Intelligent construction site safety management platform and method thereof

Abstract

The invention discloses an intelligent construction site safety management platform and a method thereof. The management platform comprises a data acquisition module, a data processing module, a machine learning prediction module, an intelligent sonar analysis module, an early warning module, a display and notification module and a feedback and learning module. The machine learning prediction module comprises a dynamic feature selection unit, a multi-model integration unit, a context awareness reasoning unit, a self-learning and model updating unit and an uncertainty evaluation unit; the intelligent sonar analysis module comprises a multi-frequency sonar signal processing unit, an environment perception and compensation unit, a target identification and classification unit, a dynamic signal enhancement and noise reduction unit, a real-time position tracking and trajectory prediction unit and a self-learning and system optimization unit. According to the invention, an early warning signal can be rapidly and accurately sent out in a high-risk environment, in addition, an early warning decision value can be adjusted in real time, a potential dangerous situation can be rapidly responded, and the safety risk on a construction site is effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of construction site safety management, and in particular to a smart construction site safety management platform and method thereof. Background Art

[0002] In the current construction industry, as the scale and complexity of engineering projects continue to expand, construction site safety management faces unprecedented challenges. Especially in water construction sites involving ship operations, such as port construction and cross-sea bridge construction, the navigation safety of ships is directly related to the safety and stability of the entire construction site.

[0003] However, traditional safety management methods have obvious shortcomings in predicting ship behavior. Since ship navigation is affected by many factors, such as wind direction, current, tide, and the ship's own performance status, its navigation trajectory and behavior pattern are highly complex and uncertain. Traditional methods are difficult to accurately capture these dynamic changes, resulting in low prediction accuracy of ship behavior, and unable to provide timely and effective decision-making support for managers, thereby failing to effectively reduce safety risks. Summary of the invention

[0004] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a smart construction site safety management platform, mainly comprising:

[0006] A data acquisition module, which is responsible for collecting real-time dynamic data of the ship, historical navigation data, environmental data, and real-time sound wave data from the intelligent sonar system;

[0007] A data processing module, wherein the data processing module is used to pre-process the collected multi-source data;

[0008] A machine learning prediction module, wherein the machine learning prediction module is used to predict the future behavior of the ship;

[0009] An intelligent sonar analysis module, which is used to analyze underwater sound wave reflections and monitor the precise position of the ship;

[0010] An early warning module, which is used to integrate the results from the machine learning prediction module and the intelligent sonar analysis module, and decide whether to issue an early warning message according to predetermined rules and algorithms;

[0011] A display and notification module, which is used to present the warning information in a visual form on the monitoring platform of the construction site management personnel and send alarm notifications to relevant personnel;

[0012] Feedback and learning module, which is used to collect actual feedback after early warning information, and use these feedback data to optimize and update the machine learning module to improve future prediction accuracy.

[0013] As a preferred solution of the smart construction site safety management platform of the present invention, wherein: the machine learning prediction module includes a dynamic feature selection unit, a multi-model integration unit, a context-aware reasoning unit, a self-learning and model updating unit, and an uncertainty assessment unit;

[0014] The dynamic feature selection unit can automatically select the most critical features for ship behavior prediction from a large number of input features, and can dynamically adjust the feature selection strategy according to real-time environmental changes;

[0015] The multi-model integration unit integrates a plurality of different prediction models and automatically selects or weightedly combines the prediction results of different models according to the characteristics of the input data to improve the accuracy and robustness of the prediction;

[0016] The context-aware reasoning unit is used to combine current environmental data and ship behavior history to analyze the current navigation scenario in real time and perform context-aware reasoning on possible actions that the ship may take;

[0017] The self-learning and model updating unit continuously updates and optimizes the prediction model by continuously monitoring and analyzing system feedback data;

[0018] The uncertainty assessment unit is used to perform uncertainty assessment on each prediction result and provide a confidence interval for the prediction result.

[0019] As a preferred solution of the smart construction site safety management platform of the present invention, the intelligent sonar analysis module includes a multi-frequency sonar signal processing unit, an environmental perception and compensation unit, a target recognition and classification unit, a dynamic signal enhancement and noise reduction unit, a real-time position tracking and trajectory prediction unit, and a self-learning and system optimization unit;

[0020] The multi-frequency sonar signal processing unit is used to receive and process signals from the multi-frequency sonar, distinguish the acoustic characteristics of different objects by separating the sound wave reflection signals of different frequency bands, and especially to accurately identify the type, size and position of the ship forever;

[0021] The environmental perception and compensation unit is used to perceive the underwater environmental conditions in real time and to compensate and correct the sonar signals accordingly, ensuring that the sonar data still has high reliability and accuracy in various complex environments;

[0022] The target recognition and classification unit recognizes and classifies the detected targets based on the processed sonar signals, such as distinguishing between ships, marine life, fixed structures, etc.

[0023] The dynamic signal enhancement and noise reduction unit focuses on enhancing useful sonar signals and suppressing noise;

[0024] The real-time position tracking and trajectory prediction unit is used to track the current position of the ship in real time and predict the future trajectory of the ship based on current and historical data;

[0025] The self-learning and system optimization unit continuously updates and optimizes the sonar analysis algorithm by analyzing the results of each sonar signal processing and target recognition, so that the system can adapt to the ever-changing environment and target characteristics.

