Construction risk prediction method based on machine learning algorithm

Abstract

The invention discloses a construction risk prediction method based on a machine learning algorithm, and belongs to the technical field of risk prediction. The method comprises the following steps: acquiring current construction data and construction data when a historical risk occurs; obtaining a current construction risk coefficient and a historical construction risk coefficient; obtaining working data of current constructors and working data of historical constructors; obtaining a current constructor fatigue coefficient and a historical constructor fatigue coefficient; constructing a construction risk prediction model; and based on the current construction data, the current construction risk coefficient, the working data of the current constructor and the fatigue coefficient of the current constructor, performing construction risk prediction through a construction risk prediction model. Through the construction data and personnel behavior analysis based on the machine learning algorithm, the technical effects of effectively predicting the construction risk and timely early warning are achieved, and the problems of low construction risk prediction precision and untimely response in the prior art are solved.

Description

Technical Field

[0001] The present invention relates to the field of risk technology, and in particular to a construction risk prediction method based on a machine learning algorithm. Background Art

[0002] With the booming construction industry and the increasing scale and complexity of engineering projects, the types of risks faced during construction are increasing, and accurate prediction of construction risks has become crucial. Traditional construction risk assessment methods can no longer meet the needs of modern engineering construction. Existing construction risk prediction systems usually identify potential risks through statistical analysis of historical data, input of expert experience, and certain rule reasoning. For example, some systems predict possible risk points by analyzing environmental factors, construction progress, weather changes and other data at the construction site, combined with previous accident cases and project management experience.

[0003] In the existing technology, since expert experience is highly subjective, there may be large differences in the judgments of different experts, resulting in low reliability of risk prediction results and an inability to provide comprehensive, efficient and accurate risk warnings for the construction process. Summary of the invention

[0004] The embodiment of the present application provides a construction risk prediction method based on a machine learning algorithm, which solves the problem of low reliability of risk prediction results in the prior art and inability to provide comprehensive, efficient and accurate risk warnings for the construction process. It achieves a more comprehensive capture of various risk factors and their relationships in the construction process, greatly improving the accuracy and comprehensiveness of risk prediction.

[0005] The embodiment of the present application provides a construction risk prediction method based on a machine learning algorithm, comprising the following steps: obtaining current construction data and construction data when historical risks occur, and performing preprocessing to obtain the preprocessed current construction data and the preprocessed construction data when historical risks occur; obtaining the current construction risk coefficient and the historical construction risk coefficient based on the preprocessed current construction data and the preprocessed construction data when historical risks occur, respectively; obtaining the work data of current construction personnel and the work data of historical construction personnel, and performing preprocessing to obtain the preprocessed work data of current construction personnel and the preprocessed work data of historical construction personnel; obtaining the fatigue coefficient of current construction personnel and the fatigue coefficient of historical construction personnel based on the preprocessed work data of current construction personnel and the preprocessed work data of historical construction personnel, respectively; constructing a construction risk prediction model based on the preprocessed construction data when historical risks occur, the historical construction risk coefficient, the preprocessed work data of historical construction personnel, and the fatigue coefficient of historical construction personnel; and performing construction risk prediction through the construction risk prediction model based on the preprocessed current construction data, the current construction risk coefficient, the preprocessed work data of current construction personnel, and the current construction personnel fatigue coefficient.

[0006] Furthermore, the current construction risk coefficient and the historical construction risk coefficient are obtained in the following manner: based on the pre-processed current construction data and the pre-processed construction data when the historical risk occurred, the weather data during the current construction and the historical risk occurred, the geological data of the construction location, the construction equipment data, and the operation data of the construction personnel are extracted respectively; the weight factors of the weather data during the construction, the geological data of the construction location, the construction equipment data, and the operation data of the construction personnel for the construction risk coefficient are obtained from the database; the current construction risk coefficient and the historical construction risk coefficient are calculated respectively by the construction risk coefficient formula; the construction risk coefficient formula is: ; In the formula, is the construction risk factor, is the temperature influence coefficient, is the formation stability influence coefficient, is the equipment failure influence coefficient, is the influence coefficient of personnel operation behavior, a is right The weight factor of right The weight factor, c is right The weight factor d is right The weight factor of .

