Intelligent rescue indication method and system based on intelligent protective helmet

Through real-time monitoring and data analysis of the sensor set of intelligent protective helmets, the problem that rescuers cannot obtain key data on the site in real time is solved, and risk assessment and path planning of disaster sites are realized, and rescue efficiency and safety are improved.

CN118780963BActive Publication Date: 2024-11-15JIANGSU KETIANSHI INTELLIGENT TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411253707.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-09
Publication Date
2024-11-15
Estimated Expiration
2044-09-09

AI Technical Summary

Technical Problem

Rescue personnel find it difficult to obtain and share key data in real time at the disaster site, resulting in inefficient rescue efforts and the inability to effectively formulate and implement rescue strategies.

Method used

It adopts intelligent protective helmet equipped with a sensor group to monitor environmental data in real time, and provides real-time risk assessment and collaborative rescue instructions through data analysis and path planning modules.

Benefits of technology

Through the data collection and sharing of intelligent protective helmets, environmental perception, risk assessment, strategy formulation and path planning of the rescue site are realized, and the efficiency and safety of rescue operations are improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118780963B_ABST
    Figure CN118780963B_ABST
Patent Text Reader

Abstract

The present invention discloses an intelligent rescue indication method and system based on an intelligent protective helmet, which belongs to the field of intelligent helmets, wherein the method includes: obtaining a target intelligent protective helmet, obtaining a target rescue scene environment data stream; performing abnormal data identification, obtaining a rescue abnormal data stream, performing risk feature assessment, obtaining rescue risk feature information, matching with the rescue strategy plan space, obtaining a target rescue strategy plan, and obtaining GPS positioning information at the same time; obtaining multiple rescue environment data streams through multiple intelligent protective helmets, performing path optimization planning, obtaining target rescue path planning information, and performing collaborative rescue instructions. This application solves the problem in the prior art that rescue personnel cannot obtain and share key rescue scene data in real time, resulting in low rescue efficiency, and achieves the technical effect of collecting and sharing rescue scene data in real time through intelligent protective helmets, performing data analysis and rescue instructions, and improving rescue efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of intelligent helmets, and in particular to an intelligent rescue indication method and system based on an intelligent protective helmet. Background Art

[0002] In the rescue process of various disaster accidents, rescuers need to go deep into dangerous rescue sites to search and rescue. However, due to the complex and changeable rescue environment, there are many unpredictable risk factors, and the safety of rescuers themselves is also threatened. At the same time, key data at the rescue site, such as environmental conditions, rescue risks, etc., are crucial for formulating effective rescue strategies and path planning. In the prior art, rescuers usually use traditional protective equipment, such as ordinary helmets, walkie-talkies, etc. These equipment cannot collect and transmit key data at the rescue site in real time, resulting in difficulty for rescuers to understand the on-site situation in a timely and comprehensive manner, and unable to provide effective decision-making support and guidance for rescuers; there is also a lack of real-time information sharing and coordination means between rescuers, making it difficult to achieve efficient collaboration. These problems seriously restrict the efficiency of rescue and slow down the rescue of trapped people. Summary of the invention

[0003] This application provides an intelligent rescue indication method and system based on an intelligent protective helmet, aiming to solve the problem in the prior art that rescue personnel are unable to obtain and share key data on the rescue scene in real time, resulting in low rescue efficiency.

[0004] In view of the above problems, the present application provides an intelligent rescue instruction method and system based on an intelligent protective helmet.

[0005] The first aspect disclosed in the present application provides an intelligent rescue indication method based on an intelligent protective helmet, the method comprising: obtaining a target intelligent protective helmet, the target intelligent protective helmet being equipped with a sensor group, the sensor group comprising a temperature and humidity sensor, a gas concentration sensor, a noise sensor and a GPS locator, and obtaining a target rescue scene environment data stream through monitoring by the sensor group; performing abnormal data identification on the target rescue scene environment data stream to obtain a rescue abnormal data stream, performing risk characteristic assessment based on the rescue abnormal data stream to obtain rescue risk characteristic information; constructing a rescue strategy solution space, matching the rescue risk characteristic information with the rescue strategy solution space to obtain a target rescue strategy solution, and obtaining GPS positioning information according to the rescue scene environment data stream; obtaining multiple rescue environment data streams through multiple intelligent protective helmets, performing path optimization planning on multiple rescue environment data streams, target rescue scene environment data streams and GPS positioning information, obtaining target rescue path planning information, and performing collaborative rescue indication based on the target rescue strategy solution and the target rescue path planning information.

[0006] Another aspect disclosed in the present application provides an intelligent rescue indication system based on an intelligent protective helmet, the system comprising: an environmental data acquisition module, used to acquire a target intelligent protective helmet, the target intelligent protective helmet is equipped with a sensor group, the sensor group comprises a temperature and humidity sensor, a gas concentration sensor, a noise sensor and a GPS locator, and the target rescue scene environmental data stream is acquired through monitoring by the sensor group; a risk feature acquisition module, used to identify abnormal data of the target rescue scene environmental data stream, obtain a rescue abnormal data stream, perform risk feature assessment based on the rescue abnormal data stream, and obtain rescue risk feature information; a solution matching module, used to construct a rescue strategy solution space, match the rescue risk feature information with the rescue strategy solution space, obtain a target rescue strategy solution, and obtain GPS positioning information according to the rescue scene environmental data stream; a collaborative rescue indication module, used to acquire multiple rescue environment data streams through multiple intelligent protective helmets, perform path optimization planning on multiple rescue environment data streams, target rescue scene environmental data streams and GPS positioning information, obtain target rescue path planning information, and perform collaborative rescue indication based on the target rescue strategy solution and the target rescue path planning information.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] By integrating multiple types of sensors on the target intelligent protective helmet, the environmental parameters of the rescue site are monitored and collected in real time, and the target rescue site environmental data stream reflecting the on-site conditions is obtained, providing a basis for subsequent data analysis and rescue instructions; the collected target rescue site environmental data stream is identified for anomalies, and abnormal data reflecting dangerous situations are extracted to obtain the rescue abnormal data stream, and the risk characteristics of the rescue site are evaluated based on this, and rescue risk characteristic information reflecting the risk level and type is obtained, providing a basis for the formulation of rescue strategies; a rescue strategy solution space is constructed, and the rescue risk characteristic information obtained based on the evaluation is matched with the rescue strategy solution space to obtain the target rescue strategy solution. The system obtains GPS positioning information based on the rescue scene environmental data stream to prepare for subsequent path planning; through the collaborative data collection of multiple intelligent protective helmets, the environmental information and GPS position collected by each helmet are integrated to perform global path optimization, and the target rescue path planning information is obtained. Then, the target rescue strategy and the target rescue path planning information are combined to provide collaborative rescue instructions and guide the rescue operation. This technical solution solves the problem in the prior art that rescue personnel cannot obtain and share key rescue scene data in real time, resulting in low rescue efficiency. It achieves the technical effect of real-time collection and sharing of rescue scene data through intelligent protective helmets, data analysis and rescue instructions, and improving rescue efficiency.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A flow chart of an intelligent rescue indication method based on an intelligent protective helmet is provided for an embodiment of the present application;

