Precise earthquake rescue method in confined spaces using micro-intelligent robots

Abstract

The present application relates to the field of data processing technology, and provides a method for precise rescue in confined spaces during earthquakes using micro-intelligent robots. The method includes: obtaining millimeter-wave radar point cloud data and infrared thermal imager thermal imaging data based on the target area; formulating a three-dimensional model of the ruins environment and generating a traffic path map based on the data; deploying adaptive networking for a micro-robot cluster, and performing structural stability assessment and residual vibration risk prediction in combination with the map; allocating detection partitions and rescue tasks of the micro-robot cluster in real time based on the assessment results; introducing vital signs information, optimizing detection partitions and balanced task allocation, and outputting a parallel scheduling strategy. The present application solves the technical problem of low efficiency and accuracy in confined space rescue due to complex environments and incomplete perception data, and realizes the technical effect of using micro-robot clusters to conduct life search and rescue and task allocation in confined spaces, thereby improving the accuracy and efficiency of earthquake confined space rescue.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method for precise earthquake rescue in confined spaces using a micro-intelligent robot. Background Art

[0002] Earthquake disasters are characterized by suddenness, destructiveness, and difficulty in rescue. In particular, the collapse of buildings caused by earthquakes creates numerous small, complex spaces of debris, which are often difficult to access using traditional rescue methods, severely limiting rescue efficiency. Within the debris, trapped individuals can be extremely vulnerable, with limited living space and the constant risk of aftershocks and secondary structural collapse. Therefore, achieving accurate, rapid, and safe rescue within confined spaces has become a critical challenge in the earthquake rescue field.

[0003] Traditional rescue methods rely primarily on manual search and rescue and large mechanical equipment, but these methods have significant limitations in confined spaces: manual search and rescue is inefficient and risky, and large equipment struggles to access complex debris environments. With the development of intelligent technology, micro-robots, due to their compact size, high flexibility, and adaptability, have become a crucial tool for earthquake rescue operations in confined spaces. However, existing micro-robots still have shortcomings in environmental perception, path planning, and task allocation, making them unable to meet the demands of precise rescue in complex debris environments. Summary of the Invention

[0004] This application provides a precise earthquake rescue method in confined spaces using micro-intelligent robots, aiming to solve the technical problem of low efficiency and accuracy in confined space rescue due to complex environment and incomplete perception data.

[0005] The present application provides a method for precise rescue in confined spaces during earthquakes using micro-intelligent robots, the method comprising: obtaining point cloud data corresponding to a millimeter-wave radar and thermal imaging data corresponding to an infrared thermal imager based on a target area; formulating a three-dimensional internal model of a ruins environment corresponding to a confined space during an earthquake based on the point cloud data corresponding to the millimeter-wave radar and the thermal imaging data corresponding to the infrared thermal imager, and generating a traffic path map; deploying an adaptive network for a micro-robot cluster, and performing structural stability assessment and residual vibration risk prediction in combination with the traffic path map; allocating detection partitions and rescue tasks of the micro-robot cluster in real time using the adaptive network based on the structural stability assessment results and the residual vibration risk prediction results; introducing vital sign sampling information, optimizing the balanced distribution of detection partitions and rescue tasks of the micro-robot cluster, and outputting a parallel scheduling strategy.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The aforementioned precise earthquake rescue method for confined spaces using micro-sized intelligent robots utilizes a combination of millimeter-wave radar and infrared thermal imaging to acquire data from the target area. Based on this data, a three-dimensional model of the ruins' interior is constructed, and a path map suitable for the micro-sized robots to navigate is generated. The robot swarm, through adaptive networking combined with the path map, conducts structural stability assessments and aftershock risk predictions, ensuring that the robots operate within safe areas. Based on these assessment results, the robot swarm allocates detection zones and rescue tasks in real time, and balances and optimizes these tasks by collecting vital sign information (such as body temperature and micro-vibrations). Finally, a parallel scheduling strategy is developed to improve rescue efficiency and ensure proper task prioritization and allocation, ultimately ensuring efficient and accurate rescue operations.

[0008] 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

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0010] Figure 1 The figure is a flow chart of a method for precise earthquake rescue in a confined space using a micro-intelligent robot in one embodiment.

