A multi-dimensional positioning and safety detection method and system based on safety helmet

Abstract

The present invention discloses a multi-dimensional positioning and safety detection method and system based on a hard hat, which relates to the field of positioning and adjustment technology, including real-time collection of workers' location information through a positioning module on a hard hat, and uploading the collected data to a control center; collecting sensor data from various areas, using the real-time data of each sensor to predict the changing trend of regional danger; based on the workers' historical movement data, current location and work tasks, predicting the workers' movement paths in future time periods; combining the workers' predicted paths with the dynamic danger assessment of each area, and performing multi-level early warning. The present invention comprehensively utilizes the positioning module of the hard hat and the multi-dimensional sensor data, which can realize accurate monitoring and dynamic safety assessment of workers' activities, and has significant beneficial effects in improving workers' safety, enhancing on-site safety management, optimizing emergency response, etc.

Description

Technical Field

[0001] The invention relates to the technical field of positioning and adjustment, in particular to a multi-dimensional positioning and safety detection method and system based on a safety helmet. Background Art

[0002] With the acceleration of industrialization, worker safety protection technology has been widely developed. Traditional safety protection measures usually rely on personal protective equipment (PPE) and on-site supervision, but these methods are not effective enough in a rapidly changing work environment. In recent years, based on the development of the Internet of Things (IoT) and big data analysis technology, many companies have begun to explore the integration of smart sensors and positioning technology into personal protective equipment such as hard hats to enhance workers' safety protection capabilities. The application of these technologies makes it possible to monitor workers' locations, environmental data, and potential safety hazards in real time, thus providing new ideas for safety management. However, existing technologies still have shortcomings in data integration, real-time response, and accurate prediction. For example, many existing systems are only able to collect location information or environmental parameters, but lack the ability to deeply analyze and dynamically predict data, resulting in the inability to identify and respond to potential safety threats in a timely manner.

[0003] Although some systems can monitor workers’ locations in real time, they are still insufficient in dynamically assessing and predicting environmental risks and cannot provide early warning information in a timely manner. In addition, existing technologies are usually unable to combine workers’ historical movement data with real-time location information, resulting in a lack of effective response plans when sudden dangers occur. These deficiencies not only reduce the accuracy of safety detection, but may also delay the response time to potential dangers, increasing the safety risks faced by workers. Summary of the invention

[0004] The present invention is proposed in view of the above-mentioned background problems.

[0005] Therefore, the problem to be solved by the present invention is how to combine dynamic hazard prediction with workers' historical movement path analysis to achieve more accurate and comprehensive safety monitoring.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, an embodiment of the present invention provides a multi-dimensional positioning and safety detection method based on a safety helmet, which includes collecting the worker's location information in real time through a positioning module on the safety helmet, and uploading the collected data to a control center; collecting sensor data from each area, and using the real-time data of each sensor to predict the changing trend of the regional danger level; based on the worker's historical movement data, current location and work tasks, predicting the worker's movement path in a future time period; combining the worker's predicted path with the dynamic danger level assessment of each area, and performing multi-level warnings.

[0008] As a preferred solution of the multi-dimensional positioning and safety detection method based on a safety helmet described in the present invention, the predicted changing trend of regional danger level includes the following steps: extracting time series features from processed sensor data, analyzing the time changing laws of various types of danger sources, extracting characteristic values ​​using a sliding window method, and forming a real-time feature matrix for each region; statistically analyzing past regional danger changes in combination with historical data of the sensor; constructing a danger level change trend prediction model; dynamically adjusting the real-time danger level assessment within the region based on the results of the danger level change trend prediction; and automatically generating early warning prompts based on the results of the danger level change trend prediction and the regional danger level change trend.

[0009] As a preferred solution of the multi-dimensional positioning and safety detection method based on a helmet of the present invention, the construction of a risk change trend prediction model includes: based on historical data and real-time feature matrix , use the LSTM model to build a risk prediction model; the prediction formula of the LSTM model is:

[0010] ,

[0011] in, is the activation function, usually a sigmoid function or a tanh function, is the hidden state of the LSTM unit, and is the model weight, To predict the trend of risk, and are model parameters, is the error term, is the time step, is the bias vector of the LSTM unit; is the bias vector of the LSTM output layer.

