Construction site road maintenance prediction method and system based on intelligent safety helmet motion data

Abstract

The invention discloses a construction site road maintenance prediction method and system based on intelligent safety helmet motion data, and the method comprises the steps: carrying out the region division of a construction site, and determining different work types, health data and motion data; health indexes of the workers are calculated according to the health data, and the health indexes which do not meet preset conditions are removed through clustering; analyzing the correlation degree between the residual motion data and the work type, and removing the motion data smaller than a preset threshold value; calculating the deviation between the health index of the worker in the target working data and the health reference line, and calculating the health change trend of the worker in a preset period; and substituting into a health change trend and road condition association model to predict the probability that the road in each region needs to be overhauled, and determining a construction site road detection strategy. According to the invention, the potential influence on the health of workers caused by poor road conditions can be researched according to the health state trend of the workers, so that road maintenance is planned in advance, and the road construction safety is improved while resource allocation is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of construction safety management and control, and in particular to a construction site road maintenance prediction method and system based on intelligent safety helmet motion data. Background Art

[0002] In traditional construction site road maintenance management, manual experience and regular inspections are mainly used to determine whether roads need to be repaired. Manual judgment is highly subjective, and due to the complex environment of construction sites, the road conditions in different areas vary greatly due to factors such as construction activities and worker mobility. It is difficult to accurately grasp the actual wear and tear of roads in various areas by simply relying on regular inspections, which can easily lead to some roads being repaired too early or too late, wasting resources and possibly affecting construction safety and efficiency. Summary of the invention

[0003] In order to solve at least one of the technical problems mentioned above, the present invention provides a construction site road maintenance prediction method and system based on smart safety helmet motion data.

[0004] In a first aspect, the present invention provides a construction site road maintenance prediction method based on intelligent safety helmet motion data, the method comprising:

[0005] Divide the construction site into grid areas, determine different types of work according to the color identification of the smart helmets in each area, and collect work data of each worker in different types of work within a preset period according to the smart helmets, wherein the work data includes health data and exercise data;

[0006] Calculate the health index of workers based on health data, cluster the health indexes of the same type of work in the same area, and eliminate work data whose health index does not meet the preset conditions; analyze the correlation between motion data and corresponding types of work in the remaining work data, eliminate work data with a correlation less than a preset threshold, and obtain target work data;

[0007] Establish health baselines for different types of jobs, calculate the deviation between the health index of each worker in the target job data and the health baseline, and calculate the health change trend of workers in the same area within a preset period based on all deviations;

[0008] A correlation model between health change trends and road conditions is constructed, and the health change trends of workers are substituted to predict the probability that roads in each area need maintenance. The road inspection strategy on the construction site is determined based on the predicted probability values.

[0009] Preferably, the construction of the correlation model between health change trends and road conditions includes:

[0010] ;

[0011] ;

[0012] In the formula, represents the probability that regional roads need maintenance, Indicates the changing trend of workers' health, represents environmental impact factors; Indicates the frequency of use of heavy equipment; represents the error term; , , represents the weight factor, , Represents the weight coefficient of temperature and humidity; Respectively represent the maximum, minimum and ideal values ​​of temperature; Respectively represent the maximum, minimum and ideal values ​​of humidity.

[0013] Preferably, the analysis of the correlation between the motion data in the remaining work data and the corresponding work type includes:

[0014] Obtain position data, speed data, and acceleration data from historically collected motion data, clean and denoise the data, and mark the denoised data to generate sample data;

[0015] The random forest classifier is used to train the sample data to generate a recognition model, the motion features of the current motion data are extracted, and the recognition model is used to calculate the importance score of each motion feature. The correlation between the feature and the job type is determined according to the importance score.