[0007] Furthermore, the temperature influence coefficient is obtained by: obtaining the real-time temperature of the kth temperature monitoring point in the construction site through the temperature sensor, obtaining the historical average temperature of the corresponding monitoring point in the current season, the allowable difference of the temperature fluctuation of the corresponding monitoring point, the total number of temperature monitoring points, and performing preprocessing; obtaining the weight factor of the real-time temperature of the kth temperature monitoring point on the temperature influence coefficient from the database; obtaining the temperature influence coefficient through the temperature influence coefficient formula; the temperature influence coefficient formula is: ; In the formula, is the real-time temperature of the kth temperature monitoring point in the construction site, is the historical average temperature of the corresponding monitoring point in the current season, is the allowable difference of temperature fluctuation at the corresponding monitoring point, is the weight of the kth monitoring point, and m is the total number of temperature monitoring points.

[0008] Furthermore, the method for obtaining the formation stability influence coefficient is as follows: obtaining the soil compressive strength of the construction site through a sensor, obtaining the soil compressive strength threshold from the database according to the construction type, and obtaining the formation stability influence coefficient through the formation stability influence coefficient formula; the formation stability influence coefficient formula is: ; In the formula, is the compressive strength of the soil at the construction site, is the soil compressive strength threshold, and e is a natural constant.

[0009] Furthermore, the equipment failure influence coefficient is obtained by: collecting the number of failures of equipment at the construction site, obtaining the upper limit of the total number of failures of the equipment within the normal use cycle, and the total number of equipment types; obtaining the weight factor of the number of failures of each type of equipment on the equipment failure influence coefficient from the database; obtaining the equipment failure influence coefficient through the equipment failure influence coefficient formula; the equipment failure influence coefficient is: ; In the formula, is the number of failures of the qth type of equipment, is the upper limit of the total number of failures of the qth type of equipment in the normal use cycle, is the number of failures of the qth type of equipment The weight factor is , and r is the total number of device types.

[0010] Furthermore, the steps for obtaining the influence coefficient of personnel operation behavior are as follows: obtaining the maximum force applied by the construction personnel on the operating tool in each operation process through the pressure sensor, and obtaining the standard applied force in each operation process; obtaining the weight factor of the influence coefficient of the maximum force applied on the operating tool in each operation process on the equipment failure from the database; obtaining the influence coefficient of personnel operation behavior through the influence coefficient formula of personnel operation behavior; the influence coefficient formula of personnel operation behavior is: The influence coefficient formula of personnel operation behavior is: ; In the formula, is the maximum force applied by the construction worker on the operating tool during the nth operation. The standard applied force during the nth operation, is the weight factor of the influence coefficient of the maximum force applied to the operating tool in the nth operating process on the operating behavior of the personnel, and s is the total number of operating processes.

[0011] Furthermore, the steps for obtaining the fatigue coefficient of the current construction workers and the fatigue coefficient of the historical construction workers are as follows: based on the pre-processed work data of the current construction workers and the pre-processed work data of the historical construction workers, extract the current construction workers' continuous working time, the current construction workers' working time period, the historical construction workers' continuous working time, and the historical construction workers' working time period; obtain the weight factors of the construction workers' continuous working time and the construction workers' working time period for the construction workers' fatigue coefficient from the database; obtain the longest continuous working time stipulated at the construction site, the actual working time of the construction workers in the current time period at the construction site, and the ideal working time of the construction workers in the current time period at the construction site; obtain the current construction workers' fatigue coefficient and the historical construction workers' fatigue coefficient through the construction workers' fatigue coefficient formula.

[0012] Furthermore, the fatigue coefficient formula of the construction workers is: ; In the formula, F is the fatigue coefficient of the construction workers, t is the continuous working time of the construction workers, The maximum continuous working time is specified. is the weight factor of continuous working time on F, represents the actual working hours of the construction workers in the i-th working period, is the ideal working time in the i-th working period, is the duration of the working period. is the weight factor of the working time deviation within the working time period to F, and v is the total number of working time periods divided in a construction worker's work in one day.

[0013] Furthermore, constructing a construction risk prediction model includes the following steps: dividing a training set and a test set based on the pre-processed historical risk occurrence construction data, the historical construction risk coefficient, the historical risk occurrence construction data, and the historical construction personnel fatigue coefficient; inputting the training set into a polynomial regression model for training to obtain a preliminary construction risk prediction model; inputting the test set into the preliminary construction risk prediction model for testing to obtain a construction risk prediction model.