[0012] As a preferred solution of the multi-dimensional positioning and safety detection method based on a helmet of the present invention, the real-time risk assessment in the area is dynamically adjusted according to the result of the risk change trend prediction, including: according to the predicted future risk change and the current real-time danger level , and conduct dynamic adjustment evaluation:

[0013] ,

[0014] in, is the adjusted risk assessment result. is the adjustment weight coefficient.

[0015] As a preferred solution of the multi-dimensional positioning and safety detection method based on a safety helmet described in the present invention, the prediction of the movement path of the worker in the future time period includes the following steps: the control center collects and stores the historical movement data of the worker, and identifies the worker's regular movement pattern and abnormal behavior through the historical movement data; determines the worker's current specific location and area through positioning, and associates it with the historical data to form the worker's current status; automatically retrieves the worker's current work tasks and operating instructions, and combines the task requirements with the worker's current status according to the task priority, urgency and the work area that the worker should reach, and evaluates the possible path for task execution.

[0016] As a preferred solution of the multi-dimensional positioning and safety detection method based on a helmet of the present invention, wherein: the task requirements are combined with the current status of the worker to evaluate the possible paths for task execution, including the following steps: the task requirements of the worker include , which are the coordinates of the task target location, the priority of the task, and the urgency of the task; based on the current status of the worker and mission requirements Evaluate the set of possible paths for workers to perform tasks, and for each path Define the path cost function:

[0017] ,

[0018] in, For path Length, is the distance between the worker’s current state and the task goal, , , and is the weight parameter of the cost function, which is used to balance the influence of path length, task priority, urgency and the distance from the current position to the target; through the path planning algorithm, a set of possible paths is generated. , and according to the path cost function The cost of each path is calculated and the formula for selecting the optimal path for the task is as follows:

[0019] .

[0020] As a preferred solution of the multi-dimensional positioning and safety detection method based on a safety helmet of the present invention, wherein: the current state of the worker include: ;in, For the worker's historical movement path mode, is the current moving speed, is the worker’s current specific location coordinates, Identification of workers; task requirements include: ;in, is the coordinate of the mission target location, is the priority of the task, The urgency of the task.

[0021] In the second aspect, an embodiment of the present invention provides a multi-dimensional positioning and safety detection system based on a safety helmet, which includes a positioning module, which collects the worker's location information in real time through a GPS or wireless signal positioning device installed on the safety helmet, and sends the location information to a control center; a data acquisition module, which is used to collect sensor data from each area and upload the collected data to the control center; a hazard assessment module, which is used to calculate the hazard level of each area according to the dynamic changes of the sensor data, combined with historical data and algorithm models, and predict future changes in hazard levels; a movement path prediction module, which is used to predict the worker's future movement path based on the worker's historical movement trajectory data, current task priority, urgency and the worker's current location.

[0022] In a third aspect, an embodiment of the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the multi-dimensional positioning and safety detection method based on the safety helmet as described in the first aspect of the present invention are implemented.

[0023] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of the multi-dimensional positioning and safety detection method based on a safety helmet as described in the first aspect of the present invention are implemented.

[0024] The beneficial effects of the present invention are as follows: the present invention comprehensively utilizes the positioning module of the safety helmet and multi-dimensional sensor data to achieve accurate monitoring and dynamic safety assessment of workers' activities, and has significant beneficial effects in improving worker safety, enhancing on-site safety management, and optimizing emergency response. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of 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 work.

[0026] Figure 1 The flowchart of the multi-dimensional positioning and safety detection method based on the helmet.

[0027] Figure 2 This is the structural diagram of the multi-dimensional positioning and safety detection system based on the helmet. DETAILED DESCRIPTION

[0028] 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.

[0029] 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.

[0030] 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.

[0031] Example 1

[0032] Reference Figure 1 , which is the first embodiment of the present invention, and provides a multi-dimensional positioning and safety detection method based on a helmet, such as Figure 1 As shown, including

[0033] S1: The location information of workers is collected in real time through the positioning module (such as GPS, Bluetooth, UWB, etc.) on the helmet, and the sensor data of each area is collected, and the collected data is uploaded to the control center.

[0034] It should be noted that the workers' location data will be continuously uploaded to the control center and form the workers' movement trajectory in the work area.

[0035] Install various positioning technology modules on the helmet, such as GPS and Bluetooth, to ensure that the real-time location of workers can be accurately obtained in different environments. The specific operations are as follows:

[0036] Among them, GPS positioning is suitable for outdoor environments, especially open working areas, receiving satellite signals in real time and providing high-precision location information.

