An intelligent safety helmet management system

Abstract

The application provides an intelligent safety helmet management system, which comprises a plurality of intelligent terminals, an edge collaborative gateway and a remote monitoring center; the intelligent terminal collects local environment data in real time, and performs lightweight calculation and processing to obtain a safety risk score, a channel state, a cruising ability and a data timeliness level; the terminal processes communication information in cooperation with the edge collaborative gateway to generate a data processing strategy corresponding to the terminal; in the process of data uploading according to the strategy, the local environment data can be uploaded to the edge collaborative gateway according to the time slot specified in the strategy according to the safety risk score, the channel state, the cruising ability and the data timeliness level, or the data can be forwarded by other intelligent terminals belonging to the same collaborative cluster as collaborative terminals, so as to avoid untimely data uploading or data uploading failure due to different factors such as environment factors, network factors or terminal self factors.

Description

Technical Field

[0001] This invention relates to the field of smart wearable device technology, and in particular to a smart safety helmet management system. Background Technology

[0002] In high-risk industries such as mining, petrochemicals, and construction, smart safety helmets are gradually replacing traditional safety helmets, becoming core equipment for personnel safety monitoring. Modern smart safety helmets integrate cameras, gas sensors, temperature sensors, and positioning modules, enabling real-time monitoring of the surrounding environment and physiological state of personnel. This massive amount of multimodal data needs to be transmitted back to the monitoring center in real time via wireless networks (such as Wi-Fi or 5G private networks) for centralized alarm and dispatching.

[0003] However, existing systems face significant challenges: First, in densely populated work areas, the simultaneous transmission of high-definition video and sensor data from numerous safety helmets quickly consumes limited wireless bandwidth, leading to network congestion and delays or even loss of critical alarm information. Second, the complex industrial environment, with metal structures and equipment obstructions causing rapid wireless signal attenuation and unstable links, means that safety helmets in signal dead zones may become "information islands." Finally, most existing systems employ simple polling or contention-based access mechanisms, failing to differentiate data importance and prioritizing the transmission of high-risk alarms when bandwidth is insufficient. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies, such as intense competition for wireless resources, inability to guarantee critical alarms, and low transmission reliability in harsh channel environments, when multiple smart helmets transmit concurrently. The invention provides a smart helmet management system that enables cross-layer optimization of perception and communication and supports inter-terminal collaboration.

[0005] To achieve the above objectives, the present invention provides an intelligent safety helmet management system, comprising:

[0006] Multiple smart terminals, an edge collaboration gateway, and a remote monitoring center, wherein each of the smart terminals is mounted on a safety helmet;

[0007] For each of the aforementioned smart terminals, the smart terminal is used to collect local environmental data in real time and send communication information to the edge collaboration gateway; upon obtaining the data processing strategy fed back by the edge collaboration gateway, the target data is sent to the data receiving end based on the data processing strategy and the communication information.

[0008] The communication information includes terminal information, channel status information, and security risk score of the smart terminal. The terminal information includes the battery life and data timeliness level of the smart terminal. The data processing strategy includes a data collaboration list. The data collaboration list includes the communication information, data upload time slots, and terminal collaboration instructions of each smart terminal belonging to the same collaboration cluster in the smart helmet management system. The data receiving end is the edge collaboration gateway or a smart terminal belonging to at least one of the same target collaboration cluster. The target data includes at least the local environment data. The target collaboration cluster is a terminal set containing multiple smart terminals, including the smart terminal mentioned above.

[0009] The edge collaboration gateway is used to, when receiving communication information from each of the smart terminals in the smart helmet management system, determine the priority of each smart terminal based on the communication information, and set the data upload time slot for each smart terminal based on the priority; generate one or more collaboration clusters and a data collaboration list corresponding to each collaboration cluster based on the communication information; generate a data processing strategy corresponding to the data collaboration list, and send the data processing strategy to each smart terminal in the collaboration cluster corresponding to the data collaboration list; when receiving target data uploaded by any smart terminal, update the global security situation diagram corresponding to each smart terminal, and upload the global security situation diagram to the remote monitoring center;

[0010] The remote monitoring center is used for monitoring and issuing related instructions based on the global security situation diagram.

[0011] Optionally, in the aforementioned intelligent helmet management system, the intelligent terminal includes:

[0012] Multimodal sensing unit, wireless transmission unit, and local processing unit;

[0013] The multimodal sensing unit is used to collect local environmental data in real time;

[0014] The wireless transmission unit is used to communicate with the edge collaboration gateway and each terminal within the target collaboration cluster;

[0015] The local processing unit is used to detect the terminal information and channel status information of the smart terminal in real time, and to perform a security risk score on the current environment of the smart terminal based on the local environment data to obtain communication information containing the terminal information, the channel status information, and the security risk score; when receiving a data processing strategy from the edge collaboration gateway based on the communication information from the wireless transmission unit, it determines whether the smart terminal is a collaboration terminal based on the security risk score; if the smart terminal is a non-collaboration terminal, it sets the edge collaboration gateway or a collaboration terminal within the target collaboration cluster as a data receiving end based on the security risk score and the data collaboration list in the data processing strategy; if the smart terminal is a collaboration terminal, it sets the edge collaboration gateway as a data receiving end; it generates target data carrying the local environment data, and sends the target data to the data receiving end through the wireless transmission unit.

[0016] Optionally, in the aforementioned intelligent safety helmet management system, the local processing unit determines whether the intelligent terminal is a collaborative terminal based on the safety risk score, specifically for:

[0017] Determine whether the safety risk score is lower than a first risk threshold, and whether the data collaboration list contains a collaboration identifier associated with the smart helmet management system; the collaboration identifier is the identifier of the candidate collaboration terminal marked by the edge collaboration gateway;

[0018] If the safety risk score is not lower than the first risk threshold, or if the data collaboration list does not contain a collaboration identifier associated with the smart helmet management system, the smart terminal is determined to be a non-collaboration terminal.