[0026] As a preferred solution of the smart construction site safety management platform of the present invention, the early warning module uses an early warning decision algorithm to decide whether to issue an early warning message. The formula of the early warning decision algorithm is as follows:

[0027]

[0028] in:

[0029] R t is the warning decision value, representing the comprehensive warning intensity in the time period t0 to t;

[0030] P(X) i It is the behavior prediction function after multi-model integration, and outputs the predicted value y of the ship's future behavior t , represents the predicted value of the ith model for the feature vector X, where X is the feature selection function F(D t ,H t ,E t ,S t ) provides the feature vector, where D t is a collection of real-time dynamic data, including ship speed, direction, H t is a collection of historical navigation data, including past navigation paths. Ship behavior, E t is a collection of environmental data, including wind speed, tide, and current. t Real-time sound wave data of the intelligent sonar system, including the distance, direction and speed of the target;

[0031] G(S t ,E t ) i It is an intelligent sonar analysis function that outputs the current trajectory and predicted trajectory of the ship. t , combined with U(y t ,zt ) Evaluate uncertainty σ t ;

[0032] is an exponential decay factor, where σ t is the uncertainty assessment value, which indicates the uncertainty of the prediction result. This factor is used to adjust the influence of the prediction result, based on the probability density function of the Gaussian distribution;

[0033] is the normalization factor in the Gaussian distribution, ensuring that the integral value of the probability density function is 1, and is used to adjust Make it conform to the form of standard normal distribution;

[0034] W(λ t ) is the early warning decision weight function, which is used to calculate the decision influence factor λ t The weight given to the warning result, λ t is through situational awareness and reasoning function I(y t ,z t ,σ t ), based on the current situation, the possible actions that the ship may take in the future are inferred, and the decision influencing factor λ is output in combination with the uncertainty assessment value. t ;

[0035] λ t is the decision-making influencing factor, comprehensively considering the future behavior prediction value y t , trajectory prediction value z t and uncertainty estimate σ t The value generated after the warning, which reflects the impact of the current situation on the warning;

[0036] Value range meaning:

[0037] The output of the formula R t Represents the warning intensity in the time period t0 to t. The larger the value, the higher the urgency and necessity of the warning. t If the preset threshold value θ is exceeded, the system will trigger an early warning notification.

[0038] As a preferred solution of the smart construction site safety management platform of the present invention, the specific circumstances in which the early warning decision module issues early warning information are as follows:

[0039] When the uncertainty assessment value σ t Less than a critical value σ c When the warning decision value R t If it is greater than the threshold value θ, the system will immediately issue an early warning;

[0040] When the uncertainty assessment value σ t Equal to or greater than a critical value σ cWhen R t The growth rate of the economy slowed down and early warnings were delayed or not issued;

[0041] When the decision influencing factor λ t Greater than a certain threshold λ c When the warning weight W(λ t ) will increase rapidly, making the warning decision value R t Significantly higher than the threshold θ;

[0042] When the decision influencing factor λ t Less than or equal to a threshold λ c When the warning weight W(λ t ) will decrease, making the warning decision value R t The growth rate slows down or stops;

[0043] When the ship behavior prediction value y t Equal to the trajectory prediction value z t When the warning decision value R t The growth rate reaches the highest level, and the warning signal is immediately issued;

[0044] When the ship behavior prediction value y t and trajectory prediction value z t When they are not equal, the system will t Re-evaluation, uncertain σ t increase, thereby delaying the issuance of warnings.

[0045] As a preferred solution of the smart construction site safety management platform of the present invention, the dynamic feature selection unit includes a feature screening mechanism based on fuzzy logic, which is used to dynamically adjust the priority of feature selection through fuzzy membership function in real-time environmental changes, thereby improving the prediction accuracy in a changing environment.

[0046] As a preferred solution of the smart construction site safety management platform of the present invention, the early warning information includes visual warnings, sound alarms and notifications to relevant personnel.

[0047] As a preferred solution of the smart construction site safety management platform of the present invention, the early warning module includes a multi-level decision-making unit, which calculates the decision-making influence factor λ t The changing trend of the warning decision value R is adjusted in real time t The growth rate enables the early warning module to issue early warning signals more quickly in high-risk environments based on the results of ship behavior prediction and sonar analysis.

[0048] As a preferred solution of the smart construction site safety management platform of the present invention, the early warning module also includes a collaborative evaluation unit, which compares the ship behavior prediction value y t and trajectory prediction value z t The consistency of y t and z t Equal and uncertainty σ t When the two are inconsistent, the warning module will delay the warning until a re-evaluation results in a higher confidence level.

