An Acoustic-Optic Alarm Method for Heat-Resistant Protective Clothing Based on the Internet of Things

Abstract

The present invention provides an acoustic and optical warning method for heat-resistant protective clothing based on the Internet of Things, which relates to the technical field of protection monitoring. It includes collecting the physiological parameters, surrounding environment parameters, and position parameters of the wearer; analyzing the collected data through a microprocessor or edge device built in the protective clothing to obtain a risk coefficient, where the risk coefficient includes the risk coefficients of each parameter and the comprehensive risk coefficient; determining whether to trigger the first warning condition according to the risk coefficient; determining a transmission coefficient according to the risk coefficients of each parameter and the comprehensive risk coefficient of the wearer; obtaining the risk level and risk probability of each wearer through a machine learning model; determining whether to trigger the second warning condition according to the risk level and risk probability; using wireless transmission to send warning information to the background monitoring center or designated contacts; ensuring that response measures can be taken in a timely manner at different risk levels, avoiding false alarms or missed alarms, and greatly improving safety.

Description

Technical Field

[0001] This application relates to the technical field of protection monitoring, and particularly to an audible and visual alarm method for heat-resistant protective clothing based on the Internet of Things. Background Art

[0002] Workers operating in high-temperature, harmful gas or other extreme environments face serious health and safety risks. To protect the safety of these workers, heat-resistant protective clothing has become an essential personal protective equipment. However, most traditional protective clothing only provides a physical barrier and lacks real-time monitoring and alarm functions, unable to promptly remind workers to stay away from dangerous environments or take necessary protective measures.

[0003] With the rapid development of Internet of Things technology, more and more intelligent devices are being applied in various fields, including personal protective equipment. Internet of Things technology realizes real-time monitoring and data transmission of objects or environments through sensors, microprocessors, wireless transmission, etc.; applying Internet of Things technology to heat-resistant protective clothing can real-time monitor workers' physiological parameters (such as body temperature, heart rate, blood oxygen concentration) and surrounding environmental parameters (such as temperature, humidity, harmful gas concentration), and evaluate the safety status of workers based on these data.

[0004] However, simply collecting data is not sufficient to fully ensure the safety of workers. How to promptly and accurately evaluate the degree of danger of workers based on the collected data and trigger an alarm when necessary is a key issue in the application of Internet of Things technology in heat-resistant protective clothing. In addition, due to the working environment may be very complex, the priority and efficiency of data transmission also need to be fully considered.

[0005] Currently, there are already some Internet of Things-based protective clothing products on the market, but most of these products only provide simple data monitoring and transmission functions and lack intelligent analysis and alarm mechanisms. Summary of the Invention

[0006] Based on the above problems, this application provides an audible and visual alarm method for heat-resistant protective clothing based on the Internet of Things. By collecting the physiological parameters, surrounding environmental parameters, and location information of the wearer, using a microprocessor or edge device for analysis, evaluating the degree of danger, and triggering an audible and visual alarm when necessary. At the same time, the priority and efficiency of data transmission are also considered, and a machine learning model is used to comprehensively analyze the data of multiple wearers to ensure that corresponding measures can be taken in a timely manner at different risk levels, avoiding false alarms or missed alarms, and greatly improving safety.

[0007] The object of this application is achieved by the following technical solutions:

[0008] On the one hand, this application provides an audible and visual alarm method for heat-resistant protective clothing based on the Internet of Things, and the method includes:

[0009] S1. Collect the physiological parameters of the wearer and the surrounding environment parameters through the sensors in the protective clothing; collect the position parameters of the wearer through the positioning device;

[0010] S2. Analyze the collected data through the microprocessor or edge device built in the protective clothing to obtain the risk coefficient, where the risk coefficient includes the risk coefficient of each parameter and the comprehensive risk coefficient; determine whether to trigger the first warning condition according to the risk coefficient, and if so, activate the sound and light alarm device on the protective clothing;

[0011] S3. Determine the transmission coefficient according to the risk coefficient of each parameter and the comprehensive risk coefficient of the wearer; upload the collected data and the first analysis result to the cloud platform according to the transmission coefficient; obtain the risk level and risk probability of each wearer through the machine learning model; determine whether to trigger the second warning condition according to the risk level and risk probability; if so, activate the sound and light alarm device on the protective clothing;

[0012] S4. Use wireless transmission to send the warning information to the background monitoring center or the designated contact person.

[0013] Preferably, the physiological parameters include body temperature, heart rate and blood oxygen concentration; the environmental parameters include temperature, humidity and gas concentration.

[0014] Preferably, the S2 includes:

[0015] Set the range and basic weight of each parameter; obtain the risk coefficient according to the real-time data, parameter range and basic weight of each parameter;

[0016] Determine whether to trigger the first warning condition according to the risk coefficient and the coefficient threshold;

[0017] Set up a local cache for storing the data collected within a preset time period.