[0019] If the safety risk score is lower than the first risk threshold, and the data collaboration list contains a collaboration identifier associated with the smart helmet management system, the smart terminal is determined to be a collaboration terminal.

[0020] Optionally, in the aforementioned intelligent safety helmet management system, when the intelligent terminal is a non-cooperative terminal, the local processing unit is specifically used for:

[0021] The local environment data is encoded to generate target data, and it is determined whether the security risk score exceeds a second risk threshold, wherein the second risk threshold is greater than or equal to the first risk threshold.

[0022] If the security risk score does not exceed the second risk threshold, the edge collaboration gateway is set as the data receiving end, and when the data upload time slot corresponding to the smart terminal is reached, the target data is sent to the edge collaboration gateway through the wireless transmission unit.

[0023] If the security risk assessment exceeds the second risk threshold, the communication information and terminal information of each collaborative terminal in the terminal collaboration list of the target collaborative cluster are obtained; based on the communication information and terminal information, the optimal collaborative terminal is selected as the data receiving end, and the target data generated in real time is sent to the optimal collaborative terminal through the wireless transmission unit.

[0024] Optionally, in the aforementioned intelligent safety helmet management system, when the intelligent terminal is a collaborative terminal, the local processing unit is specifically used for:

[0025] Upon receiving target environment data sent by other smart terminals in the target collaborative cluster, load detection is performed on the smart terminal, where the target environment data is the local environment data of the other smart terminals;

[0026] When the load detection result indicates that the smart terminal is not fully loaded, the target environment data and the local environment data are fused to generate target data;

[0027] When the data upload time slot corresponding to the smart terminal is reached, the target data is sent to the edge collaboration gateway through the wireless transmission unit.

[0028] Optionally, in the aforementioned smart helmet management system, the edge collaboration gateway includes:

[0029] Resource scheduling unit, security situation assessment unit, uplink interface unit, and prediction and optimization unit;

[0030] The resource scheduling unit communicates with each of the intelligent terminals; the uplink interface unit communicates with the remote monitoring center.

[0031] The prediction and optimization unit is used to obtain the communication information of each of the smart terminals and the historical communication information of each smart terminal received by the resource scheduling unit; based on the current communication information and the historical communication information, it updates the weights corresponding to the terminal information, channel state information and security risk score in the communication information of each smart terminal, so as to obtain the current weight information corresponding to each smart terminal.

[0032] The resource scheduling unit is configured to calculate the priority of each smart terminal based on its communication information and current weight information, and set the data upload time slot for each smart terminal according to its priority; divide the smart terminals into multiple collaborative clusters according to their locations; mark the candidate collaborative terminals in each collaborative cluster according to the communication information of each smart terminal, and generate a collaborative instruction corresponding to each smart terminal; generate a data collaborative list for each collaborative cluster, and send a data processing strategy carrying the data collaborative list to each smart terminal, wherein the data collaborative list includes at least the priority, data upload time slot, communication information, and collaborative instruction of each smart terminal in the same collaborative cluster;

[0033] The security situation assessment unit is used to obtain target data of each of the smart terminals from the resource scheduling unit, and parse each of the target data to obtain local environment data of each smart terminal in the smart safety helmet management system. Based on each of the local environment data, a global security situation diagram corresponding to each smart terminal is generated, and the global security situation diagram and the local environment data of each smart terminal are sent to the remote monitoring center through the uplink interface unit.

[0034] Optionally, in the aforementioned intelligent safety helmet management system, the resource scheduling unit marks candidate collaborative terminals in each collaborative cluster based on the communication information of each intelligent terminal, specifically for:

[0035] Smart terminals with a security risk score below the third risk threshold are identified as the first candidate terminals.

[0036] Based on the terminal information and channel status information of each first candidate terminal, the battery life and channel quality of each first candidate terminal are determined.

[0037] The first candidate terminal whose battery life exceeds the target battery life and whose channel quality is excellent is marked as a candidate cooperative terminal, and a cooperative identifier corresponding to each candidate cooperative terminal is generated.

[0038] Optionally, in the aforementioned intelligent safety helmet management system, the prediction and optimization unit includes a cross-attention module;

[0039] The cross-attention module is used to identify the correlation between the current communication information and historical communication information of the same smart terminal, and dynamically capture the changing trends of the terminal information, channel state information and security risk score of the smart terminal according to the time series, and dynamically adjust the weights corresponding to the terminal information, channel state information and security risk score of the smart terminal according to the changing trends.

[0040] Optionally, in the aforementioned intelligent safety helmet management system, the prediction and optimization unit is further used for:

[0041] Determine the overall network load of the edge collaboration gateway;

[0042] Based on the overall network load, update the current weight information of each smart terminal.

[0043] Optionally, in the aforementioned intelligent safety helmet management system, the remote monitoring center is specifically used for:

[0044] Based on the global security situation diagram and the local environment data of each smart terminal, the system identifies whether there are high-risk areas in the global area monitored by the smart helmet management system. If a high-risk area exists, an alarm message corresponding to the high-risk area is issued, and an alarm command is sent to each smart terminal in the high-risk area through the edge collaboration gateway to trigger each smart terminal in the high-risk area to perform self-location alarm.