[0011] Figure 2 A structural schematic diagram of an intelligent rescue indication system based on an intelligent protective helmet is provided for an embodiment of the present application.

[0012] Explanation of the reference numerals: environmental data acquisition module 11, risk characteristic acquisition module 12, solution matching module 13, collaborative rescue instruction module 14. DETAILED DESCRIPTION

[0013] The overall idea of ​​the technical solution provided by this application is as follows:

[0014] The embodiments of the present application provide an intelligent rescue indication method and system based on an intelligent protective helmet.

[0015] First, obtain the target intelligent protective helmet equipped with temperature and humidity sensors, gas concentration sensors, noise sensors and GPS locators, and obtain the rescue site environment data stream through the sensor group monitoring. Perform abnormal data identification and risk feature assessment on the data stream to obtain rescue risk feature information. Then, construct the rescue strategy plan space, match the Jiyuan risk feature information with the rescue strategy plan space, obtain the target rescue strategy plan, and obtain GPS positioning information. Finally, obtain multiple rescue environment data streams through multiple intelligent protective helmets, perform path optimization planning for each data stream and GPS information, obtain the target rescue path planning information, and then provide collaborative rescue instructions based on the target rescue strategy plan and target path planning information.

[0016] To sum up, this application makes full use of the data collection, transmission and positioning functions of the smart protective helmet, and through data analysis and information sharing, realizes rescue site environment perception, risk assessment, strategy formulation, path planning and coordinated command, forms an intelligent rescue instruction plan, and provides rescue personnel with comprehensive, accurate and real-time rescue information and decision-making support, effectively improving the efficiency and safety of rescue operations.

[0017] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically described below in conjunction with the drawings in the specification.

[0018] Embodiment 1

[0019] like Figure 1 As shown, the embodiment of the present application provides an intelligent rescue indication method based on an intelligent protective helmet, the method comprising:

[0020] S1: Acquire a target smart protective helmet, wherein the target smart protective helmet is equipped with a sensor group, wherein the sensor group includes a temperature and humidity sensor, a gas concentration sensor, a noise sensor and a GPS locator, and the sensor group is used to monitor and acquire a target rescue scene environment data stream.

[0021] Specifically, the target smart protective helmet is an intelligent protective helmet used in rescue scenarios. Various types of sensors are integrated inside the smart protective helmet to form a sensor group for sensing the environmental parameters of the rescue site. The sensor group includes temperature and humidity sensors, gas concentration sensors, noise sensors and GPS locators. Among them, the temperature and humidity sensors are used to detect the temperature and humidity conditions at the rescue site; the gas concentration sensor is used to detect the air composition and harmful gas concentration at the rescue site; the noise sensor is used to detect the decibel number of sound at the rescue site and judge the noise level; the GPS locator is used to obtain the geographical coordinates of the smart protective helmet in real time. During the rescue process, the sensor group will continuously collect various environmental parameters at the rescue site to form a real-time updated data stream, namely the target rescue site environmental data stream, which provides a key data basis for subsequent data analysis and rescue instructions.

[0022] By using smart protective helmets to obtain key environmental information at the rescue site, we can grasp the temperature and humidity, harmful gas concentration, noise level and other parameters at the site in real time, providing data support for scientific rescue. At the same time, the GPS positioning function provides a location basis for planning rescue routes and coordinating the actions of multiple rescuers. The acquisition of this information lays a data foundation for the subsequent formulation of rescue plans and coordinated rescue instructions.

[0023] S2: performing abnormal data identification on the target rescue scene environment data stream to obtain a rescue abnormal data stream, performing risk feature assessment based on the rescue abnormal data stream to obtain rescue risk feature information.

[0024] Specifically, due to the complex and changeable environment of the rescue site, the data collected by the sensor can reflect the existence of certain abnormal conditions at the scene, such as sudden temperature rise, excessive concentration of harmful gases, excessive noise decibels, etc. These abnormal conditions indicate that there are potential dangers or threats at the rescue site and rescue is needed. Therefore, after obtaining the target rescue site environment data stream, the target rescue site environment data stream is firstly identified for abnormal data. For example, by setting abnormal judgment conditions, such as temperature threshold, gas concentration upper limit, noise decibel critical value, etc., the abnormal data points in the data stream are automatically identified. Once abnormal data is found, it is marked as rescue abnormal data to form a special data stream and obtain the rescue abnormal data stream. Secondly, the rescue abnormal data stream is used to evaluate the risk characteristics of the rescue site, judge the risk level and risk type of the current rescue site, and provide a basis for rescue decision-making. By analyzing the key parameters in the rescue abnormal data stream, such as the abnormal temperature, the degree of excess of harmful gas concentration, the size of the noise decibel, etc., and taking into account the duration of the abnormal situation, the trend of change and other factors, the comprehensive risk level of the rescue site is evaluated, such as high risk, medium risk, low risk, etc. At the same time, the corresponding risk types are listed, such as high temperature, toxic gas, strong noise, etc., to prompt rescuers to respond accordingly. Among them, risk feature assessment can be implemented in a variety of ways, such as experience-based risk judgment matrix, data-based risk assessment model, etc. Afterwards, the risk types and risk levels obtained from the risk feature assessment are summarized as rescue risk feature information.