[0018] Preferably, the obtaining the risk coefficient according to the real-time data, parameter range and basic weight of each parameter; includes:

[0019] ;

[0020] ;

[0021] ;

[0022] DG is the comprehensive risk coefficient, is the risk coefficient of the i-th parameter; is the current weight of the i-th parameter; is the basic weight of the i-th parameter; is the j-th sampling value of the i-th parameter currently; is the maximum value of the i-th parameter range; is the minimum value of the i-th parameter range; The j-th sub-sampling value of the i-th parameter, m is a positive integer, m < j and m 5; is the threshold of the change rate between the j-th sampling of the i-th parameter and the (j - k)-th sampling value of the i-th parameter, > 0; max() is to take the maximum value; abs() is to take the absolute value; n is the number of parameters.

[0023] Preferably, determining whether to trigger the first warning condition according to the risk coefficient and the coefficient threshold; includes:

[0024] If the risk coefficient or the comprehensive risk coefficient of any parameter is greater than its first preset threshold, a first warning of the first level is given;

[0025] If the risk coefficient or the comprehensive risk coefficient of any parameter is greater than its second preset threshold, or multiple risk coefficients are greater than its first threshold, and among the parameters whose risk coefficients are greater than its first threshold, the current weights of two or more parameters are greater than their basic weights, a first warning of the second level is given;

[0026] If the risk coefficient or the comprehensive risk coefficient of any parameter is greater than its third preset threshold, or multiple risk coefficients are greater than its second threshold, and among the parameters whose risk coefficients are greater than its second threshold, the current weights of two or more parameters are greater than their basic weights, a first warning of the third level is given;

[0027] Wherein the first threshold of the risk coefficient or the comprehensive risk coefficient of each parameter is less than its second threshold, and the second threshold of the risk coefficient or the comprehensive risk coefficient of each parameter is less than its third threshold.

[0028] Preferably, the S3 includes:

[0029] According to the location information of the wearer, obtain the risk coefficients and comprehensive risk coefficients of each parameter of each wearer in the same area;

[0030] Determine the transmission coefficient according to the risk coefficients and comprehensive risk coefficients of each parameter of the wearer;

[0031] Determine the priority transmission order of each wearer's parameters according to the size of the transmission coefficient; on the premise of network congestion, transmit the data of the wearer to the cloud platform in turn according to the priority transmission order;

[0032] Establish a machine learning model according to the physiological parameters, environmental parameters and risk coefficients of multiple wearers;

[0033] Based on the real-time parameters of the wearer, obtain the risk level and risk probability of each wearer through a machine learning model;

[0034] Determine whether to reactivate the audible and visual alarm device on the protective clothing according to the risk level and risk probability.

[0035] Preferably, the determining the transmission coefficient according to the risk coefficients of each parameter and the comprehensive risk coefficient of the wearer includes:

[0036] ;

[0037] where T is the transmission coefficient, is the first preset threshold of the risk coefficient of the i-th parameter; is the first preset threshold of the comprehensive risk coefficient.

[0038] Preferably, the establishing a machine learning model according to the physiological parameters, environmental parameters and risk coefficients of multiple wearers includes:

[0039] Extract features from the original data, and the features should reflect the physiological state, health status of the wearer and the impact of the environment on him;

[0040] Through principal component analysis, select the features that have an important impact on the model prediction performance as alternative features;

[0041] Combine physiological and environmental features to construct composite features;

[0042] Input the alternative features and composite features into the machine learning model; optimize according to the model training results.

[0043] Preferably, the risk level includes no risk, the first level, the second level and the third level;

[0044] If the risk probabilities of the first level, the second level and the third level of the wearer are all greater than their preset thresholds, select the highest level for alarm.

[0045] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the method described in any one of this application.

[0046] The beneficial effects of the present invention include: by continuously collecting the physiological parameters (such as body temperature, heart rate, blood oxygen concentration) and environmental parameters (such as temperature, humidity, gas concentration) of the wearer and performing real-time analysis, potential risks can be detected in advance, and audible and visual alarm devices at different levels can be triggered in a timely manner to remind the wearer to pay attention to the current situation; according to the risk coefficient and preset threshold, the first, second, or third-level alarm is automatically triggered to ensure that appropriate countermeasures can be taken at different risk levels, avoiding false alarms or missed alarms, and greatly improving safety; the microprocessor or edge device built into the protective clothing is used to perform real-time analysis on the collected data, reducing the time delay of data transmission and cloud processing, and improving the response speed of the system; the transmission coefficient is determined according to the risk coefficient of each parameter and the comprehensive risk coefficient, and the data is uploaded in order of priority to ensure that key information can still be transmitted to the cloud platform in a timely manner under network congestion, further accelerating the emergency response speed; a local cache is set up to store the data collected within a preset time period to ensure that the data will not be lost during network interruption and will be uploaded after the network is restored to ensure data integrity; by combining the physiological parameters, environmental parameters, and historical data of multiple wearers, a machine learning model is established to predict the risk levels and probabilities of each wearer, enhancing the predictability and reliability of the system; through wireless transmission technology, the alarm information is sent to the background monitoring center or designated contacts, facilitating the timely intervention of external rescue forces and forming a multi-level security guarantee system; the cloud platform is used to centrally manage and analyze the data of multiple wearers, supporting large-scale deployment and remote monitoring, and improving management efficiency; by obtaining the risk coefficients of each parameter and the comprehensive risk coefficient of each wearer in the same area through location information, collaborative work and unified scheduling within the area are realized, and the overall work efficiency is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 FIG. is a schematic diagram of an audible and visual alarm method for a heat-resistant protective clothing based on the Internet of Things provided by an embodiment of the present application; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] Next, in combination with the drawings and specific embodiments, the present application will be further described. It should be noted that, on the premise of no conflict, the following-described embodiments or technical features can be combined arbitrarily to form new embodiments.

