Tunnel construction intelligent supervision system and method based on intelligent safety helmet

Abstract

The invention relates to the technical field of tunnel construction intelligent supervision, in particular to a tunnel construction intelligent supervision system and method based on an intelligent safety helmet. The sensing unit is used for collecting data of positions and working environments of constructors; the communication unit is used for transmitting the collected data through a wireless communication protocol; the edge calculation unit performs edge calculation on the data transmitted by the communication unit for real-time analysis, analyzes positioning information through a centroid method, and introduces personnel density and material stacking weight to optimize calculation of centroid coordinates, so as to reduce the influence of signal change in positioning information calculation, the edge calculation unit comprises a trajectory prediction module, a contour construction module and a collision early warning module. According to the design, positioning information is analyzed through a centroid method, the positions of constructors can be tracked, and each worker can be rapidly positioned especially in a complex and changeable tunnel environment.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent supervision of tunnel construction, and in particular to an intelligent supervision system and method for tunnel construction based on an intelligent safety helmet. Background Art

[0002] The tunnel construction environment is complex and changeable, and there are many safety hazards, such as collapse, gas explosion, rock burst, etc., which pose a threat to the personal safety of workers; there are many people on the construction site, and it is difficult to grasp the location and status of each employee in real time, which makes it difficult to evacuate people in an emergency; the traditional management method relies on manual recording and monitoring, which is not only inefficient, but also prone to omissions; the information transmission on the construction site is often not timely and accurate, making it difficult for management to make effective decisions quickly; environmental factors such as gas concentration, temperature and humidity inside the tunnel need to be continuously monitored to ensure the safety of the working environment. Therefore, an intelligent supervision system and method for tunnel construction based on a smart helmet is provided. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent supervision system and method for tunnel construction based on an intelligent safety helmet, so as to solve the problems of high safety risks, difficult personnel management, low work efficiency and insufficient environmental monitoring proposed in the above background technology.

[0004] To achieve the above objectives, on the one hand, the present invention aims to provide an intelligent monitoring system for tunnel construction based on an intelligent helmet, comprising: A sensor unit, which is used to collect data on the location of construction personnel and working environment; A communication unit, the communication unit is used to transmit the collected data via a wireless communication protocol; An edge computing unit, wherein the edge computing unit performs edge computing on the data transmitted by the communication unit for real-time analysis, analyzes the positioning information by the centroid method, and introduces the weight of personnel density and material stacking to optimize the calculation of the centroid coordinates, so as to reduce the influence of signal changes in the calculation of the positioning information; wherein the edge computing unit includes a trajectory prediction module, a contour construction module, and a collision warning module; The trajectory prediction module uses a particle filtering method to predict the movement trajectory of construction personnel; The contour construction module constructs an environment contour including positioning information based on a clustering algorithm; The collision warning module integrates the movement trajectory and the environment contour, and performs collision warning based on the collision time model.

[0005] As a further improvement of the technical solution, the sensing unit includes a positioning module and an image acquisition module; Among them, the positioning module is used to track and determine the position of the smart helmet; The image acquisition module is used to monitor the environmental conditions of the construction environment.

[0006] As a further improvement of the technical solution, the edge computing unit analyzes the positioning information by the centroid method, including the following steps: S1.1. Deploy multiple reference points with known positions m in the tunnel, each of which has a known coordinate position; S1.2. Use the smart helmet to measure the value; S1.3. Utilization The relationship between the value and the distance is used to estimate the approximate distance between the smart helmet and the reference point; There is an exponential decay relationship with distance: ; in, Indicates the received signal strength indicator; is the reference point at a distance Where value; represents the path loss exponent; represents the estimated distance between the smart helmet and the reference point; Indicates the reference distance, which is 1 meter; S1.4. Use the positions of all reference points and the corresponding estimated distances to calculate the center of mass coordinates of the smart helmet, and introduce the weights of personnel density and material stacking to optimize the calculation of the center of mass coordinates, so as to reduce the impact of signal propagation changes caused by high-density personnel and material stacking on the accuracy of positioning calculations.