[0045] Compared with existing technologies, the intelligent safety helmet management system provided by this invention involves multiple intelligent terminals collecting local environmental data in real time and performing lightweight calculations and processing to obtain terminal communication information such as terminal safety risk scores, channel status, battery life, and data timeliness levels. This information is then processed in conjunction with an edge collaborative gateway to generate data processing strategies for each intelligent terminal. During data upload according to these strategies, the system can select the data upload time slot specified in the strategy to upload the local environmental data of each intelligent terminal to the edge collaborative gateway, based on the safety risk score, channel status, battery life, and data timeliness level. Alternatively, other intelligent terminals within the same collaborative cluster can act as collaborative terminals to forward the data, thus avoiding untimely or failed data uploads due to environmental, network, or terminal-specific factors.

[0046] This invention has the following advantages:

[0047] a) Resource allocation that maximizes security effectiveness: By directly mapping multiple factors such as security risks and terminal status to communication priorities, it ensures that high-risk and high-timeliness alarms can always obtain transmission resources first when the network is congested, thereby transforming limited wireless bandwidth into the highest security guarantee capability.

[0048] b) A business-driven immune self-organizing network mechanism was constructed: Unlike traditional routing technology based solely on network load balancing, this invention innovatively introduces the "security potential" and "node immunity" mechanisms to force high-risk terminals to exit relay services, which not only protects the terminal resources of high-risk personnel but also ensures the robustness of the relay network, reflecting the deep control of business logic over the communication topology.

[0049] c) Achieved cross-modal joint prediction with physical environment awareness: Unlike general LSTM timing prediction, this invention utilizes the strong physical coupling characteristics of "disaster-channel" in industrial scenarios, and uses risk characteristics to assist in calibrating channel prediction, which significantly improves the accuracy of link state prediction in sudden severe environments, thereby achieving more accurate preventive scheduling.

[0050] d) Cross-layer joint optimization and system self-evolution are achieved: the local processing unit can adaptively adjust the coding according to the network state prediction, realizing cross-layer optimization of perception and communication; the scheduling model can continuously learn and evolve based on historical data, making the system more and more intelligent with use. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0052] Figure 1 A system structure diagram of an intelligent safety helmet management system provided in an embodiment of the present invention;

[0053] Figure 2 This is a schematic diagram of the structure of an intelligent terminal in an intelligent safety helmet management system provided in an embodiment of the present invention;

[0054] Figure 3 This is a schematic diagram of the edge collaboration gateway in an intelligent safety helmet management system provided in an embodiment of the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] In this application, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0057] This invention can be used in a wide variety of general-purpose or special-purpose computing environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor devices, distributed computing environments including any of the above devices, etc.

[0058] This invention provides an intelligent safety helmet management system, such as... Figure 1 As shown, it specifically includes:

[0059] Multiple smart terminals, an edge collaboration gateway, and a remote monitoring center, wherein each of the smart terminals is mounted on a safety helmet;

[0060] For each of the aforementioned smart terminals, the smart terminal is used to collect local environmental data in real time and send communication information to the edge collaboration gateway; upon obtaining the data processing strategy fed back by the edge collaboration gateway, the target data is sent to the data receiving end based on the data processing strategy and the communication information.

[0061] The communication information includes terminal information, channel status information, and security risk score of the smart terminal. The terminal information includes the battery life and data timeliness level of the smart terminal. The data processing strategy includes a data collaboration list. The data collaboration list includes the communication information, data upload time slots, and terminal collaboration instructions of each smart terminal belonging to the same collaboration cluster in the smart helmet management system. The data receiving end is the edge collaboration gateway or a smart terminal belonging to at least one of the same target collaboration cluster. The target data includes at least the local environment data. The target collaboration cluster is a terminal set containing multiple smart terminals, including the smart terminal mentioned above.

[0062] The edge collaboration gateway is used to, when receiving communication information from each of the smart terminals in the smart helmet management system, determine the priority of each smart terminal based on the communication information, and set the data upload time slot for each smart terminal based on the priority; generate one or more collaboration clusters and a data collaboration list corresponding to each collaboration cluster based on the communication information; generate a data processing strategy corresponding to the data collaboration list, and send the data processing strategy to each smart terminal in the collaboration cluster corresponding to the data collaboration list; when receiving target data uploaded by any smart terminal, update the global security situation diagram corresponding to each smart terminal, and upload the global security situation diagram to the remote monitoring center;

[0063] The remote monitoring center is used for monitoring and issuing related instructions based on the global security situation diagram.

[0064] In this invention, each smart terminal is mounted as a lightweight device on a smart safety helmet, serving as a front-end device for data collection and simple data processing and calculation. The smart terminal includes a multimodal sensing unit, a wireless transmission unit, and a local processing unit (such as...). Figure 2 (As shown in the image) The multimodal sensing unit consists of multiple data acquisition devices (e.g., cameras, temperature sensors, gas sensors, and positioning modules). Each data acquisition device performs different data acquisition tasks to obtain local environmental data about the smart terminal's current environment and location. Local environmental data may include images, videos, audio, current location, current ambient gas information, and the user's current status information, such as the user wearing a smart safety helmet. After data collection by the multimodal sensing unit, the data is transmitted to the local processing unit, where it undergoes simple processing and calculations.