[0049] See Figure 1 , some embodiments of the present application provide an audible and visual alarm method for a heat-resistant protective clothing based on the Internet of Things, and the method includes:

[0050] S1. Collect the physiological parameters of the wearer and the surrounding environmental parameters through the sensors in the protective clothing; collect the position parameters of the wearer through the positioning device;

[0051] S2. Analyze the collected data through the microprocessor or edge device built into the protective clothing to obtain a risk coefficient, where the risk coefficient includes the risk coefficients of each parameter and the comprehensive risk coefficient; determine whether to trigger the first warning condition based on the risk coefficient, and if so, activate the audible and visual alarm device on the protective clothing;

[0052] S3. Determine the transmission coefficient according to the risk coefficients of each parameter and the comprehensive risk coefficient of the wearer; upload the collected data and the first analysis result to the cloud platform according to the transmission coefficient; obtain the risk level and risk probability of each wearer through a machine learning model; determine whether to trigger the second warning condition according to the risk level and risk probability; if so, activate the audible and visual alarm device on the protective clothing;

[0053] S4. Use wireless transmission to send the warning information to the background monitoring center or designated contact person.

[0054] In some embodiments, the physiological parameters include but are not limited to body temperature, heart rate, and blood oxygen concentration; the environmental parameters include but are not limited to temperature, humidity, and gas concentration.

[0055] The working principle of the above technical solution is as follows: The sensors embedded in the protective clothing can continuously monitor physiological parameters of the wearer, such as body temperature, heart rate, blood oxygen concentration, etc. These parameters are crucial for evaluating the physical condition of the wearer. At the same time, the sensors also collect environmental parameters such as the temperature, humidity, and harmful gas concentration of the environment where the wearer is located to assess the potential threats that the external environment may pose to the wearer. Through the built-in GPS or other positioning devices, the system can accurately obtain the location information of the wearer. The microprocessor or edge device in the protective clothing will perform real-time analysis on the collected data, calculate the risk coefficients of each parameter and the comprehensive risk coefficient. The risk coefficient reflects the risk level that the wearer currently faces. According to the preset warning conditions (such as the risk coefficient threshold), the system determines whether to trigger the first warning condition. Once the condition is met, the audible and visual alarm device on the protective clothing will be immediately activated to visually alert the wearer to pay attention and take corresponding measures. Based on the risk coefficient of the wearer, the system calculates the transmission coefficient to determine the priority transmission order of the data. In this way, under the limited network bandwidth, it can ensure that the most important data is uploaded to the cloud platform first. The data uploaded to the cloud platform will undergo further analysis and processing, and machine learning models are used to predict the risk level and risk probability of the wearer. These prediction results help the back-end monitoring center or designated contact person to more comprehensively understand the safety status of the wearer. The cloud platform determines whether to trigger the second warning condition based on the predicted risk level and risk probability. If the condition is met, it will send an instruction to the protective clothing to activate the audible and visual alarm device again and send the warning information to the back-end monitoring center or designated contact person through wireless transmission. Once the warning condition is triggered, the system will send detailed warning information (including the location, physiological parameters, environmental parameters, and risk level of the wearer, etc.) to the back-end monitoring center or designated contact person through wireless transmission. After receiving the warning information, the back-end monitoring center or designated contact person can quickly take actions, such as dispatching a rescue team, adjusting the working environment, or providing necessary medical support, etc., to ensure the safety and health of the wearer.

[0056] The effects of the above technical solution are as follows: By collecting the physiological parameters of the wearer (such as body temperature, heart rate, etc.), the surrounding environment parameters (such as temperature, humidity, harmful gas concentration, etc.), and the location information in real time, the comprehensive monitoring of the safety status of the wearer is ensured. Once a dangerous situation is detected, the acoustic and optical alarm device on the protective clothing will be immediately activated, providing immediate and intuitive warning information for the wearer, which helps the wearer quickly take countermeasures and reduce the accident risk. By quickly analyzing the collected data through the built-in microprocessor or edge device, calculating the risk coefficients and comprehensive risk coefficients of each parameter, the rapid and intelligent assessment of potential risks is realized; combined with the machine learning model, it can further predict the risk level and risk probability of the wearer, providing more accurate early warning information for the background monitoring center or designated contact person, which helps to take preventive measures in advance. Determine the transmission coefficient according to the risk coefficient of the wearer, and give priority to uploading important data to the cloud platform to ensure the efficient transmission and effective utilization of data. The cloud platform can remotely monitor the safety status of multiple wearers. Once an abnormal situation is found, it can quickly notify the background monitoring center or designated contact person to achieve remote monitoring and rapid response; this method can not only trigger the acoustic and optical alarm locally, but also send the warning information to the background monitoring center or designated contact person through wireless transmission, which helps to expand the warning range and improve the emergency response speed. The background monitoring center or designated contact person can quickly take actions according to the warning information, such as dispatching rescue teams, adjusting the working environment, etc., to ensure the safety of the wearer to the greatest extent.