[0007] As a further improvement of the technical solution, in S1.4, the weight of personnel density and material stacking is introduced to optimize the calculation of the centroid coordinates as follows: ; in, Indicates The weight after introducing the density of people and the stacking of materials at each reference point is: ; Indicates that the smart helmet is The estimated distance between reference points; Indicates The density factor of people near a reference point; Indicates Material stacking factor near a reference point; , , Indicates The coordinates of the reference points; Represents the coordinates of the center of mass of the smart helmet; Indicates the number of reference points.

[0008] As a further improvement of the technical solution, the trajectory prediction module predicts the movement trajectory of the construction personnel, including the following steps: S2.1. Collect the location coordinates of the construction workers, record the timestamp of each location data, and collect the speed and acceleration data of the construction workers; S2.2, preprocess the data; S2.3. Construct a time series containing the position coordinates of the construction workers at each time point, calculate the speed of the construction workers, and calculate the acceleration of the construction workers by dividing the distance between adjacent position coordinates by the corresponding time interval; S2.4. Analyze the movement trends of construction personnel, including straight-line movement and turning; S2.5. Use the particle filtering method to estimate the location probability distribution of construction workers and predict their movement trajectories.

[0009] As a further improvement of the technical solution, the particle filtering method is used in S2.5 to estimate the position probability distribution of the construction personnel and predict the movement trajectory, including the following steps: S2.51. Set an initial particle set, each particle contains the location of the construction worker , as well as direction and velocity, where the particles represent probabilistic estimates of the construction worker’s location; S2.52. Update the position of each particle using the state transition model: ; ; ; in, Represents particles exist Moment Position coordinates of the direction; Represents particles exist Moment Position coordinates of the direction; Represents particles exist Moment Position coordinates of the direction; Represents particles At the previous moment of Position coordinates of the direction; Represents particles At the previous moment of Position coordinates of the direction; Represents particles At the previous moment of Position coordinates of the direction; Represents particles exist Direction speed; Represents particles exist Direction speed; Represents particles exist Direction speed; Indicates a time interval; Indicates Particles in The random noise in the direction; Indicates Particles in The random noise in the direction; Indicates Particles in The random noise in the direction; S2.53. Use the Gaussian distribution observation model to weight each particle: ; in, Indicated in Moment Particle The weight of is the actual observed value, is the observed value predicted from the particle state, is the standard deviation of the observation noise; S2.54, resampling according to the weight of the particles; S2.55. Repeat steps S2.52-S2.54. After each step, the expected position is calculated based on the positions of all particles as the current best estimate, thereby forming a predicted trajectory.

[0010] As a further improvement of the technical solution, the contour construction module constructs an environment contour including positioning information, comprising the following steps: S3.1. The entrance of the tunnel is the origin of the coordinate system. The direction along the extension of the tunnel is the positive x-axis, the direction perpendicular to the main axis of the tunnel is the y-axis, and the direction perpendicular to the ground is the z-axis. S3.2, mapping the position information of all objects into a three-dimensional coordinate system; S3.3, classify the data by object type, including smart helmets, vehicles, and building materials; S3.4. Use clustering algorithm to group each type of object, and introduce the vibration intensity parameters generated by heavy machinery to optimize the clustering algorithm to reduce the impact of vibration on the data, thereby improving the accuracy and reliability of clustering results. For each clustering result, calculate their boundaries; use the convex hull method to approximate the outline of each cluster, and merge the outlines of all objects into a large outline to represent the scope of the entire construction environment; S3.5. Use the 3D visualization tool Matplotlib to display the position and outline of each object.

[0011] As a further improvement of the technical solution, in S3.4, the vibration intensity parameters generated by heavy machinery are introduced to optimize the clustering algorithm as follows: ; in, Indicates The cluster centers of the clusters; Indicates data points, of which , Indicates The location coordinates of the data points, Indicates the vibration intensity; Indicates The set of all data points of a cluster; Indicates The weight of the data point is used to reflect the impact of vibration intensity on the cluster center; , A tuning parameter that indicates the degree of influence of vibration intensity on weight.

[0012] As a further improvement of the technical solution, the collision warning module performs collision warning, including the following steps: S4.1. Update the environment contour based on real-time data, especially for moving objects, including construction workers and vehicles, detect the position changes of dynamic obstacles and update their contours; S4.2. Use a collision time model to assess potential collision risks, predict the proximity between objects based on distance, speed, and acceleration information, and trigger a collision warning when the distance between objects is less than a safety threshold k; S4.3. When a potential collision risk is detected, a warning signal is sent to relevant personnel.