[0025] Through abnormal data identification and risk feature assessment, real-time monitoring and assessment of rescue site risks are achieved, which helps to quickly discover abnormal conditions and potential risks at the rescue site, provide key basis for decision-making on rescue operations, improve the targetedness and effectiveness of rescue, maximize the safety of rescue personnel, and improve rescue efficiency.

[0026] S3: Construct a rescue strategy plan space, match the rescue risk feature information with the rescue strategy plan space, obtain a target rescue strategy plan, and obtain GPS positioning information based on the rescue scene environment data stream.

[0027] Specifically, first, a rescue strategy space is constructed, which is a collection of strategies and contains a variety of rescue strategies for different risk feature information. Among them, each strategy is pre-designed and optimized based on previous rescue experience and expert knowledge, covering all aspects of the rescue operation, such as personnel dispatch, equipment selection, action route, communication coordination, etc. The construction of the rescue strategy space adopts the method of knowledge engineering to systematically summarize, abstract and formalize the experience and knowledge of rescue experts to form a structured strategy library. Secondly, using the obtained rescue risk feature information, a matching search is performed in the rescue strategy space to find the optimal rescue strategy. Specifically, the rescue risk feature information is used as the retrieval condition to find the matching strategy in the rescue strategy space. The matching principle is to find the strategy that can most effectively deal with the current rescue risk feature information, such as matching through multiple dimensions, such as risk type, risk level, on-site environment, etc. After matching, the optimal strategy is selected from the rescue strategy space as the target rescue strategy. At the same time, the GPS positioning information of the rescue site environmental data stream is read, accurate to the specific longitude and latitude coordinates, reflecting the real-time location of the rescue personnel, providing basic data for subsequent rescue route planning and multi-party collaboration, and ensuring the time and space synchronization of the rescue operation.

[0028] By constructing a rescue strategy plan space and matching strategies according to rescue risk feature information, the target rescue strategy plan is automatically generated to provide rescuers with an actionable action guide. At the same time, the location information of the rescue site is obtained to lay the foundation for subsequent path planning.

[0029] S4: Acquire multiple rescue environment data streams through multiple smart protective helmets, perform path optimization planning on the multiple rescue environment data streams, the target rescue scene environment data stream and the GPS positioning information, obtain target rescue path planning information, and provide collaborative rescue instructions based on the target rescue strategy plan and the target rescue path planning information.

[0030] Specifically, in a rescue operation, there is not just one rescuer, but multiple rescuers working at the same time. Each rescuer is equipped with a smart protective helmet, and there are multiple smart protective helmets. The target rescue site environment data stream is the environmental data of the target area where the rescue operation needs to be carried out. The data stream collected by a specific rescuer using the smart protective helmet he wears reflects the real-time environmental conditions of the target rescue site. Multiple rescue environment data streams refer to the environmental data streams collected by other rescuers in their respective rescue areas. Therefore, in addition to the target rescue site environment data stream, multiple smart protective helmets are used to obtain multiple rescue environment data streams to reflect the rescue site conditions of other rescue areas, providing a more comprehensive and dynamic information basis for global rescue decisions. Secondly, multiple rescue environment data streams, target rescue site environment data streams, and GPS positioning information are summarized and integrated to form a comprehensive data view. On this basis, the optimal planning of the rescue path is carried out. The goal of path planning is to find one or more optimal paths so that the rescue operation can be completed at the fastest speed, shortest distance, and highest safety. Among them, path planning needs to consider multiple factors, such as the topography, road conditions, obstacle distribution, etc. of the rescue site, as well as the risk assessment results of different paths; at the same time, factors such as the travel speed of different rescuers and the performance of rescue equipment should be considered. Path planning can adopt a variety of optimization algorithms, such as graph search algorithm, heuristic search algorithm, dynamic programming algorithm, etc., and the optimal path plan can be searched by establishing a road network model and cost function. After optimal planning, the target rescue path planning information is obtained to guide rescuers to perform rescue in the best way and avoid potential risks and obstacles as much as possible. Finally, according to the target rescue strategy plan and the target rescue path planning information, collaborative rescue instructions are given to multiple rescuers, so that multiple rescuers can act and cooperate with each other to achieve the best rescue effect. The collaborative rescue instructions comprehensively consider the task division, action route, communication coordination and other aspects of different rescuers. Through the communication module of the smart protective helmet, the rescue strategy plan and rescue path planning information are sent to each rescuer in real time, and the tasks and actions of each rescuer are dynamically adjusted to adapt to the changes at the rescue site. At the same time, it supports information sharing and real-time communication between rescuers to promote the coordination and cooperation of rescue operations.

[0031] By aggregating data from multiple smart protective helmets, combined with environmental data and positioning information, the rescue path is optimized and planned to obtain the optimal rescue path plan. On this basis, according to the rescue strategy plan and path planning information, multiple rescue personnel are given coordinated rescue instructions to achieve optimal dispatch and joint action of rescue forces, maximize the role of rescue forces, and improve rescue efficiency.

[0032] Furthermore, the embodiment of the present application also includes:

[0033] A rescue scene database is acquired through big data technology, wherein the rescue scene database includes rescue risk data sets of various scene attribute categories; the rescue risk data sets of various scene attribute categories are respectively identified and trained using a deep learning network structure to obtain a rescue risk feature analysis model library; the target rescue scene is attributed and target scene attribute information is determined; based on the target scene attribute information, a target risk feature analysis model is called from the rescue risk feature analysis model library, and risk feature assessment is performed on the rescue abnormal data stream based on the target risk feature analysis model to output the rescue risk feature information.

[0034] In a preferred embodiment, in order to obtain rescue risk feature information, a rescue risk feature analysis model library based on different rescue scenarios is constructed, and the risk features of the target rescue site are quickly and accurately evaluated based on this to obtain rescue risk feature information. First, using big data acquisition and storage technology, massive data related to rescue site risk assessment is collected and summarized from multiple channels such as historical rescue cases, expert experience, sensor data, etc., and organized into a structured database as a rescue site database. The rescue site database is classified according to different rescue scene attributes (such as disaster type, geographical environment, meteorological conditions, etc.) to form several sub-datasets, which are rescue risk data sets for each scene attribute category, and each sub-dataset corresponds to a specific rescue scene. The sub-dataset contains historical data, monitoring data and statistical analysis data of various risk factors (such as fire spread, toxic gas diffusion, building collapse, etc.) in the scene, which comprehensively reflects the rescue risk characteristics in the scene. Then, a special deep learning network structure is designed for the rescue risk data set of each scene attribute category, such as a convolutional neural network, a long short-term memory network, etc., and the network is trained using the data set. During the training process, the network continuously adjusts its internal parameters to learn how to identify and extract key features related to rescue risks from raw data, such as high temperature, smoke, and dangerous gases. After training, each deep learning network becomes a specialized rescue risk feature analysis model that can quickly and accurately analyze risk features for specific rescue scenarios. All trained models are aggregated to form a rescue risk feature analysis model library that covers various rescue scenarios.