[0057] In some embodiments, the S2 includes:

[0058] Set the range and basic weight of each parameter; obtain the risk coefficient according to the real-time data, parameter range, and basic weight of each parameter;

[0059] According to the risk coefficient and coefficient threshold, determine whether to trigger the first warning condition;

[0060] Set up a local cache for storing the data collected within a preset time period.

[0061] In some embodiments, the obtaining the risk coefficient according to the real-time data, parameter range, and basic weight of each parameter; includes:

[0062] ;

[0063] ;

[0064] ;

[0065] DG is the comprehensive risk coefficient, is the risk coefficient of the i-th parameter; is the current weight of the i-th parameter; is the basic weight of the i-th parameter; is the j-th sampling value of the i-th parameter currently; is the maximum value of the range of the i-th parameter; is the minimum value of the range of the i-th parameter; the j-th sampling value of the i-th parameter, m is a positive integer, m < j and m 5; is the threshold of the change rate between the j-th sampling of the i-th parameter and the j-th sampling value of the i-th parameter, > 0; max() is to take the maximum value; abs() is to take the absolute value; n is the number of parameters; Deltayk = Deltay1 * f(k); f(k) is a functional relationship, f(k) can be α×(k−1), or can also be + bk + c; where α, a, b, c are constants, and different parameters correspond to different f(k).

[0066] The working principle of the above technical solution is as follows: First, a reasonable range (maximum value Cimax and minimum value Cimin) and a basic weight wi are set for each parameter (such as body temperature, heart rate, ambient temperature, etc.). These ranges and weights are set based on expert experience, knowledge base, and / or historical models, and are based on an in-depth understanding of the wearer's safety status, aiming to reflect the influence degree of each parameter on the overall danger level. For each parameter, a danger coefficient is calculated according to the relationship between its real-time sampling value and the set range. This calculation process considers three situations:

[0067] When the sampling value exceeds the maximum value, the danger coefficient is calculated based on the relative size of the exceeded part; when the sampling value is lower than the minimum value, the danger coefficient is also calculated based on the relative size of the lower part; of course, if the parameter only has a maximum value range or a minimum value range, only the situation of exceeding the maximum value or being lower than the minimum value is considered for this part of the operation;

[0068] When the sampled value is within the set range, consider the change rate of the sampled value compared with its previous several sampled values (up to the previous 5 times, that is, k = 1 to 5), and calculate the risk coefficient based on the relationship between these change rates and the set threshold Deltayk; for example, if there were two samplings before the current sampling, then compare the current sampled value with the first sampling and the second sampling respectively to obtain two change rates, and obtain two values according to the relationship between the two change rates and their corresponding thresholds Deltay1 and Deltay2, and take the maximum value of the two values as the current risk coefficient of this parameter; another example is that if there were 10 samplings before the current sampling, compare the current sampled value with the adjacent 5 sampled values respectively to obtain 5 change rates, and obtain five values according to the relationship between the two change rates and their corresponding thresholds Deltay1~Deltay5, and take the maximum value of the five values as the current risk coefficient of this parameter; during the calculation process, the absolute value function abs() is used to ensure the accuracy and rationality of the calculation result; among them, Deltayk is obtained through the knowledge base, expert experience and / or historical model; Deltayk = Deltay1*f(k); f(k) is a functional relationship, and f(k) can be α×(k−1), or it can be +bk + c; where α, a, b, and c are constants, and different parameters correspond to different f(k).

[0069] According to the risk coefficient DGi and the base weight wi of each parameter, calculate a current weight wci. This current weight reflects the degree of influence of this parameter on the overall risk level in the current situation; by adjusting the current weight, the importance of different parameters in different situations can be more flexibly reflected. Multiply the current weights wci of all parameters by their corresponding risk coefficients DGi and add the results to obtain a comprehensive risk coefficient DG. This comprehensive risk coefficient reflects the overall risk level faced by the wearer currently.

[0070] Compare the calculated comprehensive risk coefficient DG with the preset coefficient threshold. If DG exceeds this threshold, it is considered that the first alarm condition is triggered. At this time, activate the audible and visual alarm device on the protective clothing to remind the wearer to pay attention and take corresponding measures.

[0071] For the convenience of subsequent analysis and troubleshooting, a local cache is set to store the data collected within a preset time period. These data include the real-time sampled values of each parameter, the calculated risk coefficients and comprehensive risk coefficients, etc.