[0037] Bluetooth positioning is suitable for indoor or confined space positioning. Through Bluetooth beacons deployed in each work area, the Bluetooth module on the helmet is positioned through signal strength (RSSI) or arrival time difference (TDOA).

[0038] At specified time intervals (e.g. 1 second or 0.5 seconds), the positioning system automatically obtains the worker's location information, including longitude and latitude and accuracy indicators, and packages the collected location information together with the timestamp to form a standardized data packet.

[0039] At the same time, environmental sensors installed in the workers' work areas will continuously collect safety parameters of the area. These sensors include but are not limited to: temperature sensors, toxic gas sensors, vibration sensors, smoke and flame sensors; each sensor collects environmental data regularly or according to event triggers, and uploads the data to the control center through a communication network (such as 5G, Wi-Fi or LoRa).

[0040] The sensor data and the workers’ location information are uploaded simultaneously. The control center integrates the data to form a real-time status snapshot of the area, and cleans and preprocesses the real-time data uploaded by each sensor.

[0041] S2: Utilize the real-time data from each sensor through time series prediction algorithms or machine learning models to predict the changing trend of regional danger.

[0042] S2.1: Extract time series features from the processed sensor data, analyze the temporal variation patterns of various types of hazardous sources, use the sliding window method to extract feature values, and form a real-time feature matrix for each area.

[0043] Preferably, assuming that the sensor data is , where t is time, using sliding window length W to extract feature values ​​in the time series, for sensor data ,..., , the feature matrix can be expressed as:

[0044] ,

[0045] Where n is the number of sensors, is the characteristic matrix at the corresponding moment, μ is the mean, is the variance, is the maximum value, is the minimum value.

[0046] S2.2: Combine historical data from sensors to conduct statistical analysis of past changes in regional hazards.

[0047] By comparing historical hazard change patterns with real-time data, environmental features associated with historical incidents can be identified, e.g., setting historical time periods The regional hazard change data within , , by comparing with the real-time data at the current time t, identify the relevant environmental features .

[0048] S2.3: Construct a model to predict the trend of risk changes.

[0049] Preferably, based on historical data and real-time feature matrix , use machine learning models such as LSTM (Long Short-Term Memory Network) or GRU (Gated Recurrent Unit) to build a risk prediction model, where the prediction formula of the LSTM model is:

[0050] ,

[0051] in, is the activation function, usually a sigmoid function or a tanh function, is the hidden state of the LSTM unit, and is the model weight, To predict the trend of risk, and are model parameters, is the error term, is the time step, is the bias vector of the LSTM unit; is the bias vector of the LSTM output layer.

[0052] S2.4: Dynamically adjust the real-time hazard assessment in the area based on the prediction results of hazard changes.

[0053] Based on the predicted future risk changes and the current real-time danger level , and conduct dynamic adjustment evaluation:

[0054] ,

[0055] in, is the adjusted risk assessment result. It is an adjustment weight coefficient that combines the current risk and the predicted value.

[0056] If the hazard prediction model predicts that the hazard level will have a significant upward trend in the future period, it will be corrected based on the real-time hazard assessment to generate an expected hazard level, which will be combined with the real-time hazard level to form a more accurate hazard status prediction.

[0057] S2.5: Automatically generate early warning alerts based on prediction results and regional hazard change trends.

[0058] If the future danger level of an area exceeds the safety threshold, the system will send out a warning signal and alert the worker through the sound or vibration of the hard hat.

[0059] S3: Based on the workers’ historical movement data, current location, and work tasks, predict the workers’ possible movement paths in the future time period.

[0060] S3.1: The control center regularly collects and stores the historical movement data of workers. Through the above historical data, the regular movement patterns and abnormal behaviors of workers are identified. The movement data of workers is timestamped. , location coordinates , Movement Speed and worker status Construct, organize and store this data into a set of historical movement sequences .

[0061] Based on the workers’ daily movement data, clustering algorithms (such as K-means) are used to analyze regular movement patterns.

[0062] Furthermore, the detection of abnormal behavior can be determined by the degree of deviation from the normal pattern, and if the deviation exceeds a threshold, it is considered abnormal behavior.

[0063] S3.2: Determine the worker’s current specific location and area through positioning, and associate it with historical data to form the worker’s current status.