[0035] When facing a specific rescue mission, the target rescue site is first analyzed to determine the type of rescue scene it belongs to and to mark its attributes. Taking into account multiple characteristics of the target rescue site, such as disaster type (fire, earthquake, mudslide, etc.), geographical environment (plateau, jungle, city, etc.), meteorological conditions (high temperature, strong wind, heavy rain, etc.), building conditions (intact, damaged, collapsed, etc.), etc., the target scene attribute information of the target rescue site is obtained, that is, the scene attribute category of the site, such as "high-rise building fire", "mountain mudslide", etc. After determining the target scene attribute information, the corresponding target risk feature analysis model is called from the rescue risk feature analysis model library. Next, the obtained rescue abnormal data stream is input into the model, the data stream is analyzed and identified, and the rescue risk feature information is extracted, including the risk type of the target rescue site (such as fire, explosion, collapse, etc.), the danger level of each risk feature (such as high risk, medium risk, low risk), etc., to provide a basis for the subsequent rescue strategy formulation and rescue operation.

[0036] Through big data technology and deep learning methods, a rescue strategy clustering solution library is constructed to quickly and accurately identify and evaluate the risk characteristics of the target site for different rescue scenarios, provide support for the acquisition of rescue strategies, and improve the targetedness and effectiveness of rescue.

[0037] Furthermore, the embodiment of the present application also includes:

[0038] Perform rescue risk feature analysis on the rescue strategy solution space to obtain rescue risk type feature parameters and rescue risk level feature parameters; cluster and divide the rescue strategy solution space according to the rescue risk type feature parameters and the rescue risk level feature parameters to obtain a rescue strategy cluster solution library; perform similarity analysis on the rescue risk feature information and the rescue strategy cluster solution library to obtain a risk feature similarity set; screen and match the rescue strategy cluster solution library based on the risk feature similarity set to determine the target rescue strategy solution.

[0039] In a feasible implementation, the rescue strategy scheme space contains a set of multiple rescue strategy schemes, and each strategy scheme is designed for specific rescue risk characteristics. First, the rescue risk characteristics of each rescue strategy scheme in the rescue strategy scheme space are analyzed, and two types of characteristic parameters are extracted from each rescue strategy scheme, namely, rescue risk type characteristic parameters and rescue risk level characteristic parameters. Among them, the rescue risk type characteristic parameters reflect the main risk types targeted by the rescue strategy scheme, such as fire, explosion, collapse, mudslide, etc., to determine which types of rescue risks its main goal is to deal with; the rescue risk level characteristic parameters reflect the severity of the risk applicable to the rescue strategy scheme, such as high risk, medium risk, low risk, etc., to determine which levels of rescue risks it is applicable to. By performing feature analysis on each rescue strategy scheme in the rescue strategy scheme space, the rescue risk type characteristic parameters and the rescue risk level characteristic parameters are obtained, laying the foundation for subsequent strategy clustering and matching.

[0040] After obtaining the rescue risk type characteristic parameters and the rescue risk level characteristic parameters, the rescue strategy plan space is clustered and divided according to the rescue risk type characteristic parameters and the rescue risk level characteristic parameters, and the rescue strategy plans with similar characteristics are classified into the same category to form a rescue strategy clustering plan library. For example, clustering algorithms such as k-means, hierarchical clustering, and DBSCAN are used to cluster the rescue strategy plans in the rescue strategy plan space based on the rescue risk type characteristic parameters and the rescue risk level characteristic parameters. During the clustering process, the characteristic similarity between the rescue strategy plans is calculated, and they are classified into different clusters according to the similarity to form several rescue strategy clusters. Each cluster contains a group of rescue strategy plans with similar risk types and risk levels. Several rescue strategy clusters are aggregated to form a rescue strategy clustering plan library. After obtaining the rescue risk characteristic information of the target rescue site, the rescue strategy clustering plan library is used to match and select the strategy plans. The risk type and risk level parameters in the rescue risk feature information are extracted, and compared with the rescue risk type feature parameters and rescue risk level feature parameters of each cluster in the rescue strategy clustering solution library. The similarity between them is calculated using Euclidean distance, cosine similarity, etc., to obtain a risk feature similarity set, which contains the similarity value between the rescue risk feature information and each rescue strategy cluster. The higher the similarity value, the closer the rescue risk feature information is to the risk feature represented by the cluster, that is, the higher the match between the rescue strategy solution in the cluster and the target rescue site. After that, the risk feature similarity set is sorted, and the cluster with the highest similarity is selected. The rescue strategy solution contained in the cluster is the solution that best matches the risk feature of the target rescue site, and it is used as the target rescue strategy solution.

[0041] By analyzing the features, clustering and similarity matching of the rescue strategy space, the rescue strategy is selected, so that the most matching target rescue strategy can be found quickly and accurately from a large number of rescue strategy solutions, thus improving the efficiency and quality of rescue decision-making.

[0042] Furthermore, the embodiment of the present application also includes:

[0043] Path influencing parameters are extracted from the multiple rescue environment data streams and the target rescue scene environment data stream to obtain a rescue-related path parameter set; a path planning target is defined, and a path cost function is constructed according to the path planning target; the rescue-related path parameter set and the GPS positioning information are converted into a graph representation to establish a rescue scene path graph model; a path planning iteration is performed within the rescue scene path graph model using a Dijkstra planning algorithm to obtain multiple rescue path planning information; based on the path cost function, the multiple rescue path planning information is evaluated, calculated and optimized to determine the target rescue path planning information.