[0016] In a second aspect, the present invention further provides a construction site road maintenance prediction system based on intelligent safety helmet motion data, the system comprising:

[0017] A region division unit is used to divide the construction site into grid regions, determine different types of work according to the color identification of the smart helmet in each region, and collect work data of each worker in different types of work within a preset period according to the smart helmet, wherein the work data includes health data and exercise data;

[0018] The abnormal elimination unit is used to calculate the health index of the workers based on the health data, cluster the health indexes of the same type of work in the same area, and eliminate the work data whose health index does not meet the preset conditions; analyze the correlation between the motion data and the corresponding type of work in the remaining work data, eliminate the work data whose correlation is less than the preset threshold, and obtain the target work data;

[0019] The trend prediction unit is used to establish health baselines for different types of work, calculate the deviation between the health index of each worker in the target work data and the health baseline, and calculate the health change trend of workers in the same area within a preset period based on all deviations;

[0020] The maintenance assessment unit is used to build a correlation model between health change trends and road conditions, substitute the health change trends of workers to predict the probability that roads in each area need maintenance, and determine the construction site road inspection strategy based on the predicted probability values.

[0021] Preferably, the inspection and evaluation unit is used to construct a correlation model between health change trends and road conditions, including:

[0022] ;

[0023] ;

[0024] In the formula, represents the probability that regional roads need maintenance, Indicates the changing trend of workers' health, represents environmental impact factors; Indicates the frequency of use of heavy equipment; represents the error term; , , represents the weight factor, , Represents the weight coefficient of temperature and humidity; Respectively represent the maximum, minimum and ideal values ​​of temperature; Respectively represent the maximum, minimum and ideal values ​​of humidity.

[0025] Preferably, the abnormal elimination unit is used to analyze the correlation between the motion data in the remaining work data and the corresponding work type, including:

[0026] Obtain position data, speed data, and acceleration data from historically collected motion data, clean and denoise the data, and mark the denoised data to generate sample data;

[0027] The random forest classifier is used to train the sample data to generate a recognition model, the motion features of the current motion data are extracted, and the recognition model is used to calculate the importance score of each motion feature. The correlation between the feature and the job type is determined according to the importance score.

[0028] In a third aspect, the present invention further provides an electronic device comprising a processor and a memory, wherein the memory is used to store computer program code, and the computer program code comprises computer instructions. When the processor executes the computer instructions, the electronic device executes the method as described in the first aspect above and any possible implementation thereof.

[0029] In a fourth aspect, the present invention further provides a computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, wherein the computer program includes program instructions, and when the program instructions are executed by a processor of an electronic device, the processor executes the method as described in the first aspect above and any possible implementation thereof.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] The present invention provides a construction site road maintenance prediction method based on smart helmet motion data, comprising dividing a construction site into grid areas, determining different types of work according to the color identification of smart helmets in each area, collecting work data of each worker in different types of work within a preset period according to the smart helmet, wherein the work data comprises health data and motion data; calculating the health index of the worker according to the health data, clustering the health indexes belonging to the same type of work in the same area, and eliminating the work data whose health index does not meet the preset conditions; analyzing the correlation between the motion data and the corresponding type of work in the remaining work data, eliminating the work data whose correlation is less than a preset threshold, and obtaining the target work data; establishing health baselines for different types of work, calculating the deviation between the health index of each worker in the target work data and the health baseline, and calculating the health change trend of the workers in the same area within the preset period according to all deviations; constructing a correlation model between the health change trend and the road condition, substituting the health change trend of the workers into the probability that the roads in each area need to be repaired, and determining the construction site road detection strategy according to the predicted probability value.

[0032] The present invention realizes refined management of different types of work on the construction site by grid management of the construction site and using the color identification of the smart safety helmet to distinguish different types of work. The health data is processed by cluster analysis and other methods to ensure that the work data involved in the subsequent analysis has a high quality and reduce the errors caused by abnormal values. By analyzing the correlation between motion data and the corresponding types of work, useful work data is further screened out, providing a reliable basis for the establishment of subsequent health baselines. Health baselines are established according to different types of work, and the health change trends of workers are predicted based on this. Combined with road conditions, the probability of road maintenance on the construction site is finally predicted, which improves the accuracy of the prediction results, reduces the cost of road maintenance and optimizes resource allocation.

[0033] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the background technology, the drawings required for use in the embodiments of the present invention or the background technology will be described below.