[0064] Preferably, the worker's current specific location is obtained through the positioning system , and with the workers' historical data Associate to form the current status of the worker , the state includes the worker’s current position and historical path pattern, expressed as: ,in, For the worker's historical movement path mode, is the current moving speed, Identification of workers.

[0065] It should be noted that the current status of the worker not only includes the current location information, but also needs to extract the regular path pattern based on the historical movement data and compare it with the current location information in real time. The current path can be updated through the sliding window method and continuously matched with the historical path pattern.

[0066] S3.3: Automatically retrieve the worker’s current work tasks and operating instructions (such as maintenance, inspection, equipment operation, etc.), combine the task requirements with the worker’s current status based on the task priority, urgency and the work area the worker should reach, and evaluate the possible paths for task execution.

[0067] Optimally, worker task requirements include , are the coordinates of the task target location, the task priority and the task urgency respectively.

[0068] Based on the current state of the worker and mission requirements Evaluate the possible paths for workers to perform tasks, where the path set can be generated by a path planning algorithm (such as Dijkstra or A* algorithm). Define the path cost function:

[0069] ,

[0070] in, For path Length, is the distance between the worker’s current state and the task goal, , , and is the weight parameter of the cost function, which is used to balance the impact of path length, task priority, urgency, and the distance from the current location to the target.

[0071] Generate a set of possible paths through the path planning algorithm , and according to the path cost function Calculate the cost of each path, and finally select the optimal path for the task according to the following formula:

[0072] ,

[0073] Through this optimal path, workers can be guided to perform tasks in the shortest time and via the optimal route, taking into account the combined impact of priorities and emergencies.

[0074] It should be noted that if the task performed by the worker changes, or the worker needs to change the path due to environmental conditions (such as increased danger detected by sensors), the movement path prediction will be updated based on the latest information and the optimal path will be replanned. Or if the sensor detects that the area passed by a certain path is more dangerous, the path will be updated in real time and a low-risk path will be selected.

[0075] S4: Combine the predicted paths of workers with the dynamic hazard assessment of each area to provide multi-level warning.

[0076] Preferably, the dynamically assessed risk level of each area is continuously received.

[0077] Each key point in the predicted path is matched with the danger level of the area to form an association table of the path and environmental danger. The association table records the danger level corresponding to the path node and is updated according to the real-time sensor data.

[0078] After combining the worker's predicted path with the regional hazard assessment results, it is automatically determined whether each node in the worker's path is in a high-risk area. Specifically, the worker's predicted path Given by the path prediction model in S3.4, it is defined as a node sequence in the future period; assuming that the predicted risk level of each area is , each node in the path is associated with the area where it is located. The area where the node is located is found through the node coordinates, and the danger level of the area is matched to the path node: the path node and the danger level are matched through the association table of the path and environmental danger to form an association table on the time series.

[0079] Set warning thresholds, such as excessive concentration of dangerous gases, abnormal equipment vibration, or excessive temperature in high-temperature areas. Once the danger level of a node exceeds the preset threshold, the system will mark the node as a high-risk area.

[0080] When the worker's predicted path passes through a high-risk area, the warning program will be activated through the early warning mechanism. The sound reminder device or vibration reminder device built into the helmet will immediately sound an alarm, reminding the worker that he is about to enter the high-risk area and reminding him to stop the current action. At the same time, the worker's movement trajectory will be uploaded to the control center in real time through the positioning system for real-time monitoring by safety managers.

[0081] Furthermore, a multi-level warning mechanism is set up according to the different danger levels of high-risk areas. For example, when the danger level is low, only a sound prompt is issued; when the danger level is high, a vibration warning is activated; when the danger level reaches an extremely high level or a worker is about to enter an extremely dangerous area, the highest level warning will be automatically triggered and the worker will be forced to evacuate. In addition, the safety management personnel will be linked to enable them to respond quickly.

[0082] In some special cases, if the danger level in a certain area increases sharply and exceeds the set threshold, the highest level warning will be issued and the area blockade command will be automatically executed. This command can be linked to the access control system, electronic fence or other safety isolation equipment to prevent workers or other personnel from entering the dangerous area, further ensuring the safety of personnel.

[0083] This embodiment also provides a computer device, which is suitable for the multi-dimensional positioning and safety detection method based on a safety helmet, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the multi-dimensional positioning and safety detection method based on a safety helmet as proposed in the above embodiment.

[0084] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0085] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the multi-dimensional positioning and safety detection method based on a safety helmet as proposed in the above embodiment is implemented.