[0011] Figure 2 The figure is a flow chart of precision calibration of a three-dimensional model of a ruins environment in an embodiment of a precise earthquake rescue method in a confined space using a micro-intelligent robot. DETAILED DESCRIPTION

[0012] The embodiments of the present application provide a precise earthquake rescue method in a small space using a miniature intelligent robot, thereby solving the technical problem of low efficiency and accuracy in rescue in a small space due to complex environment and incomplete perception data.

[0013] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only some of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0014] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0015] Examples, such as Figure 1 As shown, the present application provides a method for precise earthquake rescue in confined spaces using a micro-sized intelligent robot, the method comprising:

[0016] Based on the target area, obtain the point cloud data corresponding to the millimeter wave radar and the thermal imaging data corresponding to the infrared thermal imager.

[0017] In an embodiment of the present application, a millimeter-wave radar is first used to scan the target area to obtain point cloud data reflecting the distribution of objects in the area. This data can provide accurate distance and spatial information, which helps to identify structures, obstacles, and gaps in the ruins. At the same time, an infrared thermal imager is used to obtain thermal imaging data of the target area. This data can show the temperature differences in the ruins, helping to identify potential living spaces and heat source signals of trapped people. By combining these two different types of data, the spatial structure and temperature distribution of the ruins environment can be fully understood, providing accurate basic information for subsequent modeling and rescue work.

[0018] Based on the point cloud data corresponding to the millimeter-wave radar and the thermal imaging data corresponding to the infrared thermal imager, a three-dimensional internal model of the ruins environment corresponding to the narrow space of the earthquake is drawn up, and a traffic path map is generated.

[0019] In one embodiment, data from millimeter-wave radar and infrared thermal imagers are used to conduct a detailed analysis of the ruin environment and create an accurate three-dimensional spatial model. Specifically, the point cloud data from the millimeter-wave radar is first aligned and fused with the thermal imaging data from the infrared thermal imager. The millimeter-wave radar data provides the location, shape, and spatial structure of objects within the ruins, while the thermal imaging data reveals temperature variations within the ruins, particularly the location of heat sources, which can help locate trapped personnel or critical equipment. Data fusion technology comprehensively considers information from both sources, reducing the potential error introduced by a single data source. Subsequently, using the fused data, computer modeling software converts the point cloud data into a three-dimensional spatial model. This model depicts structural features within the ruins, such as walls, rubble accumulations, passageways, and voids. Furthermore, based on the thermal imaging data, heat source areas are marked in the model. These areas may be the locations of trapped personnel or areas requiring priority rescue. Through continuous iteration and optimization of the model, the model is ensured to reproduce the actual conditions of the ruins as faithfully as possible, resulting in a final three-dimensional model of the ruin environment corresponding to the earthquake's confined space. Next, based on the constructed 3D model, a path map is generated for the microrobot to safely navigate through the ruins. This process first defines the robot's motion constraints, such as body size and joint range of motion. Combined with the spatial structure of the 3D ruins model, a preliminary feasible path is calculated, avoiding obstacles and unstable structures within the ruins. The safety and efficiency of this path are then evaluated to ensure the robot can avoid dangerous areas and complete its mission efficiently in the shortest possible time. Finally, the resulting path map guides the microrobot to the optimal route, ensuring its safe and efficient movement through the ruins, enabling it to effectively carry out rescue missions in complex ruin environments.

[0020] Further, if Figure 2 As shown, the present application provides a method for developing a three-dimensional model of the interior of a ruined environment corresponding to a narrow earthquake space based on point cloud data corresponding to the millimeter-wave radar and thermal imaging data corresponding to the infrared thermal imager. The method further includes:

[0021] The internal cavity structure of the ruins in the target area is detected, and a dielectric constant distribution map is generated to identify the potential survival space of the three-dimensional model of the interior of the ruin environment; based on the potential survival space of the three-dimensional model of the interior of the ruin environment, the structural deformation caused by aftershocks is monitored and the danger zone markers are dynamically updated; and the updated danger zone markers are used to perform precision calibration on the three-dimensional model of the interior of the ruin environment.