[0072] The effects of the above technical solution are as follows: By setting reasonable ranges and weights for each parameter and calculating the risk coefficient based on real-time sampling values, this technology can accurately evaluate the current safety status of the wearer. Considering the change rate of the sampling value compared with the previous sampling values and combining with the set threshold to calculate the risk coefficient further improves the accuracy and sensitivity of the alarm. The calculation process of the risk coefficient not only considers whether the sampling value exceeds the set range but also the change trend of the sampling value, which enables this technology to adapt to the safety monitoring requirements in different environments and working conditions. By adjusting the current weight, it can more flexibly reflect the importance of different parameters in different situations, enhancing the adaptability and flexibility of the technology. Once the comprehensive risk coefficient exceeds the preset threshold, the audible and visual alarm device will be immediately activated to provide timely alarm information for the wearer. This immediate response mechanism helps the wearer quickly take countermeasures and reduce potential safety risks. A local cache is set to store the data collected within a preset time period, including the real-time sampling values of each parameter, the calculated risk coefficient, and the comprehensive risk coefficient, etc.; these data provide strong support for subsequent analysis and troubleshooting, helping to continuously improve and optimize the alarm system; by comprehensively considering the real-time data of each parameter, the parameter range, and the basic weight, the risk coefficient DGi of each parameter can be accurately calculated; not only the absolute value of the parameter is considered but also its change relative to the set range, so as to more accurately reflect the current safety status of the wearer; the current weight wci in the formula is calculated based on the risk coefficient DGi and the basic weight wi, which means that as the real-time data of the parameter changes, the current weight will also be adjusted accordingly. This mechanism of dynamically adjusting the weight enables the comprehensive risk coefficient DG to more accurately reflect the importance of each parameter in the current situation. By multiplying the current weight wci of each parameter by its corresponding risk coefficient DGi and adding the results, the comprehensive risk coefficient DG is obtained; this comprehensive risk coefficient can comprehensively reflect the overall risk level faced by the wearer currently; when DG exceeds the preset coefficient threshold, the alarm condition can be triggered, thus improving the accuracy and timeliness of the alarm. The change rate threshold Deltayk in the formula is adjusted through the functional relationship f(k), which enables the system to optimize the threshold setting according to historical data and expert experience.

[0073] In some embodiments, determining whether to trigger the first alarm condition according to the risk coefficient and the coefficient threshold includes:

[0074] If the risk coefficient or the comprehensive risk coefficient of any parameter is greater than its first preset threshold, a first-level first alarm is given.

[0075] If the risk coefficient or the comprehensive risk coefficient of any parameter is greater than its second preset threshold, or multiple risk coefficients are greater than its first threshold, and among the parameters with risk coefficients greater than its first threshold, the current weights of two or more parameters are greater than their base weights, then a first alarm at the second level is given;

[0076] If the risk coefficient or the comprehensive risk coefficient of any parameter is greater than its third preset threshold, or multiple risk coefficients are greater than its second threshold, and among the parameters with risk coefficients greater than its second threshold, the current weights of two or more parameters are greater than their base weights, then a first alarm at the third level is given;

[0077] Among them, the first threshold of the risk coefficient or the comprehensive risk coefficient of each parameter is less than its second threshold, and the second threshold of the risk coefficient or the comprehensive risk coefficient of each parameter is less than its third threshold.

[0078] The working principle of the above technical solution is as follows: First, according to the real-time data, parameter range, and base weight of each parameter, calculate the risk coefficient DGi of each parameter, multiply the current weight wci of each parameter by its corresponding risk coefficient DGi, and add the results to obtain the comprehensive risk coefficient DG.

[0079] Set multiple preset thresholds for the risk coefficient and the comprehensive risk coefficient of each parameter, including the first preset threshold, the second preset threshold, and the third preset threshold. These thresholds are set according to expert experience, knowledge base, and / or historical models, aiming to reflect the alarm requirements under different risk levels. Among them, the first preset threshold is the lowest, and the third preset threshold is the highest, that is, the first threshold < the second threshold < the third threshold.

[0080] First-level alarm: When the risk coefficient or the comprehensive risk coefficient of any parameter exceeds its first preset threshold, the system will trigger a first-level alarm. This indicates that there is a relatively low level of risk currently, and the wearer needs to pay attention.

[0081] Second-level alarm: When the risk coefficient or the comprehensive risk coefficient of any parameter exceeds its second preset threshold, or multiple risk coefficients exceed its first threshold and the current weights of at least two of them are greater than their base weights, the system will trigger a second-level alarm. This indicates that the current risk level is relatively high, and the wearer needs to take more proactive countermeasures.

[0082] Third-level alarm: When the risk coefficient or the comprehensive risk coefficient of any parameter exceeds its third preset threshold, or multiple risk coefficients exceed its second threshold and the current weights of at least two of them are greater than their base weights, the system will trigger a third-level alarm. This indicates that there is an extremely high risk currently, and the wearer needs to take immediate action to avoid potential risks.

[0083] Once the system determines that an alarm needs to be triggered, it will activate the audible and visual alarm device on the protective clothing to alert the wearer and prompt them to take corresponding measures. The higher the alarm level, the stronger the response of the audible and visual alarm device will be to attract the wearer's sufficient attention.