[0086] In summary, the present invention comprehensively utilizes the positioning module of the safety helmet and multi-dimensional sensor data to achieve accurate monitoring of workers' activities and dynamic safety assessment, which has significant beneficial effects in improving worker safety, enhancing on-site safety management, and optimizing emergency response.

[0087] Example 2

[0088] Reference Figure 2 This embodiment provides a multi-dimensional positioning and safety detection system based on a helmet, including:

[0089] The positioning module collects the worker's location information in real time through the GPS or wireless signal positioning device installed on the helmet, and sends the location information to the control center;

[0090] Data acquisition module, used to collect sensor data in each area and upload the collected data to the control center;

[0091] The danger assessment module is used to calculate the danger level of each area based on the dynamic changes of sensor data, combined with historical data and algorithm models, and predict possible changes in danger levels in the future;

[0092] The mobile path prediction module is used to predict the worker's possible future movement path through a path regression model or other algorithms based on the worker's historical movement trajectory data, current task priority, urgency and the worker's current location.

[0093] Example 3

[0094] Referring to Table 1, which is the third embodiment of the present invention, this embodiment provides a multi-dimensional positioning and safety detection method based on a helmet. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0095] The experiment selected six hard hats (Hard Hat 1 to Hard Hat 6) in different construction site environments as test objects. Each hard hat is equipped with multiple positioning modules (such as GPS and Bluetooth) and multiple environmental sensors (temperature sensor, toxic gas sensor, vibration sensor, etc.) to monitor the safety status of workers in real time.

[0096] The specific operations are as follows:

[0097] At each construction site, hard hats are equipped with GPS and Bluetooth modules to provide accurate location information in different environments. At the same time, environmental sensors are deployed in key areas to monitor safety parameters such as temperature, toxic gas concentration and vibration.

[0098] Every second, the hard hat collects the worker's location data in real time through the built-in positioning module and uploads it to the control center together with the timestamp. At the same time, the environmental sensor collects relevant data at specified time intervals. These data are transmitted to the control center via the 5G network to form a real-time status snapshot.

[0099] The control center uses machine learning models (such as LSTM) to analyze the collected sensor data and predict the trend of changes in the danger level in the area. Through time series analysis, the danger sources under different environmental conditions are identified and an early warning mechanism is generated. Based on the workers' historical movement data and current task requirements, clustering algorithms are used to identify regular movement patterns, and combined with real-time location and task requirements, the optimal task path is generated. This process ensures that workers can complete their work in the shortest time and with the lowest risk.

[0100] In the worker's predicted path, each key node is matched with the regional danger level. Once the danger level exceeds the preset threshold, the system immediately issues an alarm to remind the worker to pay attention to safety. Some experimental results are shown in the table below:

[0101] Table 1 Experimental results data

[0102] Experimental subjects Positioning accuracy(m) Temperature(℃) Toxic gas concentration (ppm) Vibration frequency (Hz) Security alert triggering times Task completion rate (%) Helmet 1 1.5 35 5 30 1 95 Helmet 2 2.0 36 7 28 2 90 Helmet 3 1.2 34 4 32 0 98 Helmet 4 1.8 37 6 29 1 92 Helmet 5 2.5 33 8 31 3 85 Helmet 6 1.0 36 3 27 0 100

[0103] From the above experimental data, it can be seen that the multi-dimensional positioning and safety detection method based on the helmet has significant advantages. First, in terms of positioning accuracy, the helmet 3 performed best, with a positioning error of only 1.2 meters, indicating that the device can still maintain a high degree of accuracy in complex environments. This is crucial for the safety of workers on the construction site, because accurate positioning can help managers to grasp the location of workers in a timely manner and rescue them when necessary.

[0104] Secondly, from the environmental monitoring data, helmet 5 has the highest toxic gas concentration (8 ppm), and the number of safety alarms triggered is 3, indicating that it is in a relatively dangerous environment. Helmet 6 has the best performance in terms of toxic gas concentration, which is only 3 ppm, and no safety alarms are triggered, indicating that its working environment is relatively safe. In terms of task completion rate, helmet 6 performed perfectly, reaching 100%, which shows that under good safety conditions, the workers' work efficiency is guaranteed.

[0105] In addition, through the analysis of the number of times the safety alarm was triggered, it can be found that the helmet equipped with multiple sensors and precise positioning functions can effectively warn of potential dangers and prevent accidents. The concentration of toxic gases in helmets 1 and 4 was kept at a relatively low level and only triggered the alarm once, indicating that they can maintain high work efficiency while ensuring safety.