[0065] The multimodal perception module can also preprocess the collected raw data (i.e., local environment data), such as performing timestamp alignment, noise filtering, and format standardization. After preprocessing, the local processing unit scores the security risk of the smart terminal's current environment based on the local environment data. Simultaneously, the local processing unit also monitors terminal information and channel status information in real time. Based on the terminal information, the unit can determine the smart terminal's battery life and data timeliness level; based on the channel status information, it can determine the channel quality between the smart terminal and the edge collaborative gateway. The local processing unit uses the terminal information, channel status information, and security risk assessment as communication information and sends them to the edge collaborative gateway via the wireless transmission unit. The wireless transmission unit can communicate with the edge collaborative gateway and other terminals within the target collaborative cluster to which the smart terminal belongs. Specifically, the local processing unit sends a channel detection command to the next wireless transmission unit according to a time period to detect the channel quality between the smart terminal and the edge collaborative gateway. The wireless transmission unit periodically measures the physical layer parameters of the communication link with the edge collaborative gateway according to the instructions, including the signal-to-noise ratio, signal strength and the success rate of the most recent data transmission, and calculates a channel quality index that characterizes the quality of the current transmission conditions, and sends the index to the local processing unit.

[0066] Optionally, the smart terminal may also include an alarm unit, which can autonomously issue a location alarm message when the security risk score is high, and can also be used to receive monitoring instructions issued by upstream devices (e.g., edge collaborative gateways or remote monitoring centers).

[0067] Optionally, the local processing unit can be configured with a lightweight neural network model to perform real-time fusion analysis on the preprocessed data, identify preset security risk events, and calculate and generate a quantitative comprehensive security risk score based on the type and severity of the events.

[0068] When communication information needs to be sent to the edge collaborative gateway, each terminal, through contention for access or a preset initial time slot, encapsulates its communication information into a lightweight status report message via a wireless transmission unit and uploads it to the edge collaborative gateway. Upon receiving the communication information, the edge collaborative gateway calculates the priority of each smart terminal. The formula for calculating the smart terminal priority P_i is: P_i = Σ(w_k * F_k), where F_k includes multi-dimensional competition factors including security risk score R_i, channel quality index C_i, terminal battery life B_i, and data timeliness level T_i; w_k is the dynamic weight coefficient corresponding to each factor, which is dynamically adjusted and optimized by an online reinforcement learning model based on real-time network load, historical scheduling effects, and task type (alarm / normal). For example, the more congested the network, the higher the weight of the security risk score R_i; the lower the smart terminal's battery life, the higher its corresponding weight can be appropriately increased to avoid connection loss due to terminal battery life delays. After calculating the priority, the edge collaboration gateway sets the time slot for the terminal to send data. This time slot can be one or more, and multiple collaboration clusters can be set at the same time to generate a data collaboration list corresponding to each collaboration cluster.

[0069] When a smart terminal receives a data processing strategy containing a data collaboration list, it sends target data carrying local environment data according to the strategy. The data collaboration list includes communication information, data upload time slots, and terminal collaboration instructions for each smart terminal belonging to the same collaboration cluster within the smart helmet management system. Upon receiving the data processing strategy, the local processing unit determines the timing for uploading local environment data to the edge collaboration gateway based on the data upload time slot corresponding to the smart terminal in the strategy. Simultaneously, it determines whether it can act as a collaboration terminal. A collaboration terminal refers to a terminal that can act as middleware to forward data to other smart terminals. Specifically, the smart terminal's local processing unit determines its ability to act as a collaboration terminal based on the current security risk score and the data collaboration list. Specifically, it checks whether the security risk score is lower than a first risk threshold N1 and whether the data collaboration list contains a collaboration identifier associated with the smart helmet management system. This collaboration identifier is the identifier of a candidate collaboration terminal marked by the edge collaboration gateway. If the security risk score is not lower than the first risk threshold, or the data collaboration list does not contain a collaboration identifier, the smart terminal is determined to be a non-collaboration terminal. If the security risk score is lower than the first risk threshold and the data collaboration list contains a collaboration identifier, the smart terminal is determined to be a collaboration terminal.

[0070] When the smart terminal is used as a non-cooperative terminal, if the security risk score does not exceed the second risk threshold N2 (N2≥N1), it indicates that the environment in which the smart terminal is located is temporarily secure, and it can upload its local environment data according to the data upload time slot given by the edge collaborative gateway. If the security risk score exceeds the second risk threshold, it indicates that the environment in which the smart terminal is currently located is not secure. In the same target collaborative cluster, other collaborative terminals superior to the smart terminal are selected as data receivers, and the data from the smart terminal is forwarded through the data receivers.

[0071] The process of determining the data receiving end is as follows: determine whether the security risk score exceeds the second risk threshold, which is greater than or equal to the first risk threshold; if the security risk score does not exceed the second risk threshold, set the edge collaborative gateway as the data receiving end; if the security risk assessment exceeds the second risk threshold, obtain the communication information and terminal information of each collaborative terminal in the terminal collaborative list of the target collaborative cluster; based on the communication information and terminal information, select the optimal collaborative terminal as the data receiving end.

[0072] Among these, the optimal cooperating terminal is the one with the best channel quality among multiple cooperating terminals belonging to the target cooperating cluster and whose battery life exceeds the target battery life. That is, based on the battery life in the terminal information, cooperating terminals in the cooperating cluster whose battery life exceeds the target battery life are selected, and based on the channel quality in the communication information, the cooperating terminal with the highest channel quality (e.g., the strongest signal strength) among the selected cooperating terminals is chosen as the optimal cooperating terminal.

[0073] It should be noted that smart terminals whose battery life does not exceed the target battery life have relatively low battery capacity. To ensure that the smart terminal can successfully upload data in its corresponding data upload slot, it is necessary to reduce the processing of additional data by the smart terminal. Therefore, in the process of selecting the optimal collaborating terminal, it is necessary to avoid smart terminals with low battery life to prevent them from processing additional data sent by other smart terminals, which would lead to insufficient battery life to support data upload in the data upload slot.