[0014] Furthermore, the construction risk prediction includes the following steps: inputting the pre-processed current construction data, the current construction risk coefficient, the work data of the current construction personnel, and the fatigue coefficient of the current construction personnel into the construction risk prediction model; the model predicts according to the input data, and outputs a quantitative value representing the construction risk, the value range is between 0 and 1, where 0 represents no risk and 1 represents the highest risk; according to the output risk quantification value, it is mapped to a specific risk level, and thresholds A, B, C, and D are set. The risk quantification value is not less than threshold A and is lower than threshold B for low risk, the risk quantification value is not less than threshold B and is lower than threshold C for medium risk, and the risk quantification value is not less than threshold C and is lower than threshold D for high risk; when the prediction result is low risk, routine monitoring is required; when it is medium risk, the construction process needs to be adjusted; when it is high risk, an alarm to suspend construction is issued to notify relevant personnel.

[0015] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. By combining current construction data, historical construction data and construction personnel's work data, a construction risk prediction model is generated based on a machine learning algorithm, thereby achieving accurate prediction of construction risks, effectively solving the problems of reliance on manual experience, incomplete data processing and insufficient prediction accuracy in existing technologies.

[0016] 2. By comprehensively calculating the construction risk coefficient based on factors such as temperature, formation stability, equipment failure and personnel operation, the construction risk is predicted through a polynomial regression model, thereby achieving timely feedback on the dynamic changes of the construction site and accurate prediction, effectively solving the problem of the lack of real-time and dynamic update capabilities of traditional risk assessment models in existing technologies.

[0017] 3. By extracting the continuous working time and working time period data of construction workers and combining the fatigue coefficient formula to calculate the fatigue degree of construction workers, the impact of fatigue on construction risks is considered, and then a quantitative assessment of the impact of construction workers' fatigue is achieved, which effectively solves the problem of ignoring the impact of construction workers' fatigue status on construction safety in the existing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flow chart of a construction risk prediction method based on a machine learning algorithm provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] The embodiment of the present application solves the problems of low reliability of risk prediction and inaccurate early warning in the prior art by providing a construction risk prediction method based on a machine learning algorithm. By collecting and preprocessing construction and personnel data, extracting influencing factors, calculating risks and fatigue coefficients, and constructing a prediction model, reliable prediction of construction risks can be achieved, thereby improving the reliability and practicality of risk prediction.

[0020] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0021] like Figure 1 As shown, it is a flow chart of a construction risk prediction method based on a machine learning algorithm provided by an embodiment of the present application, the method comprising the following steps: obtaining current construction data and construction data when historical risks occur, and preprocessing them to obtain the preprocessed current construction data and the preprocessed construction data when historical risks occur; obtaining the current construction risk coefficient and the historical construction risk coefficient based on the preprocessed current construction data and the preprocessed construction data when historical risks occur, respectively; obtaining the work data of current construction personnel and the work data of historical construction personnel, and preprocessing them to obtain the preprocessed work data of current construction personnel and the preprocessed work data of historical construction personnel; obtaining the fatigue coefficient of current construction personnel and the fatigue coefficient of historical construction personnel based on the preprocessed work data of current construction personnel and the preprocessed work data of historical construction personnel, respectively; constructing a construction risk prediction model based on the preprocessed construction data when historical risks occur, the historical construction risk coefficient, the preprocessed work data of historical construction personnel, and the fatigue coefficient of historical construction personnel; and predicting construction risks through the construction risk prediction model based on the preprocessed current construction data, the current construction risk coefficient, the preprocessed work data of current construction personnel, and the current construction personnel fatigue coefficient.

[0022] In this embodiment, the construction risk coefficient is used to quantify the potential risk level in the construction process and can provide real-time risk warnings for construction management personnel; while the construction worker fatigue coefficient helps determine whether the construction workers are working in a fatigued state by evaluating their working status, thereby preventing safety problems caused by worker fatigue.

[0023] By integrating multi-dimensional information such as current construction data, historical construction data, and the working status of construction personnel, a more comprehensive perspective is provided for risk prediction during the construction process. Through preprocessing and machine learning algorithms, potential risk factors can be efficiently identified, improving the accuracy of risk management during the construction process.