[0049] The management method using the above-mentioned smart construction site safety management platform includes the following steps:

[0050] Step 1: Start the data acquisition module and configure the data acquisition equipment to receive real-time dynamic data, historical navigation data, environmental data and real-time acoustic wave data from the ship's intelligent sonar system; obtain data from various data sources regularly or on demand, and ensure the integrity and accuracy of the data; store the collected raw data in the data storage system for subsequent processing and analysis;

[0051] Step 2: Start the data processing module to pre-process the collected multi-source data, including data cleaning, denoising, normalization and format conversion; integrate information from different data sources and convert it into a format suitable for machine learning and intelligent sonar analysis modules;

[0052] Step 3: Start the machine learning prediction module and perform dynamic feature selection to screen out the most critical features for ship behavior prediction; use the multi-model integration unit to select or weightedly combine the results of different prediction models according to the characteristics of the input data; perform situational awareness reasoning to predict the future behavior of the ship by combining the current environmental data and historical behavior; start the self-learning and model update unit to continuously update and optimize the prediction model based on system feedback; use the uncertainty assessment unit to assess the uncertainty of the prediction results and determine the confidence interval of the prediction results;

[0053] Step 4: Start the intelligent sonar analysis module to process multi-frequency sonar signals, separate and analyze sound wave reflection signals in different frequency bands; use the environmental perception and compensation unit to compensate and correct the sonar signals to ensure data accuracy; identify and classify underwater targets, including ships, marine life and fixed structures, through the target recognition and classification unit; the dynamic signal enhancement and noise reduction unit enhances useful signals and suppresses noise to improve signal quality; use the real-time position tracking and trajectory prediction unit to track the current position of the ship and predict its future trajectory; start the self-learning and system optimization unit to continuously update the sonar analysis algorithm to improve system adaptability and accuracy;

[0054] Step 5: Start the warning module, calculate the warning decision value, and apply the warning decision algorithm formula based on the results of the machine learning prediction module and the intelligent sonar analysis module; evaluate whether the calculated warning decision value exceeds the preset threshold; if the warning decision value exceeds the threshold, trigger the warning signal, including visual warning, sound alarm and notification to relevant personnel; adjust the warning decision according to the specific situation of the warning information, taking into account the changing trend of the uncertainty assessment value and decision influencing factors;

[0055] Step 6: Start the display and notification module to present the warning information in a visual form on the monitoring platform of the construction site manager; send alarm notifications to relevant personnel to ensure that they receive the warning information in time and take corresponding measures; provide a visual dashboard to display the current ship position, predicted trajectory, warning status and historical warning records;

[0056] Step 7: Start the feedback and learning module to collect feedback information after the actual warning, including the accuracy of the warning and the system response; analyze the feedback data and use this data to optimize and update the machine learning prediction module; adjust and optimize the prediction model based on the feedback information to improve the accuracy of future predictions and the overall performance of the system.

[0057] Beneficial effects of the present invention:

[0058] 1. The smart construction site safety management platform of the present invention realizes high-precision prediction and real-time monitoring of ship behavior by integrating technologies such as dynamic feature selection, machine learning prediction, multi-model integration and intelligent sonar analysis. In particular, the multi-level decision-making unit and collaborative evaluation unit of the early warning module enable the platform to quickly and accurately issue early warning signals in high-risk environments. In addition, by comprehensively considering the predicted value and sonar analysis results, the system can adjust the early warning decision value in real time, quickly respond to potential dangerous situations, and effectively reduce safety risks on the construction site.

[0059] 2. The present invention can improve the adaptability of the platform in complex and changeable environments by introducing a feature screening mechanism based on fuzzy logic and a multi-frequency signal processing function of intelligent sonar. The dynamic feature selection unit can automatically adjust the feature selection strategy according to real-time environmental changes, thereby improving the accuracy of the prediction. The intelligent sonar analysis module ensures high reliability and accuracy in various environments by sensing underwater environmental conditions in real time and compensating and correcting sonar signals. The application of these technologies enables the system to intelligently adapt to environmental changes and continuously optimize itself based on new data, thereby providing more accurate early warning services and enhancing the intelligence level and application scope of the platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0061] Figure 1 This is a structural block diagram of a smart construction site safety management platform of the present invention. DETAILED DESCRIPTION

[0062] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0063] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0064] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0065] Secondly, the present invention is described in detail with reference to the schematic diagram. When describing the embodiments of the present invention in detail, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0066] Example 1

[0067] Reference Figure 1 , which is the first embodiment of the present invention, provides a smart construction site safety management platform, which mainly includes:

[0068] A data acquisition module, which is responsible for collecting real-time dynamic data of the ship, historical navigation data, environmental data (such as wind speed, tide, visibility), and real-time sound wave data from the intelligent sonar system;

[0069] When in use, the data acquisition module is connected to sensors, cameras, AIS systems, weather stations and sonar equipment to transmit the collected data to the data processing module.

[0070] A data processing module is used to pre-process the collected multi-source data, including data cleaning, format conversion, data fusion and other operations to ensure the consistency and integrity of the data;

[0071] When in use, the data processing module receives data from the data acquisition module and transmits the processed data to the machine learning prediction module and the intelligent sonar analysis module.

[0072] A machine learning prediction module, which is used to predict the future behavior of the ship, such as turning, deceleration or acceleration. The prediction results are used to determine in advance whether the ship is likely to enter the construction area and issue an early warning;

[0073] When in use, the machine learning prediction module receives data from the data processing module, makes predictions based on historical data and current environmental conditions, and transmits the prediction results to the early warning module.