[0084] At the same time as triggering the alarm, the system will also store relevant data (including real-time sampling values of various parameters, calculated risk coefficients, and comprehensive risk coefficients, etc.) in the local cache for subsequent analysis and troubleshooting.

[0085] The effects of the above technical solution are as follows: By setting three different levels of alarm conditions (the first, second, and third levels), it can more precisely reflect the degree of danger faced by the wearer and ensure appropriate response measures can be taken at different risk levels. When the risk coefficient of any parameter or the comprehensive risk coefficient exceeds the first preset threshold, it is triggered to promptly alert the wearer to pay attention to the current situation and prevent potential risks from deteriorating further. When the risk coefficient of any parameter or the comprehensive risk coefficient exceeds the second preset threshold, or when the risk coefficients of multiple parameters exceed the first threshold simultaneously and the current weights of two or more of these parameters are greater than their base weights, it is triggered to prompt the wearer to take more urgent measures, such as immediately evacuating the dangerous area. When the risk coefficient of any parameter or the comprehensive risk coefficient exceeds the third preset threshold, or when the risk coefficients of multiple parameters exceed the second threshold simultaneously and the current weights of two or more of these parameters are greater than their base weights, it indicates that the situation is extremely critical and the highest-level emergency response needs to be initiated immediately. Different application scenarios can set different thresholds to make the system more flexible in adapting to various working environments. For example, in a high-temperature environment, the threshold for the temperature parameter can be appropriately relaxed, while in an environment with toxic gases, the threshold for the gas concentration parameter should be more stringent.

[0086] It not only focuses on the risk coefficient of a single parameter but also combines the comprehensive influence of multiple parameters. Especially when multiple parameters exceed the threshold simultaneously, it can more accurately evaluate the overall degree of danger and avoid false alarms or missed alarms. By setting different levels of alarm conditions, the system can issue warnings at an early stage, thus avoiding waste of resources caused by overreaction. Only when there is indeed a relatively high risk will a higher-level alarm be triggered, improving the reliability and resource utilization rate of the system. The condition of "the current weight is greater than the base weight" is introduced to ensure that a higher-level alarm is triggered only when the key parameters are truly abnormal, reducing false alarms caused by fluctuations in a single parameter and improving the accuracy of the system.

[0087] In some embodiments, S3 includes:

[0088] According to the location information of the wearer, obtain the risk coefficients and comprehensive risk coefficients of each parameter of each wearer in the same area;

[0089] Determine the transmission coefficient according to the risk coefficients of various parameters and the comprehensive risk coefficient of the wearer;

[0090] Determine the priority transmission order of each wearer parameter according to the magnitude of the transmission coefficient; on the premise of network congestion, transmit the data of the wearer to the cloud platform in sequence according to the priority transmission order;

[0091] Establish a machine learning model based on the physiological parameters, environmental parameters and risk coefficients of multiple wearers;

[0092] Obtain the risk level and risk probability of each wearer through the machine learning model according to the real-time parameters of the wearer;

[0093] Determine whether to reactivate the audible and visual alarm device on the protective clothing according to the risk level and risk probability.

[0094] In some embodiments, the determining the transmission coefficient according to the risk coefficients of various parameters and the comprehensive risk coefficient of the wearer includes:

[0095] ;

[0096] where T is the transmission coefficient, is the first preset threshold of the risk coefficient of the i-th parameter; is the first preset threshold of the comprehensive risk coefficient.

[0097] The working principle and effects of the above technical solution are as follows: By using the location information of the wearers, all the wearers in the same area are determined, and their respective parameter risk coefficients and comprehensive risk coefficients are obtained; the location information of each wearer is obtained through positioning devices such as GPS, Wi-Fi, and BLE beacons. The working area is divided into several small areas (such as rooms, floors, work areas) according to the actual application scenario for more refined management. The physiological parameters, environmental parameters, and the calculated respective parameter risk coefficients and comprehensive risk coefficients are synchronized among all the wearers in the same area; according to the risk coefficients and comprehensive risk coefficients of the respective parameters, the transmission coefficient T is calculated to determine the priority of data transmission. The formula takes into account the deviation of the risk coefficient of each parameter from its first preset threshold (DGiy1), and the deviation of the comprehensive risk coefficient from its first preset threshold (DGy1); by adding these deviations, a transmission coefficient is obtained that can reflect the degree of deviation of the current risk state of the wearer from the preset threshold; the magnitude of the transmission coefficient directly determines the priority transmission order of the wearer's data, and the larger the transmission coefficient, the higher the priority. In the case of network congestion, the system will preferentially transmit the data of those wearers with higher transmission coefficients (i.e., greater degree of deviation, which may represent a higher risk state); collect the physiological parameters, environmental parameters, and risk coefficients of multiple wearers, and use these data to train a machine learning model; this model can learn the complex relationship between the wearer's parameters and the risk level and risk probability, so as to achieve accurate prediction of new data. By continuously updating and optimizing the model, the system can continuously improve the prediction ability of the wearer's risk state. The system inputs the real-time parameters of the wearer into the trained machine learning model, and the model will calculate the risk level and risk probability of the wearer according to these parameters; the risk level is a qualitative assessment indicating the degree of risk the wearer is currently in; while the risk probability is a quantitative assessment indicating the likelihood of the wearer experiencing a dangerous event in a future period of time; according to the calculated risk level and risk probability, they are compared with the preset warning conditions. If the wearer's risk level is high or the risk probability exceeds a certain threshold, the audible and visual alarm device on the protective clothing will be activated again to remind the wearer to pay attention and take corresponding measures; this mechanism ensures that the system can issue an alarm in a timely manner when the wearer faces potential risks, providing effective safety protection for the wearer.