[0013] On the other hand, the present invention provides a method for intelligent supervision of tunnel construction based on a smart helmet, which is used in the above-mentioned intelligent supervision system for tunnel construction based on a smart helmet, and comprises the following steps: S5.1. Collect data on the location and working environment of construction workers; S5.2, transmitting the collected data via a wireless communication protocol; S5.3. Perform edge computing on the transmitted data for real-time analysis; analyze the positioning information through the centroid method to obtain accurate location information; use the particle filter method to predict the movement trajectory of construction workers; construct an environmental contour including positioning information based on the clustering algorithm; fuse the movement trajectory and environmental contour, and perform collision warning based on the collision time model.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. In the intelligent supervision system and method for tunnel construction based on smart safety helmets, the location of construction personnel can be tracked by analyzing positioning information through the centroid method, especially in a complex and changeable tunnel environment, each worker can be quickly located; the centroid method determines the location of the target through multiple reference points (such as base stations). Even if a reference point fails or the signal is unstable, it can be located through the data of other reference points, thereby improving the stability and accuracy of the system; signal reflection and multipath effects in the tunnel environment may cause positioning errors, and the centroid method can reduce the influence of these factors by increasing the number of reference points. Due to the flexibility of the centroid method, the position and number of reference points can be adjusted according to actual needs to adapt to different tunnel construction scenarios.

[0015] 2. In the intelligent supervision system and method for tunnel construction based on smart safety helmets, a particle filtering method is used to predict the movement trajectory of construction workers. The particle filtering method can automatically adapt to model and observation errors through a resampling mechanism, and can maintain good robustness even in a complex and changeable tunnel construction environment. The particle filtering method can handle missing data and incomplete information, making the prediction results more reliable. It can quickly adapt to the movement changes of construction workers by updating the particle set in real time; it can automatically identify different areas in the environment, such as construction areas, rest areas, dangerous areas, etc. through clustering algorithms, which helps to simplify the process of environmental modeling and can dynamically update the environmental contour as the construction progress changes, thereby adapting to the ever-changing construction environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flowchart of the overall process of the present invention.

[0017] The meaning of each number in the figure is: 1. Sensing unit; 2. Communication unit; 3. Edge computing unit. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not 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.

[0019] Example 1: Please refer to Figure 1 As shown, a tunnel construction intelligent supervision system based on an intelligent helmet is provided, comprising: Sensor unit 1, which is used to collect data on the location and working environment of construction personnel; The construction personnel position includes the coordinates of the construction personnel, and the working environment includes the environment contour; In this embodiment, the sensing unit 1 includes a positioning module and an image acquisition module; Among them, the positioning module is used to track and determine the position of the smart helmet; The image acquisition module is used to monitor the environmental conditions of the construction environment.

[0020] Communication unit 2, communication unit 2 is used to transmit the collected data via a wireless communication protocol; The edge computing unit 3 performs edge computing on the data transmitted by the communication unit 2 for real-time analysis, analyzes the positioning information by the centroid method, and introduces the weight of the personnel density and the material stacking to optimize the calculation of the centroid coordinates, so as to reduce the influence of the signal change in the positioning information calculation; wherein, the edge computing unit 3 includes a trajectory prediction module, a contour construction module, and a collision warning module; The trajectory prediction module uses a particle filtering method to predict the movement trajectory of construction personnel; The contour construction module constructs an environment contour including positioning information based on a clustering algorithm; The collision warning module integrates the movement trajectory and the environment contour, and performs collision warning based on the collision time model.