[0074] An intelligent sonar analysis module, which uses intelligent sonar technology to analyze underwater sound wave reflections, especially to detect the precise position of ships in haze, at night or in low visibility conditions. The module determines the speed and direction of the ship through sonar data to enhance early warning capabilities in harsh environments;

[0075] When in use, the intelligent sonar analysis module receives sonar data from the data processing module, performs real-time analysis, and transmits the analysis results to the early warning module.

[0076] An early warning module, which integrates the results from the machine learning prediction module and the intelligent sonar analysis module, and decides whether to issue an early warning message based on predetermined rules and algorithms. The early warning message may include visual warnings, sound alarms, and notifications to relevant personnel;

[0077] When in use, the warning module receives data from the machine learning prediction module and the intelligent sonar analysis module, makes judgments based on the synthesized results, and transmits the warning information to the display and notification module.

[0078] A display and notification module, which is used to present the warning information in a visual form on the monitoring platform of the construction site management personnel and send alarm notifications to relevant personnel through SMS, email, etc.;

[0079] When in use, the display and notification module receives the warning information from the warning module and presents it to the user in the form of graphics, text or audio.

[0080] Feedback and learning module, which is used to collect actual feedback after warning information, such as the response of the ship, whether an accident has occurred, etc., and use these feedback data to optimize and update the machine learning model to improve the accuracy of future predictions;

[0081] When in use, the feedback and learning module receives feedback data from the display and notification module, and transmits it to the machine learning prediction module after processing to continuously optimize the algorithm.

[0082] Specifically, the machine learning prediction module includes a dynamic feature selection unit, a multi-model integration unit, a context-aware reasoning unit, a self-learning and model updating unit, and an uncertainty assessment unit;

[0083] The dynamic feature selection unit can automatically select the most critical features for ship behavior prediction from a large number of input features, and can dynamically adjust the feature selection strategy according to real-time environmental changes, such as wind speed, tide, ship load and other factors, to ensure the optimal feature combination for prediction model training and reasoning;

[0084] The dynamic feature selection unit includes a feature screening mechanism based on fuzzy logic, which is used to dynamically adjust the priority of feature selection through fuzzy membership function in real-time environmental changes, thereby improving the prediction accuracy in a changing environment;

[0085] The multi-model integration unit integrates multiple different prediction models (such as time series models, deep neural networks, Bayesian networks, etc.), and automatically selects or weightedly combines the prediction results of different models according to the characteristics of the input data to improve the accuracy and robustness of the prediction;

[0086] The situational awareness reasoning unit is used to combine the current environmental data and the ship's behavior history to analyze the current navigation scenario in real time and perform situational awareness reasoning on the actions that the ship may take. For example, in strong winds, the prediction unit will be more inclined to consider the possibility of the ship slowing down or adjusting its course.

[0087] The self-learning and model updating unit continuously updates and optimizes the prediction model by continuously monitoring and analyzing system feedback data. The actual situation after each warning is input into the unit as new training data to help the model learn new patterns and trends.

[0088] The uncertainty assessment unit is used to perform uncertainty assessment on each prediction result and provide a confidence interval for the prediction result. This unit can help the early warning module determine the reliability of the prediction result, thereby deciding whether to issue an early warning or conduct further analysis.

[0089] Specifically, the intelligent sonar analysis module includes a multi-frequency sonar signal processing unit, an environment perception and compensation unit, a target recognition and classification unit, a dynamic signal enhancement and noise reduction unit, a real-time position tracking and trajectory prediction unit, and a self-learning and system optimization unit;

[0090] The multi-frequency sonar signal processing unit is used to receive and process signals from the multi-frequency sonar, distinguish the acoustic characteristics of different objects by separating the sound wave reflection signals of different frequency bands, and especially to accurately identify the type, size and position of the ship forever;

[0091] The environmental sensing and compensation unit is used to sense underwater environmental conditions in real time, such as water temperature, salinity, noise, etc., and to compensate and correct the sonar signal accordingly, ensuring that the sonar data still has high reliability and accuracy in various complex environments;

[0092] The target recognition and classification unit recognizes and classifies detected targets based on the processed sonar signals, such as distinguishing between ships, marine life, fixed structures, etc. The unit uses a deep learning model combined with acoustic features for accurate classification;

[0093] The dynamic signal enhancement and noise reduction unit focuses on enhancing useful sonar signals and suppressing noise, especially in low visibility or high noise environments, ensuring signal clarity and reliability;

[0094] The real-time position tracking and trajectory prediction unit is used to track the current position of the ship in real time and predict the future trajectory of the ship based on current and historical data. This unit is highly integrated with the machine learning prediction module to jointly complete the estimation of the possible path of the ship;

[0095] The self-learning and system optimization unit continuously updates and optimizes the sonar analysis algorithm by analyzing the results of each sonar signal processing and target recognition, so that the system can adapt to the ever-changing environment and target characteristics.

