algorithms, including geometric moments, color moments, and motion trajectories; perform multimodal feature fusion on visual features and temperature features to generate multimodal feature vectors; By utilizing the neural network model in the machine learning model, anomaly detection is performed on the extracted multimodal feature vectors; Collect historical thermal image data and video surveillance images, including samples of normal and abnormal conditions, and perform temperature feature extraction and visual feature extraction on each sample to generate a multimodal feature vector dataset; The multimodal feature vector data set is annotated to indicate whether each sample is normal or abnormal; the abnormal type and abnormal location are annotated for the abnormal samples, and the abnormal types include: abnormal temperature, abnormal flow of people, and equipment failure; Divide the annotated multimodal feature vector dataset into a training set and a test set; The neural network model is trained using the training set. During the training process, the generalization ability of the neural network model is evaluated by using cross-validation technology. The trained neural network model is evaluated using the test set to test whether the anomaly detection ability is normal. The real-time multimodal feature vector is input into the trained neural network model, and the neural network model performs anomaly detection and outputs real-time thermal image data and video surveillance images to see whether there are any abnormalities.

7. The intelligent exhibition hall control method according to claim 6, characterized in that: In step S4, the exhibition hall is monitored and warned according to the abnormality detection results, including the following steps: when an abnormality is identified in the real-time thermal image data and video surveillance images, the type and location of the abnormality are output to generate warning information for warning, otherwise the exhibition hall continues to be monitored and detected in real time.

8. The intelligent exhibition hall control method according to claim 1, characterized in that: The step S5 comprises the following steps: According to the comprehensive evaluation results and anomaly detection results of the exhibition hall, the hot spots and abnormal areas in the exhibition hall are obtained; Obtain the exhibition hall layout plan and abstract the exhibition hall layout into a graph model, in which the exhibition areas are nodes and the passages between the exhibition areas are edges; by using the shortest path algorithm Dijkstra algorithm on the graph model, calculate the shortest path from the starting point to the hot spot area and generate the best visiting route; Based on the exhibition hall layout plan, a digital map is created; the best tour route generated by the Dijkstra algorithm is displayed on the digital map, and hot spots and abnormal areas are marked.

9. An intelligent exhibition hall control system, using an intelligent exhibition hall control method as claimed in any one of claims 1 to 8, characterized in that: Includes the following modules: Data collection module: real-time collection of monitoring data in the exhibition hall, including: environmental data, video surveillance images, audience behavior data and equipment operation data; use thermal imaging cameras to collect thermal image data in the exhibition hall in real time; Data preprocessing module: preprocess the collected monitoring data including: denoising, data cleaning and standardization; preprocess the collected thermal image data and video surveillance images including: denoising, calibration and image enhancement; Data analysis module: Analyze and calculate the environmental comfort index, the heat index of the i-th exhibit and the equipment operation efficiency index based on the collected and pre-processed monitoring data; conduct a comprehensive evaluation of the exhibition hall based on the calculation results of the analysis of the monitoring data; establish a neural network prediction model based on the comprehensive evaluation results of the exhibition hall to conduct analysis and prediction, and perform intelligent scheduling and control of the exhibition hall based on the prediction results; Monitoring and early warning module: perform anomaly detection based on pre-processed thermal image data combined with video surveillance images; monitor and warn the exhibition hall based on the anomaly detection results; Route planning module: Generates the best visiting route based on the comprehensive evaluation results and anomaly detection results of the exhibition hall.

Images