[0021] In this embodiment, the calculation process of the centroid method is relatively simple. It only needs to perform weighted averaging on the sensor data to obtain the final position estimate. It fuses the data of multiple sensors by weighted averaging, which can effectively reduce the impact of single sensor error on the overall positioning accuracy. The centroid method can easily integrate data from multiple sensors, which is very beneficial for improving positioning accuracy, especially in complex environments such as tunnel construction. Multi-sensor fusion can provide more comprehensive location information. The centroid method can quickly adapt to environmental changes and optimize the positioning results by adjusting weights. The edge computing unit 3 analyzes the positioning information by the centroid method, including the following steps: S1.1. Deploy multiple reference points with known positions m in the tunnel, each of which has a known coordinate position; S1.2. Use the smart helmet to measure the value; The value reflects the signal strength and can indirectly reflect the distance between the smart helmet and the reference point; S1.3. Utilization The relationship between the value and the distance is used to estimate the approximate distance between the smart helmet and the reference point; There is an exponential decay relationship with distance: ; in, Indicates the received signal strength indicator, i.e. value; is the reference point at a distance Where value; Represents the path loss index, which is related to the environment; represents the estimated distance between the smart helmet and the reference point; Indicates the reference distance, which is 1 meter; S1.4. Use the positions of all reference points and the corresponding estimated distances to calculate the centroid coordinates of the smart helmet, i.e., the location information of the smart helmet; and introduce the weights of personnel density and material stacking to optimize the calculation of the centroid coordinates, so as to reduce the impact of signal propagation changes caused by high-density personnel and material stacking on the accuracy of positioning calculation; Among them, in densely populated areas, the signal may be more interfered with. By increasing the weights of these areas, the impact of signal changes can be better reflected and the positioning accuracy can be improved. In material stacking areas, the signal may be blocked or reflected. By increasing the weights of these areas, the positioning error caused by this can be reduced. The weights can be dynamically adjusted according to the on-site conditions, which helps to maintain the stability and reliability of the positioning system in a changing construction environment. In areas with high population density or more material stacking, by increasing the weights, the impact of outliers on the positioning results can be reduced. The weights of population density and material stacking are introduced to optimize the calculation of the centroid coordinates. Specifically: ; in, Indicates The weight after introducing the density of people and the stacking of materials at each reference point is: ; Indicates that the smart helmet is The estimated distance between reference points; Indicates The density factor of people near a reference point; Indicates Material stacking factor near a reference point; , , Indicates The coordinates of the reference points; Represents the coordinates of the center of mass of the smart helmet; Indicates the number of reference points.

[0022] Furthermore, the trajectory prediction module predicts the movement trajectory of the construction personnel, including the following steps: S2.1. Collect the location coordinates of the construction workers, record the timestamp of each location data, and collect the speed and acceleration data of the construction workers; S2.2. Preprocess the data to remove invalid or abnormal data points. If there is data loss, interpolation methods can be used to fill in the missing values; S2.3. Construct a time series containing the position coordinates of the construction workers at each time point, calculate the speed of the construction workers, calculate the acceleration of the construction workers by dividing the distance between adjacent position coordinates by the corresponding time interval, and calculate the acceleration by the rate of change of the speed; S2.4. Analyze the movement trends of construction personnel, including straight-line movement and turning; S2.5. Use particle filtering method to estimate the location probability distribution of construction workers and predict their movement trajectory; Among them, the particle filter method is used to estimate the probability distribution of the position of the construction workers and predict the movement trajectory. The particle filter method is very suitable for processing nonlinear dynamic systems, which is especially important in the construction environment because the movement pattern of the construction workers is often nonlinear. The particle filter can automatically adapt to the model and observation errors through the resampling mechanism and has good robustness. The particle filter method can adapt to changes in the environment. For example, the movement pattern of the construction workers changes with time and tasks. The particle filter can naturally handle uncertainty and express the uncertainty of the state through the distribution of particles. The following steps are included: S2.51. Set an initial particle set, each particle contains the location of the construction worker , as well as direction and velocity, where the particles represent probabilistic estimates of the construction worker’s location; S2.52. Update the position of each particle using the state transition model: ; ; ; in, Represents particles exist Moment Position coordinates of the direction; Represents particles exist Moment Position coordinates of the direction; Represents particles exist Moment Position coordinates of the direction; Represents particles At the previous moment of Position coordinates of the direction; Represents particles At the previous moment of Position coordinates of the direction; Represents particles At the previous moment of Position coordinates of the direction; Represents particles exist Direction speed; Represents particles exist Direction speed; Represents particles exist Direction speed; Indicates a time interval; Indicates Particles in The random noise in the direction; Indicates Particles in The random noise in the direction; Indicates Particles in The random noise in the direction; S2.53. Use the Gaussian distribution observation model to weight each particle. The weight depends on the match between the observation data and the particle position: ; in, Indicated in Moment Particle The weight of is the actual observed value, is the observed value predicted from the particle state, is the standard deviation of the observation noise; S2.54, resampling is performed according to the weight of the particles to reduce particle degradation; S2.55. Repeat steps S2.52-S2.54. After each step, the expected position is calculated based on the positions of all particles as the current best estimate, thereby forming a predicted trajectory.