[0044] In a preferred embodiment, to carry out rescue path planning, firstly, the obtained multiple rescue environment data streams and target rescue site environment data streams are deeply analyzed to extract path influencing parameters. Path influencing parameters refer to factors that may affect the selection of rescue paths, such as road conditions, terrain features, obstacle distribution, location of dangerous sources, etc. By analyzing various sensor data in multiple rescue environment data streams and target rescue site environment data streams, the path influencing parameters are extracted using image recognition, geographic information analysis, data mining and other technologies. The extracted results form a rescue-related path parameter set, which includes the path influencing parameters under various rescue environments, providing a decision-making basis for subsequent path planning. Subsequently, the path planning goals are clarified, such as minimizing rescue time, minimizing rescue costs, maximizing path safety, etc. According to the determined path planning goals, a path cost function is constructed to quantitatively evaluate the advantages and disadvantages of different path plans. Afterwards, the rescue-related path parameter set and GPS positioning information are graphically converted. Among them, GPS positioning information is converted into graph nodes, representing key locations in the rescue path, such as the starting point, end point, and passing points; and path impact parameters are converted into graph edges, representing the path segment connecting two nodes; each edge is attached with corresponding path cost information to characterize the travel cost of the path segment. Through the combination of nodes and edges, a complete rescue site path graph model is constructed. This model graphically represents the geographic spatial structure and path network relationship of the rescue site, providing basic data for subsequent path search.

[0045] After obtaining the rescue site path graph model, the graph search algorithm is applied to solve the path planning, and the Dijkstra algorithm is selected as the core algorithm for path planning. The Dijkstra algorithm starts from the starting point and gradually expands outward to explore, each time selecting the node with the smallest cost to add to the explored set until the end point is reached. In the scenario of rescue path planning, the rescue starting point is taken as the source node and the rescue target point is taken as the end point. The Dijkstra algorithm is used to search on the rescue site path graph model, considering the cost information of each path, and selecting the path with the smallest comprehensive cost as the optimal solution. In order to improve the robustness and adaptability of path planning, multiple rounds of iterative search are performed, and different search parameters and constraints are set in each round of iteration, such as search depth, time limit, node access strategy, etc., so as to obtain multiple candidate rescue path planning information. After obtaining multiple candidate rescue path planning information, the node sequence and edge sequence in each candidate rescue path planning information are substituted into the path cost function, and the comprehensive cost value of the rescue path planning information is calculated. The cost values ​​of all candidate rescue path planning information are compared and sorted, and the rescue path planning information with the smallest cost is selected as the target rescue path planning information.

[0046] Through technical means such as environmental data analysis, graph model construction, and shortest path search, automatic planning and optimization of rescue routes are achieved. The data perception and processing capabilities of smart protective helmets are fully utilized, and target rescue route planning information is dynamically generated based on real-time rescue site conditions, providing accurate and efficient navigation services for rescue personnel's actions.

[0047] Furthermore, the embodiment of the present application also includes:

[0048] The multiple rescue environment data streams and the target rescue scene environment data stream are filtered preprocessed and data normalized to obtain multiple standard rescue environment data streams and a target standard rescue scene environment data stream; rescue path influencing factors are obtained, and the rescue path influencing factors include meteorological conditions, path distance and traffic conditions, terrain and obstacles, and rescue restricted areas; based on the rescue path influencing factors, associated influencing parameters of the multiple standard rescue environment data streams and the target standard rescue scene environment data stream are respectively extracted to obtain multiple associated path parameter sets and target associated path parameters; the multiple associated path parameter sets and the target associated path parameters are merged to obtain the rescue associated path parameter set.

[0049] In a preferred embodiment, before the path influence parameter is extracted, the original multiple rescue environment data streams and the target rescue scene environment data stream are preprocessed to improve the data quality and availability. Among them, the preprocessing includes filtering preprocessing and data normalization. Filtering preprocessing is to use digital signal processing technology to remove noise, interference and abnormal values ​​in the data stream. Due to the complex rescue scene environment, the data collected by the sensor is affected by various factors, and there are problems such as distortion and drift. Low-pass filtering, median filtering, Kalman filtering and other algorithms are used to smooth, denoise and correct the data stream to improve the signal-to-noise ratio and stability of the data; data normalization is to unify data from different sources and different scales into the same metric space for comparison and analysis. Since different data streams and data collected by different sensors have different physical meanings and numerical ranges, directly using the original data for processing will lead to calculation imbalance and result distortion. Therefore, the data stream is normalized by normalization methods such as maximum and minimum value normalization, zero mean normalization, and decimal calibration normalization, and mapped to the standard interval. After filtering preprocessing and data normalization, standardized rescue environment data streams are obtained, namely multiple standard rescue environment data streams and target standard rescue site environment data streams, which lay a good foundation for subsequent parameter extraction.

[0050] At the same time, based on professional knowledge and experience in the rescue field, the factors affecting the rescue route are determined, namely meteorological conditions, route distance and traffic conditions, terrain and obstacles, and rescue restricted areas. Among them, meteorological conditions include temperature, humidity, wind speed, visibility, etc., which directly affect the safety of rescue personnel and the performance of rescue equipment. For example, in a high temperature environment, rescue personnel are prone to heat stroke, dehydration and other problems. In a strong wind and low visibility environment, the driving of rescue vehicles is restricted, and the take-off and landing of helicopters are also affected; the path distance and traffic conditions determine the timeliness and accessibility of the rescue operation. The shorter the path distance, the shorter the rescue time; the better the traffic conditions, the faster the rescue speed; the terrain and obstacle information reflects the accessibility and safety of the rescue path. There may be various complex terrains at the rescue site, such as mountains, jungles, rivers, etc., as well as various obstacles, such as collapsed buildings, mudslides, etc. These factors will hinder the passage of rescue personnel and increase the risk of rescue; rescue restricted areas refer to those areas that are not allowed or suitable for rescue forces to enter due to safety, control and other reasons, such as nuclear radiation contaminated areas, military restricted areas, major hazardous sources, etc. Although these areas may be very close to the rescue target, they must be excluded in path planning due to special risks and bypass other areas.