[0074] The intelligent safety helmet management system implements an adaptive cooperative routing strategy based on "safety potential energy". It is only allowed to initiate a cooperative request when the intelligent terminal determines that its own channel quality is lower than the threshold and it is in a safe state (the safety risk score is higher than the second risk threshold). Data from non-cooperative terminals is sent to cooperative terminals that meet the safety potential energy requirements through inter-device links.

[0075] In this invention, if the security risk score of a non-cooperative terminal exceeds the second risk threshold and the channel quality is low, a cooperative terminal in the cooperative cluster with a higher channel quality than the non-cooperative terminal is selected as the data receiver. Priority is given to cooperative terminals with high channel quality and strong battery life as the optimal cooperative terminal. If the security risk score of a non-cooperative terminal does not exceed the second risk threshold, an edge cooperative gateway is selected as the data receiver.

[0076] Optionally, if the security risk score of a non-cooperative terminal is higher than the second risk threshold, but its channel quality is higher than that of other cooperative terminals, then the edge cooperative gateway is selected as the data receiving end.

[0077] For example, the smart terminal calculates a comprehensive safety risk score R every second. For instance, detecting excessive methane concentration is recorded as 0.8, detecting a person falling is recorded as 0.9, and a high-temperature alarm is recorded as 0.7. The final R is the maximum value among all currently active risks. Simultaneously, the terminal monitors channel quality C (calculated based on the success rate of recent uplink data packets and received signal strength), remaining battery power B (percentage), and the timeliness T of data to be transmitted (e.g., real-time alarm data T=1.0, historical log data T=0.3).

[0078] Understandably, when a smart terminal has a high security risk score, it has the capability to initiate a high-priority alarm. However, if its own channel quality is poor, it needs other terminals with better channel quality to forward and upload the data. When forwarding data between terminals, multi-hop relay paths based on security utility are supported. When planning paths, edge collaborative gateways or terminals aim to maximize the overall network security utility, dynamically avoiding high-risk nodes as relay hops, and supporting semantic layer feature fusion and aggregation encoding of multi-source collaborative data by collaborative terminals, rather than simply forwarding data packets.

[0079] For non-cooperative terminals, when sending data, they only need to process their own local environment data and generate target data. This can be done by encoding the local environment data. After generating the target data, if the edge collaborative gateway is the data receiving end, it sends the target data to the edge collaborative gateway via the wireless transmission unit when the data upload slot corresponding to the smart terminal arrives; if the optimal collaborative terminal is the data receiving end, it sends the real-time generated target data to the optimal collaborative terminal via the wireless transmission unit. That is, when the smart terminal is determined to be a non-cooperative terminal, it will not receive data forwarded by other terminals. Therefore, the smart terminal encodes its own real-time collected local environment data and determines the data receiving end based on the security risk score obtained from the local environment data. The security risk score is compared with a second risk threshold. If the security risk score does not exceed the second risk threshold, it indicates that the current environment of the smart terminal is not a high-risk environment, and the target data can be uploaded according to the data upload time slot specified by the edge collaboration gateway. If the security risk score exceeds the second risk threshold, it indicates that the current environment of the smart terminal is a high-risk environment. In this case, the local environment data needs to be uploaded to the edge collaboration gateway in a short period of time. However, since the terminal needs to upload data according to the allocated time slot, the smart terminal can find a high-priority collaboration terminal with strong battery life through the collaboration list to forward the data. The collaboration terminal then uploads the data to the edge collaboration gateway in advance to ensure that the smart terminal can upload data in a timely manner in a high-risk environment.

[0080] When forwarding target data from a smart terminal to a collaborating terminal, devices designated as blind spots can send target data to the collaborating terminal via Wi-Fi Direct. If a smart terminal is selected as a collaborating terminal, upon receiving target environmental data (i.e., local environmental data from other smart terminals) from neighboring smart terminals or other smart terminals, it performs semantic-level feature aggregation instead of simple IP packet forwarding. For example, if a collaborating terminal receives "fire warning" features from two adjacent smart terminals and also detects high temperature, its NPU (Neural Processing Unit) will perform spatiotemporal fusion of these three sets of features to generate a comprehensive "area fire alarm" data packet, which will then be uploaded using its allocated high-quality time slot. This mechanism significantly reduces bandwidth consumption in multi-hop networks while ensuring that critical information is not lost, avoiding broadcast storms caused by traditional multi-hop networks.

[0081] Specifically, when the smart terminal is a collaborative terminal, it may not receive local environment data forwarded by other smart terminals before the data upload time slot. Therefore, when the data upload time slot arrives, the target data sent by the collaborative terminal to the edge collaborative gateway only carries its own local environment data. If the collaborative terminal receives target environment data sent by other smart terminals in the target collaborative cluster before the data upload time slot, it fuses the target environment data with the local environment data to generate target data. This target environment data is the local environment data of other smart terminals in the target collaborative cluster that are sent to the collaborative terminal. When the data upload time slot corresponding to the smart terminal arrives, the target data is sent to the edge collaborative gateway through the wireless transmission unit. This data fusion can be time-series data fusion. Before data fusion, the smart terminal first performs a corresponding load detection. If the load detection result indicates that the smart terminal's load is not full, it fuses the target environment data with the local environment data to generate target data, and sends the target data to the edge collaborative gateway through the wireless transmission unit when the data upload time slot corresponding to the smart terminal arrives.

[0082] Optionally, when the load detection result indicates that the smart terminal is fully loaded, the target environment data is forwarded to other collaborative terminals within the target collaborative cluster, and the other collaborative terminals complete the corresponding data fusion and data upload process, thus avoiding data upload failure due to overload of the collaborative terminals.