[0096] Specifically, the warning module uses a warning decision algorithm to decide whether to issue a warning message. The formula of the warning decision algorithm is as follows:

[0097]

[0098] in:

[0099] R t is the warning decision value, representing the comprehensive warning intensity in the time period t0 to t;

[0100] P(X0 i It is the behavior prediction function after multi-model integration, and outputs the predicted value y of the ship's future behavior t, represents the predicted value of the ith model for the feature vector X, where X is the feature selection function F(D t ,H t ,E t ,S t ) provides the feature vector, where D t is a collection of real-time dynamic data, including ship speed, direction, H t is a collection of historical navigation data, including past navigation paths. Ship behavior, E t is a collection of environmental data, including wind speed, tide, and current. t Real-time sound wave data of the intelligent sonar system, including the distance, direction and speed of the target;

[0101] G(S t ,E t ) i It is an intelligent sonar analysis function that outputs the current trajectory and predicted trajectory of the ship. t , combined with U(y t ,z t ) Evaluate uncertainty σ t ;

[0102] is an exponential decay factor, where σ t is the uncertainty assessment value, which indicates the uncertainty of the prediction result. This factor is used to adjust the influence of the prediction result, based on the probability density function of the Gaussian distribution;

[0103] is the normalization factor in the Gaussian distribution, ensuring that the integral value of the probability density function is 1, and is used to adjust Make it conform to the form of standard normal distribution;

[0104] W(λ t ) is the early warning decision weight function, which is used to calculate the decision influence factor λ t The weight given to the warning result, λ t is through situational awareness and reasoning function I(y t ,z t ,σ t ), based on the current situation, the possible actions that the ship may take in the future are inferred, and the decision influencing factor λ is output in combination with the uncertainty assessment value. t ;

[0105] λ t is the decision-making influencing factor, comprehensively considering the future behavior prediction value y t , trajectory prediction value z t and uncertainty estimate σ t The value generated after the warning, which reflects the impact of the current situation on the warning;

[0106] Value range meaning:

[0107] The output of the formula R t Represents the warning intensity in the time period t0 to t. The larger the value, the higher the urgency and necessity of the warning. t If the preset threshold value θ is exceeded, the system will trigger an early warning notification.

[0108] Through this formula, the system can comprehensively consider various data and uncertainties, and automatically decide whether to issue a warning message, thereby achieving smarter and more accurate construction area safety management.

[0109] Specifically, the specific circumstances under which the early warning decision module issues early warning information are as follows:

[0110] When the uncertainty assessment value σ t Less than a critical value σ c When the warning decision value R t If it is greater than the threshold value θ, the system will immediately issue an early warning;

[0111] When the uncertainty assessment value σ t Equal to or greater than a critical value σ c When R t The growth rate of the economy slowed down and early warnings were delayed or not issued;

[0112] Explanation: Uncertainty estimate σ t represents the confidence of the prediction. t When σ is low, it means that the prediction results are more reliable and the system can make early warning decisions faster and more confidently. t Less than the critical value σ c When the warning decision value R t It will accumulate quickly and more easily exceed the predetermined threshold θ, thus triggering an early warning. This design ensures that the system can issue an early warning in a timely manner under high confidence conditions, reducing the possibility of false alarms.

[0113] Beneficial effects: By reducing the uncertainty assessment value, the system can make early warning decisions quickly when the prediction is accurate, improving the timeliness and reliability of the early warning.

[0114] When the decision influencing factor λ t Greater than a threshold λ c When the warning weight W(λ t ) will increase rapidly, making the warning decision value R t Significantly higher than the threshold θ;

[0115] When the decision influencing factor λ t Less than or equal to a threshold λ c When the warning weight W(λt ) will decrease, making the warning decision value R t The growth rate slows down or stops;

[0116] Explanation: Decision influence factor λ t It reflects the importance of current environmental conditions to early warning decisions. t Greater than the preset threshold λ c When the warning weight W(λ t ) will increase significantly, which means that in high-risk or special situations, the system pays more attention to the impact of the current environment, thereby increasing the warning decision value R t This interactive mechanism ensures that in high-risk situations, the system can make early warning decisions quickly to avoid delays.

[0117] Beneficial effects: By dynamically adjusting the warning weights, the system can respond flexibly to environmental changes and improve the adaptability and accuracy of warnings.

[0118] When the ship behavior prediction value y t Equal to the trajectory prediction value z t When the warning decision value R t The growth rate reaches the highest level, and the warning signal is immediately issued;

[0119] When the ship behavior prediction value y t and trajectory prediction value z t When they are not equal, the system will t Re-evaluation, uncertain σ t increase, thereby delaying the issuance of warnings.

[0120] Explanation: Ship behavior prediction value y t and trajectory prediction value z t Ideally, they should match to ensure the accuracy of the prediction. When the two are equal, it means that the system's prediction is very accurate, so the warning decision value R t will accumulate quickly, and the warning signal will be issued immediately. However, if the two are not equal, it means that the system prediction may be biased, and the system will automatically increase the uncertainty assessment value σ t , thereby slowing down the warning decision value R t growth and avoid erroneous early warning decisions.

[0121] Beneficial effects: By comparing the predicted values ​​with the trajectory values, the system can self-correct, reduce false alarms, and improve the accuracy and reliability of the system.

[0122] Specifically, the early warning module includes a multi-level decision-making unit, which calculates the decision-making influence factor λ t The changing trend of the warning decision value R is adjusted in real time tThe growth rate enables the early warning module to issue early warning signals more quickly in high-risk environments based on the results of ship behavior prediction and sonar analysis.

[0123] The early warning module also includes a collaborative evaluation unit, which compares the ship behavior prediction value y t and trajectory prediction value z t The consistency of y t and z t Equal and uncertainty σ t When the two are inconsistent, the warning module will delay the warning until a re-evaluation results in a higher confidence level.