[0024] Furthermore, the current construction risk coefficient and the historical construction risk coefficient are obtained in the following manner: based on the pre-processed current construction data and the pre-processed construction data when the historical risk occurred, the weather data during the current construction and the historical risk occurred, the geological data of the construction location, the construction equipment data, and the operation data of the construction personnel are extracted respectively; the weight factors of the weather data during the construction, the geological data of the construction location, the construction equipment data, and the operation data of the construction personnel for the construction risk coefficient are obtained from the database; the current construction risk coefficient and the historical construction risk coefficient are calculated respectively by the construction risk coefficient formula; the construction risk coefficient formula is: ; In the formula, is the construction risk factor, is the temperature influence coefficient, is the formation stability influence coefficient, is the equipment failure influence coefficient, is the influence coefficient of personnel operation behavior, a is right The weight factor of right The weight factor, c is right The weight factor d is right The weight factor of .

[0025] In this embodiment, if the construction risk coefficient value is large, it means that the construction is facing a high risk. This may be caused by bad weather conditions, poor geological conditions at the construction location, frequent failures of construction equipment, or many irregular operations by construction personnel, which increases the construction risk.

[0026] Furthermore, the temperature influence coefficient is obtained by: obtaining the real-time temperature of the kth temperature monitoring point in the construction site through the temperature sensor, obtaining the historical average temperature of the corresponding monitoring point in the current season, the allowable difference of the temperature fluctuation of the corresponding monitoring point, the total number of temperature monitoring points, and performing preprocessing; obtaining the weight factor of the real-time temperature of the kth temperature monitoring point on the temperature influence coefficient from the database; obtaining the temperature influence coefficient through the temperature influence coefficient formula; the temperature influence coefficient formula is: ; In the formula, is the real-time temperature of the kth temperature monitoring point in the construction site, is the historical average temperature of the corresponding monitoring point in the current season, is the allowable difference of temperature fluctuation at the corresponding monitoring point, is the weight of the kth monitoring point, and m is the total number of temperature monitoring points.

[0027] In this embodiment, when the temperature influence coefficient value is small, it indicates that the real-time temperature of the construction site is relatively stable and close to the historical average temperature, and the temperature fluctuation is within the allowable range. This means that the temperature condition has little impact on the construction, and the performance of the construction materials will not change due to temperature problems.

[0028] The steps for obtaining the allowable temperature fluctuation difference of the corresponding monitoring point are as follows: obtain the temperature data of the same season of the monitoring point in the past three years through the local meteorological bureau, calculate the difference between the maximum and minimum temperature data of the same season of the monitoring point in the past three years to obtain the allowable temperature fluctuation difference of the corresponding monitoring point.

[0029] Furthermore, the method for obtaining the formation stability influence coefficient is as follows: obtaining the soil compressive strength of the construction site through a sensor, obtaining the soil compressive strength threshold from the database according to the construction type, and obtaining the formation stability influence coefficient through the formation stability influence coefficient formula; the formation stability influence coefficient formula is: ; In the formula, is the compressive strength of the soil at the construction site, is the soil compressive strength threshold, and e is a natural constant.

[0030] In this embodiment, when the stratum stability influence coefficient is larger, it indicates that the soil compressive strength of the construction site is much higher than the soil compressive strength threshold, which means that the stratum of the construction site has strong stability and can provide a solid and reliable foundation support for the construction.

[0031] Furthermore, the equipment failure influence coefficient is obtained by: collecting the number of failures of equipment at the construction site, obtaining the upper limit of the total number of failures of the equipment within the normal use cycle, and the total number of equipment types; obtaining the weight factor of the number of failures of each type of equipment on the equipment failure influence coefficient from the database; obtaining the equipment failure influence coefficient through the equipment failure influence coefficient formula; the equipment failure influence coefficient is: ; In the formula, is the number of failures of the qth type of equipment, is the upper limit of the total number of failures of the qth type of equipment in the normal use cycle, is the number of failures of the qth type of equipment The weight factor is , and r is the total number of device types.

[0032] In this embodiment, when the equipment failure impact coefficient value is low, it means that the actual number of failures of various equipment at the construction site is very small relative to the upper limit of the total number of failures within its normal use cycle. This indicates that the overall operation of the construction equipment is in good condition and the stability and reliability of the equipment are high.

[0033] The upper limit of the total number of equipment failures during its normal use cycle can be determined based on the theoretical number of equipment failures provided by the equipment manufacturer, actual failure statistics of similar equipment in the industry, and the operating environment and operating conditions of the equipment in this project.