[0023] Further, the contour construction module constructs an environment contour including positioning information, including the following steps: S3.1. The entrance of the tunnel is the origin of the coordinate system. The direction along the extension of the tunnel is the positive x-axis, the direction perpendicular to the main axis of the tunnel is the y-axis, and the direction perpendicular to the ground is the z-axis. S3.2, mapping the position information of all objects into a three-dimensional coordinate system; S3.3, classify the data by object type, including smart helmets, vehicles, and building materials; S3.4. Use clustering algorithm to group each type of object, and introduce the vibration intensity parameters generated by heavy machinery to optimize the clustering algorithm to reduce the impact of vibration on the data, thereby improving the accuracy and reliability of clustering results. For each clustering result, calculate their boundaries; use the convex hull method to approximate the outline of each cluster, and merge the outlines of all objects into a large outline to represent the scope of the entire construction environment; Among them, clustering algorithms can automatically identify the inherent structure in a data set without specifying category labels in advance. They do not need to label data, which reduces the cost and time consumption of manual annotation. They group large amounts of data into several clusters, which helps to simplify the subsequent data analysis process. Clustering can help identify outliers that are significantly different from most samples. Clustering can more intuitively understand the distribution of data sets and help people better interpret data. Furthermore, in a tunnel construction environment, the operation of heavy machinery and equipment will generate significant vibrations, which may affect the readings of the sensors on the smart helmets, resulting in inaccurate position data. By considering the vibration intensity, the impact of vibration on the position data can be reduced, and the noise in the data can be reduced, so that the clustering results are more accurate. Vibration may cause some data points to deviate from their true positions. If not taken into account, these deviated data points may be mistakenly assigned to incorrect clusters. After introducing the vibration intensity parameter, the algorithm can give smaller weights to data points with greater vibration intensity, which can reduce their impact on the cluster center, thereby improving the accuracy of clustering. The specific optimization of the clustering algorithm by introducing the vibration intensity parameter generated by heavy machinery and equipment is as follows: ; in, Indicates The cluster center of a cluster, that is, the weighted average position of all data points in the cluster; Indicates data points, of which , Indicates The location coordinates of the data points, Indicates the vibration intensity; Indicates The set of all data points of a cluster; Indicates The weight of the data point is used to reflect the impact of vibration intensity on the cluster center; , The adjustment parameter that indicates the influence of vibration intensity on weight. The larger the value of , the more obvious the effect of vibration intensity on weight.

[0024] S3.5. Use the 3D visualization tool Matplotlib to display the position and outline of each object.

[0025] Furthermore, the collision warning module performs collision warning, including the following steps: S4.1. Update the environment contour based on real-time data, especially for moving objects, including construction workers and vehicles, detect the position changes of dynamic obstacles and update their contours; S4.2. Use a collision time model to assess potential collision risks, predict the proximity between objects based on distance, speed, and acceleration information, and trigger a collision warning when the distance between objects is less than a safety threshold k; Among them, the collision time model is specifically as follows: ; In the formula, Indicates the time remaining before two objects collide; Indicates the relative distance between two objects, that is, the shortest distance between two objects; Represents the relative velocity between two objects. If the two objects are moving toward each other, the relative velocity will be the algebraic sum of their velocities; if they are moving away from each other, the relative velocity will be the absolute value of the difference between their velocities. S4.3. When a potential collision risk is detected, a warning signal is sent to relevant personnel. The warning signal includes sound, light, and vibration.

[0026] Example 2: The difference between Example 2 of the present invention and Example 1 is that this example introduces an intelligent supervision method for tunnel construction used by an intelligent supervision system for tunnel construction based on a smart safety helmet.