[0106] In summary, this experiment demonstrates the potential value of the multi-dimensional positioning and safety detection method based on hard hats in construction site safety management. Compared with the existing technology, this method not only improves the safety of workers but also performs outstandingly in work efficiency by integrating multiple sensors and positioning modules.

[0107] 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 multi-dimensional positioning and safety detection method based on a helmet, characterized in that: include: The positioning module on the helmet collects the worker's location information in real time and uploads the collected data to the control center; Collect sensor data from each area, and use the real-time data from each sensor to predict the changing trend of regional danger; Predict workers’ movement paths in the future based on their historical movement data, current location, and work tasks; Combine workers’ predicted paths with dynamic hazard assessments of each area and provide multi-level warnings; The prediction of the changing trend of regional danger level comprises the following steps: The processed sensor data is subjected to time series feature extraction, the time variation patterns of various types of hazardous sources are analyzed, and the sliding window method is used to extract feature values ​​to form a real-time feature matrix for each area; Combined with historical data from sensors, statistical analysis of past changes in regional hazards is performed; Construct a model to predict the trend of risk changes; Dynamically adjusting the real-time risk assessment in the area according to the results of the risk change trend prediction; Automatically generate early warning prompts based on the results of the risk change trend prediction and the regional risk change trend; The construction of the risk change trend prediction model comprises: Based on historical data and real-time feature matrix , use LSTM model to build a risk prediction model; The prediction formula of the LSTM model is: ， in, is the activation function, usually a sigmoid function or a tanh function, is the hidden state of the LSTM unit, and is the model weight, To predict the trend of risk, and are model parameters, is the error term, is the time step, is the bias vector of the LSTM unit; is the bias vector of the LSTM output layer; According to the result of the risk change trend prediction, the real-time risk assessment in the area is dynamically adjusted, including: Based on the predicted future risk changes and the current real-time danger level , and conduct dynamic adjustment evaluation: ， in, is the adjusted risk assessment result. is to adjust the weight coefficient; the prediction of the worker's movement path in the future time period includes the following steps: The control center collects and stores the workers' historical movement data, and identifies the workers' regular movement patterns and abnormal behaviors through the historical movement data; The worker's current specific location and area are determined through positioning, and then associated with historical data to form the worker's current status; Automatically retrieve the worker's current work tasks and operation instructions, combine the task requirements with the worker's current status according to the task priority, urgency and the work area the worker should reach, and evaluate the possible paths for task execution; Combining the task requirements with the worker's current status and evaluating possible paths for task execution includes the following steps: Workers' task requirements include , are the coordinates of the task target location, the task priority and the task urgency respectively; Based on the current state of the worker and mission requirements Evaluate the set of possible paths for workers to perform tasks, and for each path Define the path cost function: ， in, For path Length, is the distance between the worker’s current state and the task goal, , , and is the weight parameter of the cost function, which is used to balance the impact of path length, task priority, urgency, and the distance from the current location to the target; Generate a set of possible paths through the path planning algorithm , and according to the path cost function The cost of each path is calculated and the formula for selecting the optimal path for the task is as follows: ， The current status of the worker include: ， in, For the worker's historical movement path mode, is the current moving speed, is the worker’s current specific location coordinates, Identification of workers; Task Requirements include: ， in, is the coordinate of the mission target location, is the priority of the task, The urgency of the task.

2. A multi-dimensional positioning and safety detection system based on a safety helmet, based on the multi-dimensional positioning and safety detection method based on a safety helmet as described in claim 1, characterized in that: Also includes: The positioning module collects the worker's location information in real time through the GPS or wireless signal positioning device installed on the helmet, and sends the location information to the control center; Data acquisition module, used to collect sensor data in each area and upload the collected data to the control center; The danger assessment module is used to calculate the danger level of each area based on the dynamic changes of sensor data, combined with historical data and algorithm models, and predict future changes in danger levels; The movement path prediction module is used to predict the worker's future movement path based on the worker's historical movement trajectory data, current task priority, urgency and the worker's current location.

3. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the multi-dimensional positioning and safety detection method based on a safety helmet described in claim 1 are implemented.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the multi-dimensional positioning and safety detection method based on a safety helmet described in claim 1 are implemented.

Images