[0034] Furthermore, the steps for obtaining the influence coefficient of personnel operation behavior are as follows: obtaining the maximum force applied by the construction personnel on the operating tool in each operation process through the pressure sensor, and obtaining the standard applied force in each operation process; obtaining the weight factor of the influence coefficient of the maximum force applied on the operating tool in each operation process on the equipment failure from the database; obtaining the influence coefficient of personnel operation behavior through the influence coefficient formula of personnel operation behavior; the influence coefficient formula of personnel operation behavior is: The influence coefficient formula of personnel operation behavior is: ; In the formula, is the maximum force applied by the construction worker on the operating tool during the nth operation. The standard applied force during the nth operation, is the weight factor of the influence coefficient of the maximum force applied to the operating tool in the nth operating process on the operating behavior of the personnel, and s is the total number of operating processes.

[0035] In this embodiment, the greater the value of the personnel operation behavior influence coefficient is, the closer the operation of the construction personnel is to the standard operation during the construction process, which means that the construction personnel are skilled and standardized in operation and can perform the operation according to the established construction standards and processes.

[0036] Furthermore, the steps for obtaining the fatigue coefficient of the current construction workers and the fatigue coefficient of the historical construction workers are as follows: based on the pre-processed work data of the current construction workers and the pre-processed work data of the historical construction workers, extract the current construction workers' continuous working time, the current construction workers' working time period, the historical construction workers' continuous working time, and the historical construction workers' working time period; obtain the weight factors of the construction workers' continuous working time and the construction workers' working time period for the construction workers' fatigue coefficient from the database; obtain the longest continuous working time stipulated at the construction site, the actual working time of the construction workers in the current time period at the construction site, and the ideal working time of the construction workers in the current time period at the construction site; obtain the current construction workers' fatigue coefficient and the historical construction workers' fatigue coefficient through the construction workers' fatigue coefficient formula.

[0037] In this embodiment, by calculating the fatigue coefficient of the construction workers, it is possible to timely find out whether the construction workers are in an over-fatigue state. Knowing the fatigue status in advance can effectively avoid construction safety accidents caused by fatigue work, such as the risk of falling due to fatigue during high-altitude work, collision accidents caused by misoperation due to fatigue when operating heavy machinery, etc., to ensure the life safety of construction workers.

[0038] The fatigue coefficient of construction workers is an important reference indicator for construction parties to dispatch personnel. When the fatigue coefficient of personnel in a construction area is high, energetic personnel can be deployed from other areas in a timely manner to provide support to ensure the continuity and efficiency of construction. At the same time, it is also possible to reasonably arrange the rotation of personnel according to the fatigue coefficient to ensure that construction workers have enough rest time to recover their physical strength.

[0039] Furthermore, the fatigue coefficient formula of the construction workers is: ; In the formula, F is the fatigue coefficient of the construction workers, t is the continuous working time of the construction workers, The maximum continuous working time is specified. is the weight factor of continuous working time on F, represents the actual working hours of the construction workers in the i-th working period, is the ideal working time in the i-th working period, is the duration of the working period. is the weight factor of the working time deviation within the working time period to F, and v is the total number of working time periods divided in a construction worker's work in one day.

[0040] In this embodiment, when the fatigue coefficient value of the construction workers is lower, it means that the construction workers are currently in a relatively relaxed working state, with less fatigue accumulation. They can maintain a high level of concentration and operational accuracy during the construction process, and the possibility of mistakes or accidents due to fatigue is lower.

[0041] Furthermore, constructing a construction risk prediction model includes the following steps: dividing a training set and a test set based on the pre-processed historical risk occurrence construction data, the historical construction risk coefficient, the historical risk occurrence construction data, and the historical construction personnel fatigue coefficient; inputting the training set into a polynomial regression model for training to obtain a preliminary construction risk prediction model; inputting the test set into the preliminary construction risk prediction model for testing to obtain a construction risk prediction model.

[0042] Furthermore, the construction risk prediction includes the following steps: inputting the pre-processed current construction data, the current construction risk coefficient, the work data of the current construction personnel, and the fatigue coefficient of the current construction personnel into the construction risk prediction model; the model predicts according to the input data, and outputs a quantitative value representing the construction risk, the value range is between 0 and 1, where 0 represents no risk and 1 represents the highest risk; according to the output risk quantification value, it is mapped to a specific risk level, and thresholds A, B, C, and D are set. The risk quantification value is not less than threshold A and is lower than threshold B for low risk, the risk quantification value is not less than threshold B and is lower than threshold C for medium risk, and the risk quantification value is not less than threshold C and is lower than threshold D for high risk; when the prediction result is low risk, routine monitoring is required; when it is medium risk, the construction process needs to be adjusted; when it is high risk, an alarm to suspend construction is issued to notify relevant personnel.