[0035] Figure 1 A schematic flow chart of a construction site road maintenance prediction method based on intelligent safety helmet motion data provided by an embodiment of the present invention;

[0036] Figure 2 A schematic structural diagram of a construction site road maintenance prediction system based on smart safety helmet motion data provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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.

[0038] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0039] In traditional construction site road maintenance management, manual experience and regular inspections are mainly relied upon to determine whether roads need maintenance. This method is not only highly subjective and difficult to accurately grasp the actual wear and tear of roads in various regions, but also usually requires a lot of manpower and material resources, so there are disadvantages of inaccurate maintenance prediction results and low resource utilization. To this end, the present invention aims to provide a construction site road maintenance prediction method based on smart helmet motion data, which mainly combines the construction workers' smart helmet motion data to judge the workers' health trends, and combines the correlation model of health change trends and road conditions to predict which regional roads need maintenance, thereby improving the accuracy of prediction, reducing maintenance costs, and optimizing resource allocation.

[0040] See also Figure 1 , Figure 1The present invention provides a flowchart of a method for predicting construction site road maintenance based on intelligent helmet motion data. Figure 1 As shown, the method includes:

[0041] S10. Divide the construction site into grid areas, determine different types of work according to the color identification of the smart safety helmets in each area, and collect work data of each worker in different types of work within a preset period according to the smart safety helmets, wherein the work data includes health data and exercise data.

[0042] Usually, choose a smart helmet with color identification, health data collection (such as heart rate, body temperature monitoring, etc.) and sports data collection (such as number of steps, acceleration, displacement, etc.), and assign a unique number to each helmet to associate with the worker's identity information. According to the actual situation of the construction site, set the preset cycle of data collection, such as daily or weekly cycles, or certain time periods as sampling cycles, which can be set according to the needs of the scene. Then determine the correspondence between the type of work and the color identification of the helmet, and sort out all the types of work on the construction site, such as carpentry, rebar workers, masons, scaffolders, etc. Assign a smart helmet of a specific color to each type of work, develop a detailed color and type of work correspondence table, and train workers to ensure that workers wear the corresponding color helmet correctly.

[0043] The terrain of construction site roads is usually complex and changeable. If the traditional inspection method is used, it is easy to miss something. For this reason, this embodiment considers dividing the construction site into grid areas to ensure that accurate predictions can be made when predicting roads that need to be repaired. Specifically, the construction site is accurately measured using total stations, GPS and other measuring equipment to draw a detailed site map. The site map is divided into multiple grid areas based on factors such as the functional area of ​​the site and the construction progress, and each grid area is assigned a unique number.

[0044] Furthermore, when collecting data, workers wear smart helmets when entering the construction site, and the helmets automatically turn on the data collection function to continuously collect health data and exercise data according to the preset cycle. Use Bluetooth positioning, UWB positioning and other technologies to determine the position of the helmet in the grid area, and associate the collected data with the grid area number, worker identity information, type of work, etc. Build Wi-Fi, 4G / 5G and other data transmission networks in the construction site to ensure that the data collected by the smart helmet can be transmitted to the data server in real time. Establish a database on the data server, and classify and store the collected work data according to dimensions such as grid area, type of work, worker identity, and time. In order to improve the quality of data collection, regularly clean the stored data to remove abnormal data, duplicate data, etc. to ensure data quality.

[0045] S20. Calculate the health index of the workers based on the health data, cluster the health indexes of the same type of work in the same area, and eliminate the work data whose health index does not meet the preset conditions; analyze the correlation between the motion data and the corresponding type of work in the remaining work data, eliminate the work data whose correlation is less than the preset threshold, and obtain the target work data.

[0046] Health data may generally include indicators such as heart rate, body temperature, blood pressure, and blood oxygen concentration. In order to calculate a comprehensive health index, corresponding weights may be assigned to health data indicators such as heart rate, body temperature, blood pressure, and blood oxygen concentration, and a health index calculation formula may be established, such that the health index is the weighted sum of each indicator and its respective weight.