[0051] After clarifying the factors affecting the rescue path, parameter information related to these factors is extracted from multiple standard rescue environment data streams and target standard rescue site environment data streams. For meteorological conditions, key meteorological parameters affecting the rescue operation, such as average temperature, are extracted from the data stream; for path distance and traffic conditions, parameters such as driving distance, average speed, and congestion index are calculated from the data stream. For terrain and obstacle factors, parameters such as terrain slope, obstacle height, and road width are extracted from the data stream; for rescue restricted area factors, the boundaries and attribute information of restricted areas and control areas are identified from the data stream. Through parameter extraction, multiple associated path parameter sets are obtained from multiple standard rescue environment data streams, and target associated path parameters are obtained from the target standard rescue site environment data stream. Then, multiple associated path parameter sets and target associated path parameters are integrated to form a global path parameter view as a rescue associated path parameter set.

[0052] Furthermore, the embodiment of the present application also includes:

[0053] Extract graph nodes from the rescue-related path parameter set and the GPS positioning information to obtain a path graph model node set; connect the path graph model node set to obtain a node connection edge set; label the node connection edge set in turn based on the rescue path influencing factors to obtain a node connection edge attribute set; establish the rescue site path graph model based on the path graph model node set and the node connection edge attribute set.

[0054] In a preferred embodiment, in the process of establishing a rescue site path graph model, first, the nodes in the graph are determined. The nodes of the graph represent key positions in the rescue path, such as the current rescue position, the target rescue position, the edge points of obstacles, etc. The obtained rescue-related path parameter set is analyzed, and the path nodes are identified from these parameters. For example, special positions such as road intersections and sharp turns are marked as nodes; the edge points of obstacles such as buildings and terrain are marked as nodes; the positions of rescue targets are marked as special nodes. Then, the GPS positioning information is processed, the GPS positioning information is simplified and abstracted, and the key location points are extracted as nodes of the graph. For example, the starting point, end point, turning point, etc. in the GPS positioning information are selected as nodes, and the redundant location points in the middle are ignored. By analyzing and extracting the rescue-related path parameter set and the GPS positioning information, a path graph model node set is obtained, which contains the coordinates and attribute information of all key location points in the rescue site environment.

[0055] After obtaining the node set of the path graph model, the connection relationship between the nodes, that is, the edge of the graph, is determined. In the path planning problem, an edge represents a direct path between two nodes. First, the spatial distribution characteristics of the node set of the path graph model are analyzed, and an edge is established between different nodes to represent the existing path, thereby obtaining a node connection edge set. After obtaining the node connection edge set, the attributes of each edge are annotated to reflect the characteristics of the path passing through the edge, such as travel difficulty, distance, time, etc. Each edge in the node connection edge set is traversed in turn, and its attributes are annotated. The annotation is based on the obtained rescue path influencing factors, such as meteorological conditions, terrain characteristics, traffic conditions, etc. For each edge, firstly, according to the positions of the nodes at both ends, the path influencing factor parameters in the area covered by the edge are extracted, such as the average temperature, maximum wind speed, road grade, obstacle density, etc. in the area; then, based on these parameters, the attributes of the edge such as travel difficulty, travel distance, and travel time are estimated and assigned. After edge-by-edge labeling, a set of node connection edge attributes is obtained, which contains the multi-dimensional attribute information of each edge in the rescue site path graph model, reflects the environmental characteristics and travel costs of different paths, and is an important decision-making basis for subsequent path optimization.

[0056] Subsequently, a graph data structure, such as an adjacency matrix or an adjacency list, is selected to represent the path graph model. Each node in the node set of the path graph model is mapped to a vertex in the graph data structure, and each vertex contains the node's ID, coordinates, type and other attribute information; each edge in the node connection edge attribute set is mapped to an edge in the graph data structure, and each edge contains the IDs of the two vertices it connects, as well as the edge's attribute information such as the difficulty of travel, travel distance, and travel time. Through this mapping, an abstract data model corresponding to the rescue site environment is established, namely the rescue site path graph model. This model systematically organizes and represents all key information that affects path planning in the form of a graph, such as path nodes, node connections, path attributes, etc.

[0057] Furthermore, the embodiment of the present application also includes:

[0058] Based on the path planning target, cost functions are fitted to the rescue-related path parameter set respectively to obtain a sub-target cost function set; based on the rescue demand information, criticality evaluation is performed on each sub-target in the path planning target to obtain a sub-target critical factor set; based on the sub-target critical factor set, the sub-target cost function set is weighted and fused to construct the path cost function.

[0059] In a preferred embodiment, the objectives of path planning are multifaceted, such as minimizing path length, shortest rescue time, maximizing safety, etc. In order to better quantify and evaluate the degree to which different path plans meet these objectives, a corresponding cost function is constructed for each sub-goal. First, the path planning goal is decomposed into several sub-goals. For each sub-goal, the relevant parameters are selected from the rescue-related path parameter set, and the parameters are fitted to obtain a sub-goal cost function. For example, for the sub-goal of minimizing path length, the distance parameter of the path is selected, and a linear function is used for fitting to obtain a path length cost function; for the sub-goal of shortest rescue time, the travel time parameter of the path is selected, and an exponential function is used for fitting to obtain a time cost function. By fitting the cost function for each sub-goal, a sub-goal cost function set is obtained, which contains the cost calculation methods of all sub-goals, which is used to quantitatively evaluate the performance of a path in different goal dimensions.

[0060] At the same time, the demand information related to the current rescue mission is collected and analyzed, such as the type of rescue object, danger level, environmental conditions, etc. Then, based on this information, the criticality of each sub-goal is scored. For example, in fire rescue, due to the urgency of the rescue time, the criticality score of the sub-goal of the shortest rescue time will be very high; while in earthquake rescue, due to the risk of building collapse, the criticality score of the sub-goal of maximizing safety will be very high. Through criticality evaluation, a set of sub-goal critical factors is obtained, in which each sub-goal is assigned a critical weight. The larger the weight value, the higher the importance of the sub-goal in the current rescue mission, and the larger the proportion it should occupy in the subsequent path cost function construction. With the set of sub-goal cost functions and the set of sub-goal critical factors, the final path cost function is constructed. The construction method is to perform weighted summation of the sub-goal cost functions, and the weight coefficient comes from the sub-goal critical factors. Through weighted fusion, a path cost function that takes into account multiple sub-goals and reflects the characteristics of rescue needs is established.