[0124] In summary, the smart construction site safety management platform of the present invention realizes high-precision prediction and real-time monitoring of ship behavior by integrating technologies such as dynamic feature selection, machine learning prediction, multi-model integration and intelligent sonar analysis. In particular, the multi-level decision-making unit and collaborative evaluation unit of the early warning module enable the platform to issue early warning signals quickly and accurately in high-risk environments. In addition, by comprehensively considering the predicted value and sonar analysis results, the system can adjust the early warning decision value in real time, respond quickly to potential dangerous situations, and effectively reduce safety risks on the construction site.

[0125] The present invention can improve the adaptability of the platform in complex and changeable environments by introducing a feature screening mechanism based on fuzzy logic and a multi-frequency signal processing function of intelligent sonar. The dynamic feature selection unit can automatically adjust the feature selection strategy according to real-time environmental changes, thereby improving the accuracy of the prediction. The intelligent sonar analysis module ensures high reliability and accuracy in various environments by sensing underwater environmental conditions in real time and compensating and correcting sonar signals. The application of these technologies enables the system to intelligently adapt to environmental changes and continuously optimize itself based on new data, thereby providing more accurate early warning services and enhancing the intelligence level and application scope of the platform.

[0126] The management method using the above-mentioned smart construction site safety management platform includes the following steps:

[0127] Step 1: Data Collection

[0128] 1.1: Start the data acquisition module and configure the data acquisition equipment to receive real-time dynamic data, historical navigation data, environmental data from the ship, and real-time sound wave data from the intelligent sonar system.

[0129] 1.2: Obtain data from various data sources (such as ship sensors, environmental monitors, sonar systems) regularly or on demand, and ensure the integrity and accuracy of the data.

[0130] 1.3: Store the collected raw data in the data storage system for subsequent processing and analysis.

[0131] Step 2: Data Processing

[0132] 2.1: Start the data processing module to pre-process the collected multi-source data, including data cleaning, denoising, normalization and format conversion.

[0133] 2.2: Integrate information from different data sources and transform it into a format suitable for machine learning and smart sonar analysis modules.

[0134] Step 3: Machine Learning Prediction

[0135] 3.1: Start the machine learning prediction module and perform dynamic feature selection to filter out the most critical features for ship behavior prediction.

[0136] 3.2: Use a multi-model ensemble unit to select or weight the results of different prediction models based on the characteristics of the input data.

[0137] 3.3: Perform situational awareness reasoning, combining current environmental data and historical behavior to predict the future behavior of the ship.

[0138] 3.4: Start the self-learning and model updating unit to continuously update and optimize the prediction model based on system feedback.

[0139] 3.5: Use the uncertainty assessment unit to assess the uncertainty of the prediction results and determine the confidence interval of the prediction results.

[0140] Step 4: Smart Sonar Analysis

[0141] 4.1: Start the intelligent sonar analysis module to process multi-frequency sonar signals and separate and analyze the sound wave reflection signals in different frequency bands.

[0142] 4.2: Use the environmental perception and compensation unit to compensate and correct the sonar signal to ensure data accuracy.

[0143] 4.3: Identify and classify underwater targets, including ships, marine life and fixed structures, through the target recognition and classification unit.

[0144] 4.4: Dynamic signal enhancement and noise reduction unit enhances useful signals and suppresses noise to improve signal quality.

[0145] 4.5: Use the real-time position tracking and trajectory prediction unit to track the current position of the vessel and predict its future trajectory.

[0146] 4.6: Start the self-learning and system optimization unit to continuously update the sonar analysis algorithm and improve the system adaptability and accuracy.

[0147] Step 5: Early warning decision making

[0148] 5.1: Start the early warning module, calculate the early warning decision value, and apply the early warning decision algorithm formula based on the results of the machine learning prediction module and the intelligent sonar analysis module.

[0149] 5.2: Based on the calculated warning decision value, evaluate whether it exceeds the preset threshold.

[0150] 5.3: If the warning decision value exceeds the threshold, the warning signal is triggered, including visual warning, sound alarm and notification to relevant personnel.

[0151] 5.4: Adjust the warning decision according to the specific situation of the warning information, and consider the changing trends of uncertainty assessment values ​​and decision-making influencing factors.

[0152] Step 6: Display and Notification

[0153] 6.1: Start the display and notification module to present the warning information in a visual form on the monitoring platform of the construction site management personnel.

[0154] 6.2: Send alarm notifications to relevant personnel to ensure that they receive warning information in a timely manner and take appropriate measures.

[0155] 6.3: Provide a visual dashboard to display the current ship position, predicted trajectory, warning status and historical warning records.

[0156] Step 7: Feedback and Learning

[0157] 7.1: Start the feedback and learning module to collect feedback information after the actual warning, including the accuracy of the warning and the system response.

[0158] 7.2: Analyze the feedback data and use it to optimize and update the machine learning prediction module.

[0159] 7.3: Adjust and optimize the prediction model based on feedback information to improve the accuracy of future predictions and the overall performance of the system.