[0047] Preferably, when performing data analysis, various indicators can be screened before calculating the health index. If a certain indicator is found to be obviously abnormal, for example, a worker's blood pressure is significantly higher than the normal range, it means that the cause is largely due to his own physical condition. Such data should be eliminated and not participate in the subsequent model of association analysis with road conditions. After calculating the health index, a secondary screening can also be performed. For example, a clustering algorithm (such as K-Means clustering) can be used to perform cluster analysis on the health index of workers belonging to the same type of work in the same area, and workers with similar health indexes are divided into the same category. Then, according to the construction site health management standards, the preset conditions for the health index are set, such as a health index below 60 points is considered to be unsatisfied. The work data of workers whose health index does not meet the preset conditions are removed from the database, and this part of the data is also not involved in the subsequent analysis process.

[0048] The above elimination process can eliminate the interference of some workers' own physical conditions on the road maintenance analysis. During the construction process, factors that also affect the health status include the correlation between the worker's motion data and the type of work. If a worker is assigned to a specific type of work, but his actual work behavior (such as movement mode, operation method, etc.) does not match the standard behavior of the type of work, this may indicate that the worker is performing tasks that are not suitable for his skill set, or is completing tasks in an incorrect way. For example, if a worker engaged in heavy physical labor does not carry heavy objects in the correct posture, the risk of muscle strain or other physical injuries may increase. In the process of calculating the health index based on the data collected by the smart helmet, if the type of work actually performed by the worker and its corresponding normal movement mode are not accurately identified, the calculation result of the health index may be biased. In this case, even if there is actually a problem with the worker's health status, it may not be discovered and processed in time due to the low data correlation. Therefore, in order to reduce the interference of this factor, this embodiment will further perform data cleaning to ensure the data quality used to analyze the road maintenance process.

[0049] Specifically, in this embodiment, in order to analyze the correlation between the workers' motion data and the type of work, motion data features, such as the number of steps, acceleration, displacement, etc., will be first extracted from the remaining work data. These features may have different performances for different types of work. For example, the displacement data and acceleration data of the scaffolder may be more prominent, while the number of steps of the bricklayer is relatively small but the hand movement acceleration data is characteristic. Using algorithms such as correlation analysis, the correlation between each motion data feature and the corresponding type of work is calculated, for example, the Pearson correlation coefficient between the number of steps and the bricklayer type is calculated. According to data analysis experience and the actual situation of the construction site, a preset threshold value of the correlation is set, such as 0.3. The work data with a correlation less than the preset threshold is removed from the database, and the target work data is finally obtained.

[0050] In one embodiment, preferably, the analyzing the correlation between the motion data in the remaining work data and the corresponding work type includes:

[0051] 1) Obtain position data, velocity data, and acceleration data from historically collected motion data, clean and denoise the data, and mark the denoised data to generate sample data;

[0052] 2) Use the random forest classifier to train the sample data to generate a recognition model, extract the motion features of the current motion data, and use the recognition model to calculate the importance score of each motion feature. Determine the correlation between the feature and the type of work based on the importance score.

[0053] First, ensure that the smart helmet can accurately collect the worker's position, speed, acceleration and other motion data in different types of work. This data is usually obtained through built-in sensors (such as GPS, accelerometer). Since the raw data may contain noise or erroneous readings, the data needs to be cleaned. This step can use filtering techniques (such as Kalman filtering, low-pass filtering) to remove noise and ensure the accuracy of the data. For each cleaned record, it is necessary to mark it according to the type of task performed by the worker. For example, if a worker is tying steel bars, all relevant data in this time period should be marked as "steel tying" category. Use the labeled data set as sample data for subsequent model training. Ensure that there are enough samples for each type of work to ensure the generalization ability of the model.

[0054] When training a recognition model, you can usually choose a random forest classifier as the main machine learning model. Random forest is an ensemble learning method that improves prediction accuracy and controls overfitting by building multiple decision trees and merging their results. Use the prepared labeled sample data to train the random forest model. During the training process, the model automatically learns which features are most important for distinguishing different types of work. Use the trained random forest model to score the importance of the features of the new data. In a random forest, you can evaluate the importance of each feature by looking at its contribution to the split point in the decision tree. The importance score reflects the degree of influence of a specific feature on the correct classification of a new instance. Finally, based on the importance score of the feature, you can analyze which features are most closely related to a specific type of work. Features with higher importance scores are more likely to be key factors in distinguishing different types of work. In this way, the correlation between motion data and the corresponding type of work can be effectively analyzed.