[0027] An intelligent supervision method for tunnel construction based on an intelligent helmet, used in the above-mentioned intelligent supervision system for tunnel construction based on an intelligent helmet, comprises the following steps: S5.1. Collect data on the location and working environment of construction workers; S5.2, transmitting the collected data via a wireless communication protocol; S5.3. Perform edge computing on the transmitted data for real-time analysis; analyze the positioning information through the centroid method to obtain accurate location information; use the particle filter method to predict the movement trajectory of construction workers; construct an environmental contour including positioning information based on the clustering algorithm; fuse the movement trajectory and environmental contour, and perform collision warning based on the collision time model.

[0028] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention, and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. An intelligent tunnel construction supervision system based on an intelligent helmet, characterized in that: include: A sensor unit (1), the sensor unit (1) is used to collect data on the location of construction personnel and working environment; A communication unit (2), the communication unit (2) being used to transmit the collected data via a wireless communication protocol; An edge computing unit (3), wherein the edge computing unit (3) performs edge computing on the data transmitted by the communication unit (2) for real-time analysis, analyzes the positioning information by a centroid method, and introduces the weight of personnel density and material stacking to optimize the calculation of the centroid coordinates; wherein the edge computing unit (3) comprises a trajectory prediction module, a contour construction module, and a collision warning module; The trajectory prediction module uses a particle filtering method to predict the movement trajectory of construction personnel; The contour construction module constructs an environment contour including positioning information based on a clustering algorithm; The collision warning module integrates the movement trajectory and the environment contour, and performs collision warning based on the collision time model.

2. The intelligent monitoring system for tunnel construction based on the intelligent helmet according to claim 1 is characterized in that: The sensing unit (1) comprises a positioning module and an image acquisition module; Among them, the positioning module is used to track and determine the position of the smart helmet; The image acquisition module is used to monitor the environmental conditions of the construction environment.

3. The intelligent monitoring system for tunnel construction based on the intelligent helmet according to claim 1 is characterized in that: The edge computing unit (3) analyzes the positioning information by using the centroid method, including the following steps: S1.

1. Deploy multiple reference points with known positions m in the tunnel, each of which has a known coordinate position; S1.

2. Use the smart helmet to measure the value; S1.

3. Utilization The relationship between the value and the distance is used to estimate the approximate distance between the smart helmet and the reference point; There is an exponential decay relationship with distance: ； in, Indicates the received signal strength indicator; is the reference point at a distance Where value; represents the path loss exponent; represents the estimated distance between the smart helmet and the reference point; Indicates the reference distance; S1.

4. Use the positions of all reference points and the corresponding estimated distances to calculate the center of mass coordinates of the smart helmet, and introduce the weights of personnel density and material stacking to optimize the calculation of the center of mass coordinates, so as to reduce the impact of signal propagation changes caused by high-density personnel and material stacking on the accuracy of positioning calculations.

4. The intelligent monitoring system for tunnel construction based on the intelligent helmet according to claim 3 is characterized in that: In S1.4, the weight of personnel density and material stacking is introduced to optimize the calculation of the centroid coordinates as follows: ； in, Indicates The weight after introducing the density of people and the stacking of materials at each reference point is: ; Indicates that the smart helmet is The estimated distance between reference points; Indicates The density factor of people near a reference point; Indicates Material stacking factor near a reference point; , , Indicates The coordinates of the reference points; Represents the coordinates of the center of mass of the smart helmet; Indicates the number of reference points.

5. The intelligent monitoring system for tunnel construction based on the intelligent helmet according to claim 1 is characterized in that: The trajectory prediction module predicts the movement trajectory of the construction personnel, including the following steps: S2.

1. Collect the location coordinates of the construction workers, record the timestamp of each location data, and collect the speed and acceleration data of the construction workers; S2.2, preprocess the data; S2.

3. Construct a time series containing the position coordinates of the construction workers at each time point, calculate the speed of the construction workers, and calculate the acceleration of the construction workers by dividing the distance between adjacent position coordinates by the corresponding time interval; S2.

4. Analyze the movement trends of construction personnel, including straight-line movement and turning; S2.

5. Use the particle filtering method to estimate the location probability distribution of construction workers and predict their movement trajectories.