[0160] The above steps can ensure the efficient operation of the smart construction site safety management platform and provide timely and accurate early warning information based on ship behavior prediction and intelligent sonar analysis to ensure the safety of the construction site.

[0161] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A smart construction site safety management platform, characterized in that: include: A data acquisition module, which is responsible for collecting real-time dynamic data of the ship, historical navigation data, environmental data, and real-time sound wave data from the intelligent sonar system; A data processing module, wherein the data processing module is used to pre-process the collected multi-source data; A machine learning prediction module, wherein the machine learning prediction module is used to predict the future behavior of the ship; An intelligent sonar analysis module, which is used to analyze underwater sound wave reflections and monitor the precise position of the ship; An early warning module, which is used to integrate the results from the machine learning prediction module and the intelligent sonar analysis module, and decide whether to issue an early warning message according to predetermined rules and algorithms; A display and notification module, which is used to present the warning information in a visual form on the monitoring platform of the construction site management personnel and send alarm notifications to relevant personnel; Feedback and learning module, which is used to collect actual feedback after early warning information, and use these feedback data to optimize and update the machine learning module to improve future prediction accuracy.

2. The smart construction site safety management platform according to claim 1, characterized in that: The machine learning prediction module includes a dynamic feature selection unit, a multi-model integration unit, a context-aware reasoning unit, a self-learning and model updating unit, and an uncertainty assessment unit; The dynamic feature selection unit can automatically select the most critical features for ship behavior prediction from a large number of input features, and can dynamically adjust the feature selection strategy according to real-time environmental changes; The multi-model integration unit integrates a plurality of different prediction models and automatically selects or weightedly combines the prediction results of different models according to the characteristics of the input data to improve the accuracy and robustness of the prediction; The context-aware reasoning unit is used to combine current environmental data and ship behavior history to analyze the current navigation scenario in real time and perform context-aware reasoning on possible actions that the ship may take; The self-learning and model updating unit continuously updates and optimizes the prediction model by continuously monitoring and analyzing system feedback data; The uncertainty assessment unit is used to perform uncertainty assessment on each prediction result and provide a confidence interval for the prediction result.

3. The smart construction site safety management platform according to claim 1, characterized in that: The intelligent sonar analysis module includes a multi-frequency sonar signal processing unit, an environment perception and compensation unit, a target recognition and classification unit, a dynamic signal enhancement and noise reduction unit, a real-time position tracking and trajectory prediction unit, and a self-learning and system optimization unit; The multi-frequency sonar signal processing unit is used to receive and process signals from the multi-frequency sonar, distinguish the acoustic characteristics of different objects by separating the sound wave reflection signals of different frequency bands, and especially to accurately identify the type, size and position of the ship forever; The environmental sensing and compensation unit is used to sense the underwater environmental conditions in real time and perform corresponding compensation and correction on the sonar signal; The target recognition and classification unit recognizes and classifies the detected targets based on the processed sonar signals, such as distinguishing between ships, marine life, fixed structures, etc. The dynamic signal enhancement and noise reduction unit focuses on enhancing useful sonar signals and suppressing noise; The real-time position tracking and trajectory prediction unit is used to track the current position of the ship in real time and predict the future trajectory of the ship based on current and historical data; The self-learning and system optimization unit continuously updates and optimizes the sonar analysis algorithm by analyzing the results of each sonar signal processing and target recognition, so that the system can adapt to the ever-changing environment and target characteristics.

4. The smart construction site safety management platform according to claim 1, characterized in that: The warning module uses a warning decision algorithm to decide whether to issue a warning message. The formula of the warning decision algorithm is as follows: in: R t is the warning decision value, representing the comprehensive warning intensity in the time period t0 to t; P(X) i It is the behavior prediction function after multi-model integration, and outputs the predicted value y of the ship's future behavior t , represents the predicted value of the ith model for the feature vector X, where X is the feature selection function F(D t ,H t ,E t ,S t ) provides the feature vector, where D t is a collection of real-time dynamic data, including ship speed, direction, H t is a collection of historical navigation data, including past navigation paths. Ship behavior, E t is a collection of environmental data, including wind speed, tide, and current. t Real-time sound wave data of the intelligent sonar system, including the distance, direction and speed of the target; G(S t ,E t ) i It is an intelligent sonar analysis function that outputs the current trajectory and predicted trajectory of the ship. t , combined with U(y t ,z t ) Evaluate uncertainty σ t ; is an exponential decay factor, where σ t is the uncertainty assessment value, which indicates the uncertainty of the prediction result. This factor is used to adjust the influence of the prediction result, based on the probability density function of the Gaussian distribution; is the normalization factor in the Gaussian distribution, ensuring that the integral value of the probability density function is 1, and is used to adjust Make it conform to the form of standard normal distribution; W(λ t ) is the early warning decision weight function, which is used to calculate the decision influence factor λ t The weight given to the warning result, λ t is through situational awareness and reasoning function I(y t ,z t ,σ t ), based on the current situation, the possible actions that the ship may take in the future are inferred, and the decision influencing factor λ is output in combination with the uncertainty assessment value. t ; λ t is the decision-making influencing factor, comprehensively considering the future behavior prediction value y t , trajectory prediction value z t and uncertainty estimate σ t The value generated after the warning, which reflects the impact of the current situation on the warning; Value range meaning: The output of the formula R t Represents the warning intensity in the time period t0 to t. The larger the value, the higher the urgency and necessity of the warning. t If the preset threshold value θ is exceeded, the system will trigger an early warning notification.