[0061] In summary, the intelligent rescue indication method based on the intelligent protective helmet provided in the embodiment of the present application has the following technical effects:

[0062] Obtain the target intelligent protective helmet. The target intelligent protective helmet is equipped with a sensor group, which includes a temperature and humidity sensor, a gas concentration sensor, a noise sensor, and a GPS locator. The sensor group monitors and obtains the target rescue site environmental data stream, realizes the real-time collection of rescue site environmental parameters, and provides raw data for subsequent data analysis. Perform abnormal data identification on the target rescue site environmental data stream to obtain the rescue abnormal data stream, perform risk feature assessment based on the rescue abnormal data stream, obtain rescue risk feature information, identify abnormal data reflecting dangerous situations from environmental data, and evaluate and obtain feature information representing risk levels and types, providing a decision-making basis for rescue strategy formulation. Construct a rescue strategy solution space, match the rescue risk feature information with the rescue strategy solution space, obtain the target rescue strategy solution, and obtain GPS positioning information based on the rescue site environmental data stream. According to the risk assessment results, select the best solution from the preset strategy solutions to form a specific rescue strategy, and obtain the location information of the rescue personnel to prepare for subsequent path planning. Through multiple intelligent protective helmets, multiple rescue environment data streams are obtained, and path optimization planning is performed on multiple rescue environment data streams, target rescue site environment data streams and GPS positioning information, and target rescue path planning information is obtained. Based on the target rescue strategy and target rescue path planning information, collaborative rescue instructions are given, and the data collected by multiple helmets are integrated, combined with the location information of each rescue point, global path optimization is performed, and the optimal rescue route is obtained. Based on the optimal rescue route and strategy, the rescue personnel are coordinated and directed to carry out rescue operations to improve rescue efficiency.

[0063] Embodiment 2

[0064] Based on the same inventive concept as the intelligent rescue indication method based on the intelligent protective helmet in the aforementioned embodiment, Figure 2 As shown, the embodiment of the present application provides an intelligent rescue indication system based on an intelligent protective helmet, the system comprising:

[0065] The environmental data acquisition module 11 is used to acquire the target intelligent protective helmet, which is equipped with a sensor group, including a temperature and humidity sensor, a gas concentration sensor, a noise sensor and a GPS locator, and the target rescue scene environmental data stream is monitored and acquired through the sensor group;

[0066] The risk feature acquisition module 12 is used to identify abnormal data of the target rescue scene environment data stream, obtain the rescue abnormal data stream, and perform risk feature assessment based on the rescue abnormal data stream to obtain rescue risk feature information;

[0067] A solution matching module 13 is used to construct a rescue strategy solution space, match the rescue risk feature information with the rescue strategy solution space, obtain a target rescue strategy solution, and obtain GPS positioning information according to the rescue scene environment data stream;

[0068] The collaborative rescue instruction module 14 is used to obtain multiple rescue environment data streams through multiple smart protective helmets, perform path optimization planning on the multiple rescue environment data streams, the target rescue scene environment data stream and the GPS positioning information, obtain the target rescue path planning information, and perform collaborative rescue instructions based on the target rescue strategy plan and the target rescue path planning information.

[0069] Furthermore, the risk feature acquisition module 12 includes the following execution steps:

[0070] Acquire a rescue scene database by using big data technology, wherein the rescue scene database includes a rescue risk data set of each scene attribute category;

[0071] Using a deep learning network structure, the rescue risk data sets of each scene attribute category are respectively trained to identify and obtain a rescue risk feature analysis model library;

[0072] Mark the target rescue scene with attributes and determine the target scene attribute information;

[0073] Based on the target scene attribute information, a target risk feature analysis model is called from the rescue risk feature analysis model library, and risk feature evaluation is performed on the rescue abnormal data stream based on the target risk feature analysis model to output the rescue risk feature information.

[0074] Furthermore, the solution matching module 13 includes the following execution steps:

[0075] Performing rescue risk feature analysis on the rescue strategy solution space to obtain rescue risk type feature parameters and rescue risk level feature parameters;

[0076] Clustering and dividing the rescue strategy solution space according to the rescue risk type characteristic parameters and the rescue risk level characteristic parameters to obtain a rescue strategy clustering solution library;

[0077] Performing similarity analysis based on the rescue risk feature information and the rescue strategy clustering solution library to obtain a risk feature similarity set;

[0078] The rescue strategy clustering solution library is screened and matched based on the risk feature similarity set to determine the target rescue strategy solution.

[0079] Furthermore, the collaborative rescue instruction module 14 includes the following execution steps:

[0080] Extracting path influence parameters from the multiple rescue environment data streams and the target rescue scene environment data stream to obtain a rescue-related path parameter set;

[0081] Defining a path planning goal, and constructing a path cost function according to the path planning goal;

[0082] Convert the rescue-related path parameter set and the GPS positioning information into a graph representation to establish a rescue site path graph model;

[0083] Using the Dijkstra planning algorithm to iterate the path planning in the rescue scene path graph model, a plurality of rescue path planning information is obtained;

[0084] The plurality of rescue path planning information are evaluated and calculated based on the path cost function to determine the target rescue path planning information.

[0085] Furthermore, the collaborative rescue instruction module 14 also includes the following execution steps:

[0086] Performing filtering preprocessing and data normalization on the multiple rescue environment data streams and the target rescue scene environment data stream to obtain multiple standard rescue environment data streams and a target standard rescue scene environment data stream;

[0087] Obtaining factors affecting the rescue path, wherein the factors affecting the rescue path include weather conditions, path distance and traffic conditions, terrain and obstacles, and rescue restricted areas;

[0088] Based on the rescue path influencing factors, the associated influencing parameters of the multiple standard rescue environment data streams and the target standard rescue scene environment data stream are extracted to obtain multiple associated path parameter sets and target associated path parameters;

[0089] The multiple associated path parameter sets and the target associated path parameter are merged to obtain the rescue associated path parameter set.

[0090] Furthermore, the collaborative rescue instruction module 14 also includes the following execution steps:

[0091] Extracting graph nodes from the rescue-related path parameter set and the GPS positioning information to obtain a path graph model node set;

[0092] Connecting the node set of the path graph model by edges to obtain a node connection edge set;

[0093] Based on the rescue path influencing factors, the node connection edge set is sequentially attributed to obtain a node connection edge attribute set;

[0094] The rescue scene path graph model is established based on the path graph model node set and the node connection edge attribute set.