[0098] In some embodiments, establishing a machine learning model according to the physiological parameters, environmental parameters, and risk coefficients of multiple wearers; includes:

[0099] Extract features from the original data, and the features should reflect the physiological state, health status of the wearer, and the impact of the environment on them;

[0100] Through principal component analysis, select the features that have an important impact on the model prediction performance as alternative features;

[0101] Construct composite features by combining physiological and environmental characteristics;

[0102] Input the alternative features and composite features into a machine learning model; optimize according to the model training results.

[0103] In some embodiments, the risk levels include no risk, the first level, the second level, and the third level;

[0104] If the risk probabilities of the wearer at the first level, the second level, and the third level are all greater than their preset thresholds, select the highest level for alarm.

[0105] The working principle of the above technical solution is as follows: First, a series of features will be extracted from the original data, which are designed to comprehensively reflect the physiological state, health condition of the wearer, and the potential impact of the environment on the wearer. For example, physiological characteristics may include heart rate, blood pressure, body temperature, etc., while environmental characteristics may include temperature, humidity, air pressure, etc. Next, through techniques such as principal component analysis (PCA), the extracted features are screened to select those features that have an important impact on the model prediction performance as alternative features. The purpose of this step is to reduce the number of features, improve the model training efficiency and prediction accuracy. After selecting the alternative features, composite features will also be constructed by combining physiological and environmental characteristics. Composite features are obtained by combining or transforming multiple simple features, and they can more accurately reflect the actual state of the wearer or environmental changes. For example, combining the heart rate data of the wearer with the environmental temperature data to form a new composite feature. This feature can reflect the change of the wearer's heart rate under different environmental temperatures. Specifically, under extreme conditions (such as high or low temperature environments), this composite feature can help the system determine whether the wearer is physically uncomfortable due to too high or too low environmental temperature, and thus issue a warning in a timely manner. For example, combining the blood pressure data of the wearer with the environmental humidity data to form another composite feature. This feature can reveal the impact of humidity changes on the wearer's blood pressure; because in a high humidity environment, the human body may feel stuffy and uncomfortable, which in turn affects blood circulation and blood pressure. By monitoring this composite feature, the system can timely detect the blood pressure abnormality of the wearer caused by humidity changes and take corresponding safety measures.

[0106] Finally, input the alternative features and composite features into a machine learning model for model training. During the training process, the model parameters will be continuously adjusted to minimize the prediction error and improve the model performance. Once the model training is completed, the system will also optimize the model according to the training results to ensure its accuracy and stability in practical applications.

[0107] After the machine learning model is established and optimized, the model will calculate the probability that the wearer is in different danger levels based on the wearer's real-time parameters. Danger levels are usually divided into no danger, level one, level two, and level three, and each level corresponds to a different degree of danger and response measures. If the probability that the wearer is in multiple danger levels at the same time is greater than its preset threshold (for example, the probability of danger at the first level, the probability of danger at the second level, and the probability of danger at the third level all exceed their respective thresholds), the system will select the highest level of danger to alarm. The alarm mechanism is usually implemented by activating the sound and light alarm device on the protective clothing to remind the wearer to pay attention and take appropriate measures to deal with potential dangers.

[0108] The effect of the above technical solution is: by extracting rich features from the original data and combining the principal component analysis (PCA) and other techniques for feature screening, the system can select features that have an important impact on the model prediction performance, thereby improving the accuracy of the early warning. The construction of composite features further enhances the model's ability to reflect the actual state of the wearer or environmental changes, making the early warning more accurate and timely. By reducing the number of features, the system can reduce the complexity of model training and improve training efficiency. The optimized machine learning model shows higher stability and accuracy in practical applications, reducing the possibility of false alarms and missed alarms. It can monitor the wearer's physiological parameters and environmental parameters in real time, and calculate the probability of the wearer being at different levels of danger based on these parameters. When the wearer faces potential danger, the system can issue an early warning in time, and activate the sound and light alarm device on the protective clothing to remind the wearer to pay attention and take corresponding measures, thereby effectively avoiding or reducing the harm of dangerous events to the wearer.

[0109] The present application also provides an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of any method described in the present application are implemented.

[0110] This application is explained from the perspectives of purpose of use, effectiveness, progress and novelty, and has met the functional enhancement and usage requirements emphasized by the Patent Law. The above description and drawings of this application are only the preferred embodiments of this application, and are not intended to limit this application. Therefore, all structures, devices, features, etc. that are similar or identical to this application, that is, all equivalent replacements or modifications made in accordance with the scope of the patent application of this application, should fall within the scope of protection of the patent application of this application.