[0022] Preferably, after obtaining a three-dimensional model of the interior of the ruins environment, this three-dimensional model will be optimized. First, high-frequency electromagnetic waves (such as microwaves or radar waves) are used for detection, and their propagation characteristics (reflection, refraction, absorption, etc.) in different media are used to identify the cavity structure in the ruins. Electromagnetic waves will produce different reflection waveforms when encountering different materials. The electromagnetic wave reflection characteristics of the cavities or gaps inside the ruins are usually significantly different from those of the surrounding solid structures (such as masonry, steel bars, etc.). By measuring the intensity, time delay, and frequency variation of the reflected electromagnetic waves, the shape, location, and size of the cavity can be determined. After obtaining the electromagnetic wave reflection signal of the ruins area, the dielectric constant of each area of ​​the ruins will be calculated based on the characteristics of the reflected wave. The dielectric constant is the ability of a material to respond to an electric field and directly affects the propagation speed and reflection intensity of electromagnetic waves in a medium. In the ruins environment, different materials (such as metal, concrete, wood, etc.) have different dielectric constants, and the cavities exhibit electromagnetic reflection characteristics that are significantly different from those of the surrounding materials. Subsequently, based on the electromagnetic wave reflection information at each location, a dielectric constant distribution map of the ruins is generated. This map shows the dielectric constant values ​​of different areas within the ruins. Cavities or gaps typically appear as low dielectric constant areas, while solid materials (such as masonry, steel bars, etc.) exhibit higher dielectric constants. This distribution map can help identify potential survival spaces within the ruins. These cavities may be hiding places for trapped people. Based on the generated dielectric constant distribution map, low-dielectric constant areas are marked as potential survival spaces. These areas are typically gaps, caves, or spaces that have not completely collapsed within the ruins. These areas should be prioritized for detection and search by robots. After an earthquake, the ruins structure will continue to be affected by aftershocks. By installing equipment such as seismic sensors, accelerometers, or displacement sensors, the deformation of key structural parts in the ruins can be monitored in real time. These sensors can detect tiny vibrations and displacements within the ruins, especially those that may cause further structural collapse. Then, based on the monitored structural deformation data, combined with earthquake parameters such as the epicenter, magnitude, and duration of the earthquake, a mathematical model can be used to predict aftershocks. By simulating the possible impact of aftershocks on the ruins structure, it is assessed which areas may further collapse or be damaged in future aftershocks, and the aftershock collapse probability is generated. This prediction process can identify high-risk areas in advance, thereby guiding subsequent rescue operations. After obtaining the aftershock collapse probability, this aftershock collapse probability will be compared with the preset probability threshold. When the aftershock collapse probability is greater than or equal to the preset probability threshold, the corresponding potential survival space will be marked as a dangerous area. These marked areas are usually unstable due to structural deformation caused by aftershocks, which may lead to further collapse or crack expansion.As these danger zone markers are updated, the original 3D model undergoes a precision calibration. This involves mapping the current potential survival space onto the 3D model of the ruins environment, along with the danger zone markers, to ensure the ruins model reflects the latest structural changes. After precision calibration, the robot swarm can accurately navigate based on the updated 3D model, avoiding new danger zones while more accurately locating potential survival spaces, improving the safety and efficiency of search and rescue operations.