[0055] S30. Establish health baselines for different types of work, calculate the deviation between the health index of each worker in the target work data and the health baseline, and calculate the health change trend of workers in the same area within a preset period based on all deviations.

[0056] Usually each type of job will correspond to a different health baseline, for example, heavy machinery operators and electricians have different health baselines. For each type of job, use historical health data to determine its unique health baseline. This can be achieved through statistical analysis (such as mean, standard deviation) or machine learning algorithms to identify "healthy" patterns and ranges. Given the differences in work intensity among different types of jobs, the health baseline should reflect these differences. For example, the health baseline for heavy machinery operators may focus more on the potential health effects of long-term exposure to vibration, while construction workers may focus more on physical fatigue caused by physical labor.

[0057] After determining the health baselines for different types of jobs, calculate the deviation between the health index of each worker in the target job data and the health baseline, and collect and record the health index deviation value of each worker within a preset time period (e.g., one month). Apply time series analysis techniques (such as moving average, exponential smoothing, etc.) or other forecasting models (such as ARIMA model) to analyze the changing trends of these deviation values ​​over time.

[0058] S40, constructing a correlation model between health change trends and road conditions, substituting the health change trends of workers into the model to predict the probability that roads in each area need to be inspected, and determining the road inspection strategy on the construction site based on the predicted probability value.

[0059] Preferably, the construction of the correlation model between health change trends and road conditions includes:

[0060] ;

[0061] ;

[0062] In the formula, represents the probability that regional roads need maintenance, Indicates the changing trend of workers' health, represents environmental impact factors; Indicates the frequency of use of heavy equipment; represents the error term; , , represents the weight factor, , Represents the weight coefficient of temperature and humidity; Respectively represent the maximum, minimum and ideal values ​​of temperature; Respectively represent the maximum, minimum and ideal values ​​of humidity.

[0063] In the above formula, Indicates the trend of worker health changes assessed in the previous step, reflecting whether the health status of workers has improved or deteriorated over a period of time. If the health status of most workers in an area continues to deteriorate, this may indirectly indicate that there are potential problems in the working environment (such as road conditions). As an environmental factor, the effects of changes in temperature and humidity on road conditions and worker health are considered comprehensively. Extreme temperature and humidity not only affect the aging rate of road materials, but may also make workers more susceptible to discomfort or illness. The frequency of use of heavy equipment will cause greater wear and tear on the road and increase the risk of road damage.

[0064] Suppose that after eliminating the effects of individual physical fitness and job types, it is found that the physical health of workers in a certain area gradually deteriorates over time. One possible reason for this is that the road conditions in the area are poor, causing workers to face higher safety risks or greater physical exertion at work. For example:

[0065] Poor road quality: Uneven roads make it more difficult to walk and operate equipment, making workers more susceptible to fatigue or injury. If workers are in such an environment for a long time, their health may deteriorate.

[0066] Environmental factors: In addition to problems with the road itself, adverse weather conditions, such as extremely hot or cold temperatures and high humidity, can exacerbate the deterioration of workers’ health and accelerate the aging and deterioration of road materials.

[0067] Use of heavy equipment: Frequent use of heavy equipment not only increases wear and tear on roads, but may also generate vibration and noise, further affecting workers’ health.

[0068] Therefore, by building a correlation model between health trends and road conditions, it can effectively help identify which areas of roads may need priority maintenance. By comprehensively considering the health trends of workers, environmental factors, and the frequency of heavy equipment use, it is possible to predict the probability of roads in each area needing maintenance, and accordingly formulate a reasonable construction site road inspection strategy, which will help reduce safety accidents and health problems caused by poor road conditions.