[0083] For collaborative terminals, when receiving data from other smart terminals, they need to check their own communication quality and determine whether they have received a collaborative instruction. If the communication quality is too low or no collaborative instruction is received, the data can be forwarded to other collaborative terminals within the same collaborative cluster. This collaborative instruction is a communication token for data interaction between terminals. When a smart terminal needs to access another smart terminal, it must apply this collaborative instruction to obtain access and communication permissions.

[0084] In this invention, the edge collaboration gateway is set within the global area corresponding to each smart terminal, and includes: a resource scheduling unit, a security situation assessment unit, an uplink interface unit, and a prediction and optimization unit (e.g., ...). Figure 3 (As shown).

[0085] The resource scheduling unit communicates with each intelligent terminal; the uplink interface unit communicates with the remote monitoring center.

[0086] After receiving the communication information, the resource scheduling unit processes it and sends it to the prediction and optimization unit for weight updates.

[0087] For each smart terminal, the prediction and optimization unit updates weights as follows: It acquires the terminal's current and historical communication information; based on this information, it updates the weights corresponding to the terminal information, channel state information, and security risk score, respectively, to obtain the current weight information for the smart terminal. The prediction and optimization unit includes a poor attention module, which inputs the historical security risk score sequence and channel state sequence in parallel into the module. The cross-attention module learns the implicit correlation between the two. For example, when the model detects a "dust surge" feature in the security risk score sequence, it automatically adjusts the prediction weights for the channel state, tending to predict deep fading, even if the current channel performance has not yet deteriorated. Based on this strongly coupled prediction, the gateway can predict communication interruption risks 300-500ms earlier than traditional algorithms and trigger cooperative path establishment commands or adjust video coding rates accordingly, truly achieving proactive, preventative scheduling. Therefore, when it is necessary to update the weights of security risk scores, channel state information, and terminal information, it is only necessary to input the current security risk scores, channel state information, and terminal information, as well as the historical security risk scores, channel state information, and terminal information, into the module. The module will then capture the changing trends of terminal information, channel state information, and security risk scores based on the relationships between the sequences, and dynamically adjust the weights corresponding to the terminal information, channel state information, and security risk scores of the smart terminal based on these changing trends.

[0088] It should be noted that the edge collaborative gateway collects the actual scheduling results and prediction deviations within a time period according to the time cycle, and uses this information to update the module weights of the cross-attention module online, thereby optimizing the cross-attention module and improving its accuracy.

[0089] For the resource scheduling unit, based on the communication information and current weight information of each intelligent terminal, it calculates the priority of each intelligent terminal and sets the data upload time slot for each intelligent terminal according to the priority; it divides the intelligent terminals into multiple collaborative clusters according to their locations; based on the communication information of each intelligent terminal, it marks the candidate collaborative terminals in each collaborative cluster and generates a collaborative instruction corresponding to each intelligent terminal; it generates a data collaborative list for each collaborative cluster and sends a data processing strategy carrying the data collaborative list to each intelligent terminal. The data collaborative list includes at least the priority, data upload time slot, communication information, and collaborative instruction of each intelligent terminal in the same collaborative cluster.

[0090] The formula for calculating priority by the resource scheduling unit is as follows:

[0091] P_i=Σ(w_k*F_k)

[0092] The expansion of the formula for this priority is:

[0093] P_i=w_R*R_i + w_C*C_i+w_B*B_i+w_T*T_i

[0094] F_k includes multi-dimensional competitive factors such as security risk score R_i, channel quality index C_i, terminal battery life B_i, and data timeliness level T_i; w_k is the dynamic weight coefficient corresponding to each factor, including the weight w_R of security risk score, the weight w_C of channel quality index, the weight w_B of terminal battery life, and the weight w_T of data timeliness level. Each weight is obtained by processing the cross-attention module in the prediction and optimization unit.

[0095] Optionally, after the cross-attention module outputs the aforementioned current weight information w_k, the current weight information of each smart terminal can be updated based on the overall network load of the edge collaborative gateway. The weight coefficient w_k is dynamically adjusted according to the overall network load. For example, when the network is idle, the w_k value is reduced, so that resource allocation is more inclined towards terminals with better channel quality to improve the total throughput; when the network is congested, the w_k value is increased, so that resource allocation is more inclined towards terminals with high security risks to ensure critical alarms.

[0096] In one embodiment, the edge collaboration gateway collects a report from all smart terminals every second, consisting of (security risk score R + channel quality index C + terminal battery life B + data timeliness level T). The current network load L is calculated. The cross-attention module dynamically outputs the current optimal weight combination (w_R, w_C, w_B, w_T) based on the current load L and the historical scheduling effects of each factor. For example, when the network is idle (L=0.3), the weights are biased towards channel and throughput (higher w_C); when the network is congested (L=0.9), the weights are extremely biased towards risk (extremely high w_R); simultaneously, for terminals with less than 20% battery, their w_B will be appropriately increased. Priorities are calculated according to the formula P_i=w_R*R_i+w_C*C_i+w_B*B_i+w_T*T_i, and time slot resources for the next cycle are allocated according to priority from high to low.

[0097] After calculating the priority of each smart terminal and allocating time slots, the resource scheduling unit can group adjacent smart terminals into the same cooperative cluster based on their location. Based on the communication information of each terminal within the same cooperative cluster, candidate cooperative terminals that can serve as cooperative terminals are pre-designated. Specifically, smart terminals with security risk scores below the third risk threshold are identified as first candidate terminals. Based on the terminal information and channel state information of each first candidate terminal, the endurance and channel quality of each first candidate terminal are determined. First candidate terminals with endurance exceeding the target endurance and excellent channel quality are marked as candidate cooperative terminals, and a cooperative identifier corresponding to each candidate cooperative terminal is generated. The third risk threshold is less than the second risk threshold, and the cooperative identifier is recorded in the data cooperative list. When a smart terminal receives the data cooperative list, it can determine whether it is a cooperative terminal based on the cooperative identifier and its own security risk score.