5. The smart construction site safety management platform according to claim 4, characterized in that: The specific circumstances under which the early warning decision module issues early warning information are as follows: When the uncertainty assessment value σ t Less than a critical value σ c When the warning decision value R t If it is greater than the threshold value θ, the system will immediately issue an early warning; When the uncertainty assessment value σ t Equal to or greater than a critical value σ c When R t The growth rate of the economy slowed down and early warnings were delayed or not issued; When the decision influencing factor λ t Greater than a threshold λ c When the warning weight W(λ t ) will increase rapidly, making the warning decision value R t Significantly higher than the threshold θ; When the decision influencing factor λ t Less than or equal to a threshold λ c When the warning weight W(λ t ) will decrease, making the warning decision value R t The growth rate slows down or stops; When the ship behavior prediction value y t Equal to the trajectory prediction value z t When the warning decision value R t The growth rate reaches the highest level, and the warning signal is immediately issued; When the ship behavior prediction value y t and trajectory prediction value z t When they are not equal, the system will t Re-evaluation, uncertain σ t increase, thereby delaying the issuance of warnings.

6. The smart construction site safety management platform according to claim 2, characterized in that: The dynamic feature selection unit includes a feature screening mechanism based on fuzzy logic, which is used to dynamically adjust the priority of feature selection through a fuzzy membership function in real-time environmental changes.

7. The smart construction site safety management platform according to claim 1, characterized in that: The warning information includes visual warnings, audible alarms, and notifications to relevant personnel.

8. The smart construction site safety management platform according to claim 1, characterized in that: The early warning module includes a multi-level decision-making unit, which calculates the decision-making influence factor λ t The changing trend of the warning decision value R is adjusted in real time t The growth rate enables the early warning module to issue early warning signals more quickly in high-risk environments based on the results of ship behavior prediction and sonar analysis.

9. The smart construction site safety management platform according to claim 8, characterized in that: The early warning module also includes a collaborative evaluation unit, which compares the ship behavior prediction value y t and trajectory prediction value z t The consistency of y t and z t Equal and uncertainty σ t When the two are inconsistent, the warning module will delay the warning until a re-evaluation results in a higher confidence level.

10. The management method of the smart construction site safety management platform according to any one of claims 1 to 9, characterized in that: The following steps are involved: Step 1: Start the data acquisition module and configure the data acquisition equipment to receive real-time dynamic data, historical navigation data, environmental data and real-time acoustic wave data from the ship's intelligent sonar system; obtain data from various data sources regularly or on demand, and ensure the integrity and accuracy of the data; store the collected raw data in the data storage system for subsequent processing and analysis; Step 2: Start the data processing module to pre-process the collected multi-source data, including data cleaning, denoising, normalization and format conversion; integrate information from different data sources and convert it into a format suitable for machine learning and intelligent sonar analysis modules; Step 3: Start the machine learning prediction module and perform dynamic feature selection to filter out the most critical features for ship behavior prediction; use the multi-model integration unit to select or weightedly combine the results of different prediction models according to the characteristics of the input data; perform situational awareness reasoning, combine current environmental data and historical behavior, and predict the future behavior of the ship; Start the self-learning and model updating unit to continuously update and optimize the prediction model based on system feedback; use the uncertainty assessment unit to assess the uncertainty of the prediction results and determine the confidence interval of the prediction results; Step 4: Start the intelligent sonar analysis module to process multi-frequency sonar signals, separate and analyze sound wave reflection signals in different frequency bands; use the environmental perception and compensation unit to compensate and correct the sonar signals to ensure data accuracy; The target recognition and classification unit identifies and classifies underwater targets, including ships, marine life and fixed structures. The dynamic signal enhancement and noise reduction unit enhances useful signals and suppresses noise to improve signal quality. The real-time position tracking and trajectory prediction unit tracks the current position of the ship and predicts its future trajectory. Start the self-learning and system optimization unit to continuously update the sonar analysis algorithm and improve the system adaptability and accuracy; Step 5: Start the warning module, calculate the warning decision value, and apply the warning decision algorithm formula based on the results of the machine learning prediction module and the intelligent sonar analysis module; Based on the calculated warning decision value, assess whether it exceeds the preset threshold; If the warning decision value exceeds the threshold, the warning signal is triggered, including visual warning, sound alarm and notification to relevant personnel; according to the specific situation of the warning information, the warning decision is adjusted, taking into account the changing trend of the uncertainty assessment value and decision-making influencing factors; Step 6: Start the display and notification module to present the warning information in a visual form on the monitoring platform of the construction site manager; Send alarm notifications to relevant personnel to ensure that they receive warning information in time and take corresponding measures; provide a visual dashboard to display the current ship position, predicted trajectory, warning status and historical warning records; Step 7: Start the feedback and learning module to collect feedback information after the actual warning, including the accuracy of the warning and the system response; analyze the feedback data and use this data to optimize and update the machine learning prediction module; adjust and optimize the prediction model based on the feedback information to improve the accuracy of future predictions and the overall performance of the system.

Images