[0043] In this embodiment, the risk quantification value is mapped to a specific risk level, and different thresholds are set to divide the risk into low, medium and high levels. This hierarchical management method enables construction managers to take targeted measures according to different risk levels.

[0044] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0045] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0046] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0047] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0048] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0049] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A construction risk prediction method based on machine learning algorithm, characterized in that: The following steps are involved: Obtain current construction data and construction data when historical risks occurred, and perform preprocessing to obtain preprocessed current construction data and preprocessed construction data when historical risks occurred; Based on the pre-processed current construction data and the pre-processed construction data when historical risks occurred, the current construction risk coefficient and the historical construction risk coefficient are obtained respectively; Acquire the work data of the current construction personnel and the work data of the historical construction personnel, and perform preprocessing to obtain the preprocessed work data of the current construction personnel and the preprocessed work data of the historical construction personnel; Based on the pre-processed work data of the current construction personnel and the pre-processed work data of the historical construction personnel, the fatigue coefficient of the current construction personnel and the fatigue coefficient of the historical construction personnel are obtained respectively; A construction risk prediction model is constructed based on the pre-processed historical construction data when risks occurred, the historical construction risk coefficient, the pre-processed historical construction personnel work data, and the historical construction personnel fatigue coefficient; Based on the pre-processed current construction data, the current construction risk coefficient, the pre-processed work data of the current construction personnel, and the current construction personnel fatigue coefficient, the construction risk prediction model is used to perform construction risk prediction.

2. A construction risk prediction method based on a machine learning algorithm as claimed in claim 1, characterized in that: The current construction risk coefficient and historical construction risk coefficient are obtained as follows: Based on the pre-processed current construction data and the pre-processed construction data when historical risks occurred, the weather data, geological data of the construction location, construction equipment data, and construction personnel operation data of the current construction and the historical risks occurred are extracted respectively; Obtain the weighting factors of weather data during construction, geological data of the construction location, construction equipment data, and operation data of construction personnel on the construction risk coefficient from the database; The current construction risk coefficient and the historical construction risk coefficient are calculated respectively through the construction risk coefficient formula; The construction risk coefficient formula is: ； In the formula, is the construction risk factor, is the temperature influence coefficient, is the formation stability influence coefficient, is the equipment failure influence coefficient, is the influence coefficient of personnel operation behavior, a is right The weight factor of right The weight factor, c is right The weight factor d is right The weight factor of .

3. A construction risk prediction method based on a machine learning algorithm as claimed in claim 2, characterized in that: The temperature influence coefficient is obtained as follows: The real-time temperature of the kth temperature monitoring point in the construction site is obtained through the temperature sensor, the historical average temperature of the corresponding monitoring point in the current season, the allowable difference of the temperature fluctuation of the corresponding monitoring point, the total number of temperature monitoring points, and preprocessing is performed; Obtain the weight factor of the real-time temperature of the kth temperature monitoring point on the temperature influence coefficient from the database; The temperature influence coefficient is obtained through the temperature influence coefficient formula; The temperature influence coefficient formula is: ； In the formula, is the real-time temperature of the kth temperature monitoring point in the construction site, is the historical average temperature of the corresponding monitoring point in the current season, is the allowable difference of temperature fluctuation at the corresponding monitoring point, is the weight of the kth monitoring point, and m is the total number of temperature monitoring points.

4. A construction risk prediction method based on a machine learning algorithm as claimed in claim 2, characterized in that: The method for obtaining the formation stability influence coefficient is: The soil compressive strength of the construction site is obtained through sensors, and the soil compressive strength threshold is obtained from the database according to the construction type. Obtain formation stability influence coefficient through formation stability influence coefficient formula; The formation stability influence coefficient formula is: ； In the formula, is the compressive strength of the soil at the construction site, is the soil compressive strength threshold, and e is a natural constant.