Claims

1. A sound and light alarm method for heat-resistant protective clothing based on the Internet of Things, characterized in that: The method comprises: S1. Collect the wearer's physiological parameters and surrounding environment parameters through sensors in the protective clothing; collect the wearer's position parameters through a positioning device; S2. Analyze the collected data through the microprocessor or edge device built into the protective suit to obtain a risk factor, wherein the risk factor includes the risk factor of each parameter and the comprehensive risk factor; determine whether to trigger the first alarm condition according to the risk factor, and if so, activate the sound and light alarm device on the protective suit; S3. Determine the transmission coefficient according to the risk coefficients of various parameters and the comprehensive risk coefficient of the wearer; upload the collected data and the first analysis result to the cloud platform according to the transmission coefficient; obtain the risk level and risk probability of each wearer through the machine learning model; determine whether to trigger the second alarm condition according to the risk level and risk probability; if so, activate the sound and light alarm device on the protective clothing; S4. Use wireless transmission to send the alarm information to the background monitoring center or designated contact person; The S2 includes: Set the range and basic weight of each parameter; obtain the risk factor based on the real-time data, parameter range, and basic weight of each parameter; Determine whether to trigger the first alarm condition according to the risk factor and the factor threshold; Set up a local cache to store data collected within a preset time period; The method of obtaining the risk factor according to the real-time data, parameter range and basic weight of each parameter includes: ; ; ; DG is the comprehensive risk coefficient, is the risk coefficient of the i-th parameter; is the current weight of the i-th parameter; is the basic weight of the i-th parameter; is the j-th sampling value of the i-th parameter; is the maximum value of the range of the i-th parameter; is the minimum value of the range of the i-th parameter; The m-th sampling value of the i-th parameter, m is a positive integer, m < j and ; ; is the threshold of the change rate between the j-th sampling and the m-th sampling value of the i-th parameter, ; max() is to take the maximum value; abs() is to take the absolute value; n is the number of parameters; ; The S3 includes: According to the wearer's location information, the risk coefficients of various parameters and the comprehensive risk coefficients of various wearers in the same area are obtained; Determine the transmission coefficient based on the wearer's risk factors of various parameters and the comprehensive risk factor; Determine the priority transmission order of each wearer's parameters according to the transmission coefficient; under the premise of network congestion, transmit the wearer's data to the cloud platform in sequence according to the priority transmission order; Establish a machine learning model based on the physiological parameters, environmental parameters and risk factors of multiple wearers; According to the real-time parameters of the wearer, the danger level and danger probability of each wearer are obtained through machine learning model; Determining whether to reactivate the sound and light alarm device on the protective clothing according to the danger level and danger probability; The method of determining the transmission coefficient according to the risk coefficients of various parameters of the wearer and the comprehensive risk coefficient comprises: ; Where T is the transmission coefficient, is the first preset threshold value of the ith parameter risk factor; is the first preset threshold of the comprehensive risk factor, DG is the comprehensive risk factor, is the risk coefficient of the i-th parameter.

2. The method according to claim 1, characterized in that The physiological parameters include body temperature, heart rate and blood oxygen concentration; the environmental parameters include temperature, humidity and gas concentration.

3. The method according to claim 1, characterized in that The step of determining whether to trigger a first alarm condition according to the risk factor and the factor threshold comprises: If the risk factor or the comprehensive risk factor of any parameter is greater than its first preset threshold, a first level first alarm is issued; If the risk factor or the comprehensive risk factor of any parameter is greater than its second preset threshold, or multiple risk factors are greater than their first threshold, and among the parameters whose risk factors are greater than their first threshold, the current weights of two or more parameters are greater than their basic weights, a first alarm of the second level is issued; If the risk factor or the comprehensive risk factor of any parameter is greater than its third preset threshold, or multiple risk factors are greater than their second threshold, and among the parameters whose risk factors are greater than their second threshold, the current weights of two or more parameters are greater than their basic weights, a first alarm of the third level is issued; The first threshold of the risk factor or the comprehensive risk factor of each parameter is smaller than the second threshold, and the second threshold of the risk factor or the comprehensive risk factor of each parameter is smaller than the third threshold.

4. The method according to claim 1, characterized in that: The method of establishing a machine learning model based on the physiological parameters, environmental parameters and risk factors of multiple wearers includes: Extract features from the raw data, the features should reflect the wearer's physiological state, health status and the impact of the environment on it; Through principal component analysis, features that have a significant impact on the model prediction performance are selected as candidate features; Combine physiological and environmental features to construct composite features; Input candidate features and composite features into the machine learning model; optimize based on the model training results.

5. The method according to claim 1, characterized in that The hazard levels include non-hazardous, first level, second level and third level; If the wearer's first level danger probability, second level danger probability and third level danger probability are all greater than their preset thresholds, the highest level is selected for alarm.

6. An electronic device, characterized in that: The electronic device comprises a memory and a processor, the memory stores a computer program, and the processor implements the steps of the method according to any one of claims 1 to 5 when executing the computer program.

Images