[0098] It should be noted that, when the communication information of the smart terminal indicates that the signal strength of the smart terminal reaches the target signal threshold, the smart terminal is determined to be a smart terminal with excellent channel quality. The target signal threshold can be determined based on the chip parameters of the chip to which the wireless transmission unit in the smart terminal belongs.

[0099] In addition, the edge collaboration gateway sets a corresponding collaboration instruction for each smart terminal with a collaboration identifier. This instruction is equivalent to terminal access permission. Smart terminals within the same collaboration cluster can only send data to the corresponding collaboration terminal through the collaboration instruction.

[0100] The security situation assessment unit obtains target data from each smart terminal from the resource scheduling unit, and parses each target data to obtain local environment data of each smart terminal in the smart safety helmet management system. Based on each local environment data, it generates a global security situation diagram corresponding to each smart terminal, and sends the global security situation diagram and the local environment data of each smart terminal to the remote monitoring center through the uplink interface unit.

[0101] It should be noted that the safety situation assessment unit updates the global safety situation map in real time according to each time period. In addition, based on the global safety situation map for each time period, the safety situation assessment unit generates safety situation change curves for each area block within the global area corresponding to the smart helmet management system. The global safety situation assessment unit uploads the latest global safety situation map and change curves, along with local environmental data from each smart terminal, to the remote monitoring center. Based on the global safety situation map and the local environmental data of each smart terminal, the remote monitoring center identifies whether there are high-risk areas within the global area monitored by the smart helmet management system. If a high-risk area exists, an alarm message corresponding to the high-risk area is issued, and an alarm command is sent to each smart terminal within the high-risk area through the edge collaboration gateway to trigger self-location alarms from each smart terminal within the high-risk area.

[0102] It should be noted that the remote monitoring center can also predict high-risk areas based on the security situation change curves of each area block, and formulate corresponding high-risk prevention strategies accordingly, and then distribute and report these strategies.

[0103] The intelligent safety helmet management system provided in this invention adopts a hybrid architecture of "centralized edge scheduling + distributed terminal collaboration". The edge collaboration gateway, as the control core, collects security risk scores, channel status information, and other dynamic terminal information from all intelligent terminals, and generates corresponding data processing strategies through a multi-dimensional dynamic priority algorithm. Simultaneously, the system supports an intelligent inter-terminal collaborative forwarding mechanism, allowing and encouraging data collaboration and multi-hop forwarding between adjacent terminals via inter-device communication links (such as Wi-Fi Direct), enabling intelligent terminals located at the network edge or with poor channel conditions to utilize the superior channels of their neighbors to report critical information. Furthermore, the system introduces a cross-attention module to predict short-term risks and channel conditions for proactive resource scheduling.

[0104] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for system or system embodiments, since they are basically similar to method embodiments, the description is relatively simple, and relevant parts can be referred to the descriptions in the method embodiments. The systems and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0105] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both.

[0106] To clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described in terms of functionality above. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0107] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent safety helmet management system, characterized in that, The intelligent safety helmet management system includes: Multiple smart terminals, an edge collaboration gateway, and a remote monitoring center, wherein each of the smart terminals is mounted on a safety helmet; For each of the aforementioned smart terminals, the smart terminal is used to collect local environmental data in real time and send communication information to the edge collaboration gateway; upon obtaining the data processing strategy fed back by the edge collaboration gateway, the target data is sent to the data receiving end based on the data processing strategy and the communication information. The communication information includes terminal information, channel status information, and security risk score of the smart terminal. The terminal information includes the battery life and data timeliness level of the smart terminal. The data processing strategy includes a data collaboration list. The data collaboration list includes the communication information, data upload time slots, and terminal collaboration instructions of each smart terminal belonging to the same collaboration cluster in the smart helmet management system. The data receiving end is the edge collaboration gateway or a smart terminal belonging to at least one of the same target collaboration cluster. The target data includes at least the local environment data. The target collaboration cluster is a terminal set containing multiple smart terminals, including the smart terminal mentioned above. The intelligent terminal includes: a multimodal sensing unit, a wireless transmission unit, and a local processing unit; The multimodal sensing unit is used to collect local environmental data in real time; The wireless transmission unit is used to communicate with the edge collaboration gateway and each terminal within the target collaboration cluster; The local processing unit is used to detect the terminal information and channel status information of the smart terminal in real time, and to perform a security risk score on the current environment of the smart terminal based on the local environment data to obtain communication information; when receiving a data processing strategy from the wireless transmission unit, it determines whether the smart terminal is a cooperative terminal based on the security risk score; if it is a non-cooperative terminal, it sets the edge cooperative gateway or a cooperative terminal in the target cooperative cluster as the data receiving end based on the security risk score and the data cooperative list in the data processing strategy; if it is a cooperative terminal, it sets the edge cooperative gateway as the data receiving end; it generates target data carrying the local environment data, and sends the target data to the data receiving end through the wireless transmission unit; When the smart terminal is a non-cooperative terminal, the local processing unit is specifically used to: if the security risk score indicates that the smart terminal is in a high-risk environment, then based on the security risk score and the data cooperation list in the data processing strategy, select the optimal cooperative terminal from the target cooperation cluster as the data receiving end, wherein the optimal cooperative terminal has a higher priority than the smart terminal; if the security risk score indicates that the smart terminal is not in a high-risk environment, then select the edge cooperation gateway as the data receiving end; The edge collaboration gateway includes: a resource scheduling unit, a security situation assessment unit, an uplink interface unit, and a prediction and optimization unit; The resource scheduling unit communicates with each of the intelligent terminals; the uplink interface unit communicates with the remote monitoring center. The prediction and optimization unit is used to obtain the communication information of each of the smart terminals and the historical communication information of each smart terminal received by the resource scheduling unit; based on the current communication information and the historical communication information, it updates the weights corresponding to the terminal information, channel state information and security risk score in the communication information of each smart terminal, so as to obtain the current weight information corresponding to each smart terminal. The resource scheduling unit is configured to calculate the priority of each smart terminal based on its communication information and current weight information, and set the data upload time slot for each smart terminal according to its priority; divide the smart terminals into multiple collaborative clusters according to their locations; mark the candidate collaborative terminals in each collaborative cluster according to the communication information of each smart terminal, and generate a collaborative instruction corresponding to each smart terminal; generate a data collaborative list for each collaborative cluster, and send a data processing strategy carrying the data collaborative list to each smart terminal, wherein the data collaborative list includes at least the priority, data upload time slot, communication information, and collaborative instruction of each smart terminal in the same collaborative cluster; The security situation assessment unit is used to obtain target data of each of the smart terminals from the resource scheduling unit, and parse each of the target data to obtain local environment data of each smart terminal in the smart safety helmet management system. Based on each of the local environment data, a global security situation diagram corresponding to each of the smart terminals is generated, and the global security situation diagram and the local environment data of each of the smart terminals are sent to the remote monitoring center through the uplink interface unit. The remote monitoring center is used for monitoring and issuing related instructions based on the global security situation diagram.