[0069] In summary, the method provided by the embodiment of the present invention realizes the refined management of different types of work in the construction site by grid management of the construction site and distinguishing different types of work by the color identification of the smart safety helmet. The health data is processed by cluster analysis and other methods to ensure that the work data involved in the subsequent analysis has a high quality and reduce the errors caused by abnormal values. By analyzing the correlation between motion data and the corresponding type of work, useful work data is further screened out, providing a reliable basis for the establishment of subsequent health baselines. Health baselines are established according to different types of work, and the health change trend of workers is predicted accordingly. Combined with road conditions, the probability of road maintenance on the construction site is finally predicted, which improves the accuracy of the prediction results, reduces the cost of road maintenance and optimizes resource allocation.

[0070] See also Figure 2 In one embodiment, the present invention further provides a construction site road maintenance prediction system based on intelligent safety helmet motion data, the system comprising:

[0071] The area division unit 100 is used to divide the construction site into grid areas, determine different types of work according to the color identification of the smart helmet in each area, and collect work data of each worker in different types of work within a preset period according to the smart helmet, wherein the work data includes health data and exercise data;

[0072] The abnormal elimination unit 200 is used to calculate the health index of the worker based on the health data, cluster the health indexes of the same type of work in the same area, and eliminate the work data whose health index does not meet the preset conditions; analyze the correlation between the motion data and the corresponding type of work in the remaining work data, eliminate the work data whose correlation is less than the preset threshold, and obtain the target work data;

[0073] The trend prediction unit 300 is used to establish health baselines for different types of work, calculate the deviation between the health index of each worker in the target work data and the health baseline, and calculate the health change trend of workers in the same area within a preset period based on all deviations;

[0074] The maintenance evaluation unit 400 is used to construct a correlation model between health change trends and road conditions, substitute the health change trends of workers into the prediction of the probability that roads in various areas need maintenance, and determine the construction site road detection strategy based on the predicted probability value.

[0075] In one embodiment, the inspection and evaluation unit 400 is used to construct a correlation model between health change trends and road conditions, including:

[0076] ;

[0077] ;

[0078] In the formula, represents the probability that regional roads need maintenance, Indicates the changing trend of workers' health, represents environmental impact factors; Indicates the frequency of use of heavy equipment; represents the error term; , , represents the weight factor, , Represents the weight coefficient of temperature and humidity; Respectively represent the maximum, minimum and ideal values ​​of temperature; Respectively represent the maximum, minimum and ideal values ​​of humidity.

[0079] In one embodiment, the abnormal elimination unit 200 is used to analyze the correlation between the motion data in the remaining work data and the corresponding work type, including:

[0080] Obtain position data, speed data, and acceleration data from historically collected motion data, clean and denoise the data, and mark the denoised data to generate sample data;

[0081] The random forest classifier is used to train the sample data to generate a recognition model, the motion features of the current motion data are extracted, and the recognition model is used to calculate the importance score of each motion feature. The correlation between the feature and the job type is determined according to the importance score.

[0082] It can be understood that the functions or modules included in the system provided in this embodiment can be used to execute the method described in the above method embodiment. Its specific implementation can refer to the description of the above method embodiment. For the sake of brevity, it will not be repeated here.

[0083] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0084] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. Those skilled in the art can also clearly understand that the descriptions of the various embodiments of the present invention have different focuses. For the convenience and brevity of description, the same or similar parts may not be repeated in different embodiments. Therefore, for parts not described or not described in detail in a certain embodiment, refer to the records of other embodiments.

Claims

1. A construction site road maintenance prediction method based on intelligent helmet motion data, characterized in that: The method comprises: Divide the construction site into grid areas, determine different types of work according to the color identification of the smart helmets in each area, and collect work data of each worker in different types of work within a preset period according to the smart helmets, wherein the work data includes health data and exercise data; Calculate the health index of workers based on health data, cluster the health indexes of the same type of work in the same area, and eliminate work data whose health index does not meet the preset conditions; analyze the correlation between motion data and corresponding types of work in the remaining work data, eliminate work data with a correlation less than a preset threshold, and obtain target work data; Establish health baselines for different types of jobs, calculate the deviation between the health index of each worker in the target job data and the health baseline, and calculate the health change trend of workers in the same area within a preset period based on all deviations; A correlation model between health change trends and road conditions is constructed, and the health change trends of workers are substituted to predict the probability that roads in each area need maintenance. The road inspection strategy on the construction site is determined based on the predicted probability values.