6. The intelligent monitoring system for tunnel construction based on the intelligent helmet according to claim 5 is characterized in that: The particle filtering method is used in S2.5 to estimate the probability distribution of the location of the construction personnel and predict the movement trajectory, including the following steps: S2.

51. Set an initial particle set, each particle contains the location of the construction worker , as well as direction and velocity, where the particles represent probabilistic estimates of the construction worker’s location; S2.

52. Update the position of each particle using the state transition model: ； ； ； in, Represents particles exist Moment Position coordinates of the direction; Represents particles exist Moment Position coordinates of the direction; Represents particles exist Moment Position coordinates of the direction; Represents particles At the previous moment of Position coordinates of the direction; Represents particles At the previous moment of Position coordinates of the direction; Represents particles At the previous moment of Position coordinates of the direction; Represents particles exist Direction speed; Represents particles exist Direction speed; Represents particles exist Direction speed; Indicates a time interval; Indicates Particles in The random noise in the direction; Indicates Particles in Random noise in the direction; Indicates Particles in Random noise in the direction; S2.

53. Use the Gaussian distribution observation model to weight each particle: ； in, Indicated in Moment Particle The weight of is the actual observed value, is the observed value predicted from the particle state, is the standard deviation of the observation noise; S2.54, resampling according to the weight of the particles; S2.

55. Repeat steps S2.52-S2.

54. After each step, the expected position is calculated based on the positions of all particles as the current best estimate, thereby forming a predicted trajectory.

7. The intelligent monitoring system for tunnel construction based on the intelligent helmet according to claim 1 is characterized in that: The contour building module builds an environment contour including positioning information, including the following steps: S3.

1. The entrance of the tunnel is the origin of the coordinate system. The direction along the extension of the tunnel is the positive x-axis, the direction perpendicular to the main axis of the tunnel is the y-axis, and the direction perpendicular to the ground is the z-axis. S3.2, mapping the position information of all objects into a three-dimensional coordinate system; S3.3, classify the data by object type, including smart helmets, vehicles, and building materials; S3.

4. Use clustering algorithm to group each type of object, and introduce the vibration intensity parameters generated by heavy machinery and equipment to optimize the clustering algorithm to reduce the impact of vibration on the data. For each clustering result, calculate their boundaries; use convex hull method to approximate the outline of each cluster, and merge the outlines of all objects into a large outline to represent the scope of the entire construction environment; S3.

5. Use the 3D visualization tool Matplotlib to display the position and outline of each object.

8. The intelligent monitoring system for tunnel construction based on the intelligent helmet according to claim 7 is characterized in that: In S3.4, the vibration intensity parameters generated by heavy machinery are introduced to optimize the clustering algorithm as follows: ； in, Indicates The cluster centers of the clusters; Indicates data points, of which , Indicates The location coordinates of the data points, Indicates the vibration intensity; Indicates The set of all data points of a cluster; Indicates The weight of the data point is used to reflect the impact of vibration intensity on the cluster center; , A tuning parameter that indicates the degree of influence of vibration intensity on weight.

9. The intelligent monitoring system for tunnel construction based on the intelligent helmet according to claim 1 is characterized in that: The collision warning module performs collision warning, including the following steps: S4.

1. Update the environment contour based on real-time data, especially for moving objects, including construction workers and vehicles, detect the position changes of dynamic obstacles and update their contours; S4.

2. Use a collision time model to assess potential collision risks, predict the proximity between objects based on distance, speed, and acceleration information, and trigger a collision warning when the distance between objects is less than a safety threshold k; S4.

3. When a potential collision risk is detected, a warning signal is sent to relevant personnel.

10. A method for intelligent supervision of tunnel construction based on a smart helmet, based on the intelligent supervision system for tunnel construction based on a smart helmet as claimed in any one of claims 1 to 9, characterized in that: The steps include: S5.

1. Collect data on the location and working environment of construction workers; S5.2, transmitting the collected data via a wireless communication protocol; S5.

3. Perform edge computing on the transmitted data for real-time analysis; analyze the positioning information through the centroid method to obtain accurate location information; use the particle filter method to predict the movement trajectory of construction workers; construct an environmental contour including positioning information based on the clustering algorithm; fuse the movement trajectory and environmental contour, and perform collision warning based on the collision time model.

Images