[0095] Furthermore, the collaborative rescue instruction module 14 also includes the following execution steps:

[0096] Based on the path planning goal, cost functions are respectively fitted to the rescue-related path parameter set to obtain a sub-goal cost function set;

[0097] Performing a critical evaluation on each sub-goal in the path planning goal according to the rescue demand information to obtain a set of sub-goal critical factors;

[0098] The sub-goal cost function set is weightedly fused based on the sub-goal critical factor set to construct the path cost function.

[0099] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0100] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application is intended to include these modifications and variations.

Claims

1. An intelligent rescue indication method based on an intelligent protective helmet, characterized in that: The method comprises: Acquire a target smart protective helmet, wherein the target smart protective helmet is equipped with a sensor group, wherein the sensor group includes a temperature and humidity sensor, a gas concentration sensor, a noise sensor, and a GPS locator, and the target rescue scene environment data stream is monitored and acquired through the sensor group; Performing abnormal data identification on the target rescue scene environment data stream to obtain a rescue abnormal data stream, and performing risk feature assessment based on the rescue abnormal data stream to obtain rescue risk feature information; Constructing a rescue strategy plan space, matching the rescue risk feature information with the rescue strategy plan space to obtain a target rescue strategy plan, and acquiring GPS positioning information according to the rescue scene environment data stream; Acquire multiple rescue environment data streams through multiple smart protective helmets, perform path optimization planning on the multiple rescue environment data streams, the target rescue scene environment data stream and the GPS positioning information, obtain target rescue path planning information, and provide collaborative rescue instructions based on the target rescue strategy and the target rescue path planning information; The obtaining of rescue risk characteristic information includes: Acquire a rescue scene database by using big data technology, wherein the rescue scene database includes a rescue risk data set of each scene attribute category; Using a deep learning network structure, the rescue risk data sets of each scene attribute category are respectively trained to identify and obtain a rescue risk feature analysis model library; Mark the target rescue scene with attributes and determine the target scene attribute information; Based on the target scene attribute information, calling a target risk feature analysis model from the rescue risk feature analysis model library, and performing risk feature assessment on the rescue abnormal data stream based on the target risk feature analysis model, and outputting the rescue risk feature information; The target rescue strategy scheme is obtained, including: Performing rescue risk feature analysis on the rescue strategy solution space to obtain rescue risk type feature parameters and rescue risk level feature parameters; Clustering and dividing the rescue strategy solution space according to the rescue risk type characteristic parameters and the rescue risk level characteristic parameters to obtain a rescue strategy clustering solution library; Performing similarity analysis based on the rescue risk feature information and the rescue strategy clustering solution library to obtain a risk feature similarity set; Screening and matching the rescue strategy clustering solution library based on the risk feature similarity set to determine the target rescue strategy solution; The obtaining of target rescue path planning information includes: Extracting path influence parameters from the multiple rescue environment data streams and the target rescue scene environment data stream to obtain a rescue-related path parameter set; Defining a path planning goal, and constructing a path cost function according to the path planning goal; Convert the rescue-related path parameter set and the GPS positioning information into a graph representation to establish a rescue site path graph model; Using the Dijkstra planning algorithm to iterate the path planning in the rescue scene path graph model, a plurality of rescue path planning information is obtained; The plurality of rescue path planning information are evaluated and calculated based on the path cost function to determine the target rescue path planning information.

2. The method according to claim 1, characterized in that The step of obtaining a rescue-related path parameter set includes: Performing filtering preprocessing and data normalization on the multiple rescue environment data streams and the target rescue scene environment data stream to obtain multiple standard rescue environment data streams and a target standard rescue scene environment data stream; Obtaining factors affecting the rescue path, wherein the factors affecting the rescue path include weather conditions, path distance and traffic conditions, terrain and obstacles, and rescue restricted areas; Based on the rescue path influencing factors, the associated influencing parameters of the multiple standard rescue environment data streams and the target standard rescue scene environment data stream are extracted to obtain multiple associated path parameter sets and target associated path parameters; The multiple associated path parameter sets and the target associated path parameter are merged to obtain the rescue associated path parameter set.

3. The method according to claim 2, characterized in that The step of establishing a rescue scene path diagram model includes: Extracting graph nodes from the rescue-related path parameter set and the GPS positioning information to obtain a path graph model node set; Connecting the node set of the path graph model by edges to obtain a node connection edge set; Based on the rescue path influencing factors, the node connection edge set is sequentially attributed to obtain a node connection edge attribute set; The rescue scene path graph model is established based on the path graph model node set and the node connection edge attribute set.

4. The method according to claim 3, characterized in that The constructing a path cost function according to the path planning target includes: Based on the path planning goal, cost functions are respectively fitted to the rescue-related path parameter set to obtain a sub-goal cost function set; Performing a critical evaluation on each sub-goal in the path planning goal according to the rescue demand information to obtain a set of sub-goal critical factors; The sub-goal cost function set is weightedly fused based on the sub-goal critical factor set to construct the path cost function.

5. The intelligent rescue indication system based on the intelligent protective helmet is characterized in that: The system is used to implement the intelligent rescue indication method based on the intelligent protective helmet according to any one of claims 1 to 4, and comprises: An environmental data acquisition module, which is used to acquire a target smart protective helmet. The target smart protective helmet is equipped with a sensor group, which includes a temperature and humidity sensor, a gas concentration sensor, a noise sensor, and a GPS locator. The sensor group is used to monitor and acquire a target rescue scene environmental data stream; A risk feature acquisition module, the risk feature acquisition module is used to identify abnormal data on the target rescue scene environment data stream, obtain a rescue abnormal data stream, perform risk feature assessment based on the rescue abnormal data stream, and obtain rescue risk feature information; A solution matching module, wherein the solution matching module is used to construct a rescue strategy solution space, match the rescue risk feature information with the rescue strategy solution space, obtain a target rescue strategy solution, and simultaneously obtain GPS positioning information according to the rescue scene environment data stream; A collaborative rescue instruction module is used to obtain multiple rescue environment data streams through multiple smart protective helmets, perform path optimization planning on the multiple rescue environment data streams, the target rescue scene environment data stream and the GPS positioning information, obtain target rescue path planning information, and perform collaborative rescue instructions based on the target rescue strategy plan and the target rescue path planning information.