[0023] The mathematical model used for aftershock prediction can be constructed based on a long short-term memory (LSTM) network. The training data used in the construction process includes sample structural deformation data, sample earthquake parameters, sample ruin structural characteristics (such as structural component types and material properties), and sample aftershock collapse probabilities. During the training process, a model structure is constructed, including an input layer, an LSTM hidden layer, a fully connected layer, and an output layer. The input layer of the model is then defined, and the sample structural deformation data, sample earthquake parameters, and sample ruin structural characteristics (such as structural component types and material properties) are combined to form a feature sequence, which serves as the input data for the input layer. An LSTM hidden layer is constructed after the input layer to extract the time series features and dynamic change trends in the input data. One or more fully connected layers are then connected to further integrate the features. Finally, the output layer outputs the predicted value of the sample aftershock collapse probability. After the model structure is constructed, the LSTM model weights are initialized using a random initialization method. Training data is then fed into the initialized LSTM prediction model for forward propagation. The data is passed through the input layer to the LSTM hidden layer for feature extraction and temporal dependency learning. The feature vectors are further integrated through a fully connected layer, and the output layer generates the predicted aftershock collapse probability. After obtaining the predicted results, the mean squared error (MSE) loss function is used to calculate the loss between the predicted output and the sample aftershock collapse probability. The backpropagation algorithm is then used to calculate the gradient of the loss with respect to each neuron weight in the LSTM hidden layer, fully connected layer, and output layer. Subsequently, the Adam optimizer is used to iteratively optimize and update the model parameters, dynamically adjusting the network weights and biases to minimize the loss function. This process is repeated until the preset maximum number of iterations is reached, and the model parameters gradually converge to a stable state. After training, the model performance is evaluated using a validation set (data not used for training) to determine the accuracy of the predicted aftershock structural collapse probability and the model's generalization ability. If the verification results meet the expected requirements, the current LSTM structural model will be output as the final aftershock structural deformation prediction model; if it does not meet the expectations, the hyperparameters such as learning rate, number of hidden layer nodes, training batch size, etc. will be further adjusted to continuously improve the prediction accuracy of the model.

[0024] Furthermore, the present application provides a method for generating a traffic path map, the method comprising:

[0025] A robot motion capability constraint matrix is ​​defined, including body size, joint range of motion, and obstacle crossing quantitative parameters. An initial feasible path is calculated based on the robot motion capability constraint matrix, and the path nodes are embedded in the potential survival space of the three-dimensional model inside the ruins environment. The safety and efficiency of the initial feasible path are evaluated to determine the traversable path map.

[0026] Optionally, in order to accurately plan the movement path of the micro-robot in the earthquake ruins, it is necessary to define a constraint matrix for the robot's motion capabilities. This matrix includes the robot's body dimensions (such as length, width, and height), the maximum range of motion of the joints (such as the rotation angle and extension length of the robot arm), and quantitative parameters of the robot's obstacle crossing capabilities, such as the obstacle height and maximum slope that the robot can cross. These parameters are used to clarify the spatial limitations and capability range of the robot's movement in the actual environment. Subsequently, the previously established three-dimensional model of the ruins environment is voxelized or meshed to form a structured environmental grid. Each grid cell stores structural features such as spatial dimensions, obstacle distribution, and cavity location. After meshing, a data structure is formed that is easy to perform path search calculations, facilitating the implementation of the path planning algorithm. The potential survival space locations previously determined by millimeter-wave radar, thermal imaging, dielectric constant detection, etc. are then associated with the environmental grid data to clarify the survival space locations and corresponding grid cell coordinates. These potential survival spaces are then embedded as important path nodes in the path planning process to ensure that the generated path can cover all spatial areas that need to be searched and rescued. Next, based on the robot's motion constraint matrix, a path search algorithm (such as A*, Dijkstra, or RRT fast search tree algorithm) is used to calculate an initial feasible path within the three-dimensional ruins environment. For example, the robot's starting position is used as the path starting point; the potential survival space location is used as the node target that the path must pass through. Each path node is gradually checked using the motion constraint matrix, eliminating nodes with inconsistent dimensions, difficult obstacles to cross, restricted joints, or obvious obstructions. Based on the area the robot can actually reach, an initial path connecting each potential survival space node is calculated sequentially. This results in a set of initial feasible paths, in which the path nodes clearly mark the robot's starting point, target search and rescue space, and path turning points. The determined path nodes are then spatially matched with the potential survival space locations. That is, when the initial path passes through or approaches a potential survival space, the center of the survival space is precisely marked on the path node. This allows the robot to clearly stop at the node and conduct detailed exploration and rescue operations during subsequent mission execution. If the initial path is found to not fully pass through the potential survival space, the path nodes can be further optimized, adding or adjusting the path to cover as much of the survival space as possible. Finally, the generated initial feasible path is further evaluated and analyzed.Specifically, each path will be evaluated for safety and efficiency. The safety assessment focuses on the stability of the structures around the path and whether there is a potential risk of collapse or aftershock impact. The path safety score of the path is obtained by weighting the probability of no collapse due to aftershocks (obtained by subtracting the probability of collapse due to aftershocks from 1) at each node on the path. The weight of each node is determined based on the number of obstacles reaching the node. The efficiency assessment focuses on the path length, the estimated time to reach the target area, etc. The path efficiency score is obtained by weighting the total length and total time of the path. By weighting the path safety score and path efficiency score, one or more optimal paths are determined, thus forming a complete robot action path map to guide the robot cluster to efficiently and safely implement precise rescue operations.