2. The intelligent safety helmet management system according to claim 1, characterized in that, The local processing unit determines whether the smart terminal is a collaborative terminal based on the security risk score, specifically for: Determine whether the safety risk score is lower than a first risk threshold, and whether the data collaboration list contains a collaboration identifier associated with the smart helmet management system; the collaboration identifier is the identifier of the candidate collaboration terminal marked by the edge collaboration gateway; If the safety risk score is not lower than the first risk threshold, or if the data collaboration list does not contain a collaboration identifier associated with the smart helmet management system, the smart terminal is determined to be a non-collaboration terminal. If the safety risk score is lower than the first risk threshold, and the data collaboration list contains a collaboration identifier associated with the smart helmet management system, the smart terminal is determined to be a collaboration terminal.

3. The intelligent safety helmet management system according to claim 2, characterized in that, When the smart terminal is a non-cooperative terminal, the local processing unit is specifically used for: The local environment data is encoded to generate target data, and it is determined whether the security risk score exceeds a second risk threshold, wherein the second risk threshold is greater than or equal to the first risk threshold. If the security risk score does not exceed the second risk threshold, the edge collaboration gateway is set as the data receiving end, and when the data upload time slot corresponding to the smart terminal is reached, the target data is sent to the edge collaboration gateway through the wireless transmission unit. If the security risk assessment exceeds the second risk threshold, the communication information and terminal information of each cooperative terminal in the terminal cooperative list of the target cooperative cluster are obtained; based on the communication information and terminal information, the optimal cooperative terminal is selected as the data receiving end, and the target data generated in real time is sent to the optimal cooperative terminal through the wireless transmission unit. The optimal cooperative terminal is the cooperative terminal with the best channel quality among multiple cooperative terminals that belong to the target cooperative cluster and have a battery life exceeding the target battery life.

4. The intelligent safety helmet management system according to claim 2, characterized in that, When the smart terminal is a collaborative terminal, the local processing unit is specifically used for: Upon receiving target environment data sent by other smart terminals in the target collaborative cluster, load detection is performed on the smart terminal, where the target environment data is the local environment data of the other smart terminals; When the load detection result indicates that the smart terminal is not fully loaded, the target environment data and the local environment data are fused to generate target data; When the data upload time slot corresponding to the smart terminal is reached, the target data is sent to the edge collaboration gateway through the wireless transmission unit.

5. The intelligent safety helmet management system according to claim 1, characterized in that, The resource scheduling unit marks candidate collaborative terminals in each collaborative cluster based on the communication information of each intelligent terminal, specifically for: Smart terminals with security risk scores below the third risk threshold were identified as the first candidate terminals. Based on the terminal information and channel status information of each first candidate terminal, the battery life and channel quality of each first candidate terminal are determined. The first candidate terminal whose battery life exceeds the target battery life and whose channel quality is excellent is marked as a candidate cooperative terminal, and a cooperative identifier corresponding to each candidate cooperative terminal is generated.

6. The intelligent safety helmet management system according to claim 1, characterized in that, The prediction and optimization unit includes: a cross-attention module; The cross-attention module is used to identify the correlation between the current communication information and historical communication information of the same smart terminal, and dynamically capture the changing trends of the terminal information, channel state information and security risk score of the smart terminal according to the time series, and dynamically adjust the weights corresponding to the terminal information, channel state information and security risk score of the smart terminal according to the changing trends.

7. The intelligent safety helmet management system according to claim 6, characterized in that, The prediction and optimization unit is further configured to: Determine the overall network load of the edge collaboration gateway; Based on the overall network load, update the current weight information of each smart terminal.

8. The intelligent safety helmet management system according to claim 1, characterized in that, The remote monitoring center is specifically used for: Based on the global security situation diagram and the local environment data of each smart terminal, the system identifies whether there are high-risk areas in the global area monitored by the smart helmet management system. If a high-risk area exists, an alarm message corresponding to the high-risk area is issued, and an alarm command is sent to each smart terminal in the high-risk area through the edge collaboration gateway to trigger each smart terminal in the high-risk area to perform self-location alarm.

Images