5. A construction risk prediction method based on a machine learning algorithm as claimed in claim 2, characterized in that: The equipment failure impact coefficient is obtained as follows: Collect the number of equipment failures at the construction site, obtain the upper limit of the total number of equipment failures within the normal use cycle, and the total number of equipment types; Obtain the weight factor of the number of failures of each device on the device failure influence coefficient from the database; Obtain the equipment failure impact coefficient through the equipment failure impact coefficient formula; The equipment failure impact coefficient is: ； In the formula, is the number of failures of the qth type of equipment, is the upper limit of the total number of failures of the qth type of equipment in the normal use cycle, is the number of failures of the qth type of equipment The weight factor is , and r is the total number of device types.

6. A construction risk prediction method based on a machine learning algorithm as claimed in claim 2, characterized in that: The steps to obtain the influence coefficient of personnel operation behavior are as follows: The maximum force applied by the construction personnel on the operating tool during each operation is obtained through the pressure sensor, and the standard applied force during each operation is obtained; Obtaining from the database the weight factor of the influence coefficient of the maximum force applied to the operating tool in each operation process on the equipment failure; Obtain the personnel operation behavior influence coefficient through the personnel operation behavior influence coefficient formula; The formula for the influence coefficient of personnel operation behavior is: The formula for the influence coefficient of personnel operation behavior is: ； In the formula, is the maximum force applied by the construction worker on the operating tool during the nth operation. The standard applied force during the nth operation, is the weight factor of the influence coefficient of the maximum force applied to the operating tool in the nth operating process on the operating behavior of the personnel, and s is the total number of operating processes.

7. A construction risk prediction method based on a machine learning algorithm as claimed in claim 1, characterized in that: The steps for obtaining the current construction worker fatigue coefficient and the historical construction worker fatigue coefficient are as follows: Based on the pre-processed work data of the current construction personnel and the pre-processed work data of the historical construction personnel, the continuous working time of the current construction personnel, the working time period of the current construction personnel, the continuous working time of the historical construction personnel, and the working time period of the historical construction personnel are extracted; Obtain the weight factors of the continuous working time and working time period of the construction workers on the fatigue coefficient of the construction workers from the database; Obtain the maximum continuous working time specified at the construction site, the actual working time of the construction workers in the current time period at the construction site, and the ideal working time of the construction workers in the current time period at the construction site; The current construction worker fatigue coefficient and the historical construction worker fatigue coefficient are obtained through the construction worker fatigue coefficient formula.

8. A construction risk prediction method based on a machine learning algorithm as claimed in claim 7, characterized in that: The fatigue coefficient formula of the construction personnel is: ； In the formula, F is the fatigue coefficient of the construction workers, t is the continuous working time of the construction workers, The maximum continuous working time is specified. is the weight factor of continuous working time on F, represents the actual working hours of the construction workers in the i-th working period, is the ideal working time in the i-th working period, is the duration of the working period. is the weight factor of the working time deviation within the working time period to F, and v is the total number of working time periods divided in a construction worker's work in one day.

9. A construction risk prediction method based on a machine learning algorithm as claimed in claim 1, characterized in that: Building a construction risk prediction model includes the following steps: Based on the pre-processed historical risk occurrence construction data, historical construction risk coefficient, historical risk occurrence construction data, and historical construction personnel fatigue coefficient, the training set and the test set are divided; The training set is input into the polynomial regression model for training, and a preliminary prediction model for construction risk is obtained; The test set is input into the preliminary construction risk prediction model for testing to obtain the construction risk prediction model.

10. A construction risk prediction method based on a machine learning algorithm as claimed in claim 9, characterized in that: The construction risk forecasting includes the following steps: Input the pre-processed current construction data, current construction risk coefficient, current construction personnel's work data, and current construction personnel's fatigue coefficient into the construction risk prediction model; The model makes predictions based on the input data and outputs a quantitative value representing the construction risk, which ranges from 0 to 1, where 0 represents no risk and 1 represents the highest risk; According to the outputted risk quantification value, it is mapped to a specific risk level, and thresholds A, B, C, and D are set. If the risk quantification value is not less than threshold A and is less than threshold B, it is low risk; if the risk quantification value is not less than threshold B and is less than threshold C, it is medium risk; if the risk quantification value is not less than threshold C and is less than threshold D, it is high risk; When the predicted result is low risk, routine monitoring is required; when it is medium risk, the construction process needs to be adjusted; when it is high risk, an alarm to suspend construction is issued to notify relevant personnel.

Images