2. The construction site road maintenance prediction method based on intelligent helmet motion data according to claim 1 is characterized in that: The construction of the correlation model between health change trend and road conditions includes: ； ； In the formula, represents the probability that regional roads need maintenance, Indicates the changing trend of workers' health, represents environmental impact factors; Indicates the frequency of use of heavy equipment; represents the error term; , , represents the weight factor, , Represents the weight coefficient of temperature and humidity; Respectively represent the maximum, minimum and ideal values ​​of temperature; Respectively represent the maximum, minimum and ideal values ​​of humidity.

3. The construction site road maintenance prediction method based on intelligent helmet motion data according to claim 1 is characterized in that: The analysis of the correlation between the motion data in the remaining work data and the corresponding work type includes: Obtain position data, speed data, and acceleration data from historically collected motion data, clean and denoise the data, and mark the denoised data to generate sample data; The random forest classifier is used to train the sample data to generate a recognition model, the motion features of the current motion data are extracted, and the recognition model is used to calculate the importance score of each motion feature. The correlation between the feature and the job type is determined according to the importance score.

4. A construction site road maintenance prediction system based on intelligent helmet motion data, characterized in that: The system comprises: A region division unit is used to divide the construction site into grid regions, determine different types of work according to the color identification of the smart helmet in each region, and collect work data of each worker in different types of work within a preset period according to the smart helmet, wherein the work data includes health data and exercise data; The abnormal elimination unit is used to calculate the health index of the workers based on the health data, cluster the health indexes of the same type of work in the same area, and eliminate the work data whose health index does not meet the preset conditions; analyze the correlation between the motion data and the corresponding type of work in the remaining work data, eliminate the work data whose correlation is less than the preset threshold, and obtain the target work data; The trend prediction unit is used to establish health baselines for different types of work, calculate the deviation between the health index of each worker in the target work data and the health baseline, and calculate the health change trend of workers in the same area within a preset period based on all deviations; The maintenance assessment unit is used to build a correlation model between health change trends and road conditions, substitute the health change trends of workers to predict the probability that roads in each area need maintenance, and determine the construction site road inspection strategy based on the predicted probability value.

5. The construction site road maintenance prediction system based on intelligent helmet motion data according to claim 4 is characterized in that: The inspection and evaluation unit is used to construct a correlation model between health change trends and road conditions, including: ； ； In the formula, represents the probability that regional roads need maintenance, Indicates the changing trend of workers' health, represents environmental impact factors; Indicates the frequency of use of heavy equipment; represents the error term; , , represents the weight factor, , Represents the weight coefficient of temperature and humidity; Respectively represent the maximum, minimum and ideal values ​​of temperature; Respectively represent the maximum, minimum and ideal values ​​of humidity.

6. The construction site road maintenance prediction system based on intelligent helmet motion data according to claim 4 is characterized in that: The abnormal elimination unit is used to analyze the correlation between the motion data in the remaining work data and the corresponding work type, including: Obtain position data, speed data, and acceleration data from historically collected motion data, clean and denoise the data, and mark the denoised data to generate sample data; The random forest classifier is used to train the sample data to generate a recognition model, the motion features of the current motion data are extracted, and the recognition model is used to calculate the importance score of each motion feature. The correlation between the feature and the job type is determined according to the importance score.

7. An electronic device, characterized in that: include: A processor and a memory, wherein the memory is used to store computer program codes, wherein the computer program codes include computer instructions. When the processor executes the computer instructions, the electronic device executes the construction site road maintenance prediction method based on the motion data of the smart helmet as described in any one of claims 1 to 3.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes the construction site road maintenance prediction method based on smart helmet motion data as described in any one of claims 1 to 3.

Images