[0027] Adaptive networking is deployed for micro-robot clusters, and structural stability assessment and residual vibration risk prediction are performed in combination with the traffic path map.

[0028] In one embodiment, during a rescue mission involving a swarm of micro-robots, deploying adaptive networking is key to ensuring effective swarm collaboration and flexible response to emergencies. Adaptive networking can adjust the robots' communication and collaboration modes in real time based on mission requirements and environmental changes. First, based on mission requirements, the robot swarm connects to a central control system via a wireless communication network. Each robot dynamically connects and disconnects with other robots using adaptive networking technology, ensuring stable communication and coordinated work among the robots within the swarm. Adaptive networking can adjust the network topology based on each robot's location, mission importance, and communication quality, thereby avoiding information delays or interruptions caused by communication issues. During this process, the robots' communication network monitors and updates the network status in real time. When the robots move or the environment changes, the network topology is automatically adjusted to ensure stable and real-time data transmission. For example, if the signal in a certain area is poor, the robot swarm will dynamically select a more stable path for information transmission. Subsequently, the structural stability of the ruins is assessed based on a previously constructed path map. This map not only contains the specific routes the robots will follow but also labels the safety indicators for each area within the ruins. The robot cluster will follow the path in the map and continuously monitor environmental changes using sensors (such as millimeter-wave radar and infrared thermal imagers). If a robot detects new structural damage or instability in a certain area while passing through it, it will collect data such as strain and stress distribution in that area and transmit it back in real time. Using this real-time environmental data, the potential risk areas within the ruins can be updated on the three-dimensional model of the ruins environment, and a structural stability analysis can be performed. Specifically, a finite element solver (such as ANSYS or ABAQUS) is used to perform numerical calculations on the updated model. This solver will use numerical methods (such as direct and iterative methods) to solve for the structural displacement, stress, and strain distribution of the structure, obtaining the response of the entire ruin structure. This is used to assess the bearing capacity of the risk area and determine its stability, thereby obtaining a structural stability assessment result. Furthermore, residual vibration risk prediction will be conducted for risk areas. Using the previously constructed mathematical model, the probability of aftershock collapse will be calculated based on an updated 3D model of the interior of the ruins environment. This will be used as the residual vibration risk prediction result. The structural stability assessment results and residual vibration risk prediction results will be used to dynamically adjust task allocation to ensure efficient and safe rescue operations.

[0029] Based on the structural stability assessment results and the residual vibration risk prediction results, the detection partitions and rescue tasks of the micro-robot cluster are allocated in real time using the adaptive networking.

[0030] In one embodiment, the detection and rescue missions of a micro-robot swarm can be dynamically adjusted based on the results of a ruins structural stability assessment and residual vibration risk prediction. First, the structural stability assessment provides the bearing capacity of each risk zone in the ruins, while the residual vibration risk prediction provides the collapse probability of each risk zone. These data are combined to form a comprehensive ruins risk map. Based on this, adaptive networking technology dynamically assigns tasks based on each robot's current location, function, health status, and real-time feedback from the environment. Following instructions, the robot swarm will proceed to high-risk areas for deeper exploration while avoiding areas already marked as high-risk, ensuring the safety of the swarm and efficient execution of tasks. For example, if the stability of a particular area falls below a lower stability limit and the collapse probability exceeds an upper collapse probability limit, it will be marked as a high-risk zone. Robots in other areas will be prioritized for search and rescue missions, preventing robots in high-risk areas from becoming potential traps. At the same time, adaptive networking technology ensures that each robot can share detection data in real time, coordinate with each other, and collaborate to complete tasks. If a robot detects new danger signals (such as further structural damage or changes in vital signs), its discovered data is promptly transmitted to other robots and systems in the cluster. This feedback data immediately updates task allocation, ensuring the progress and efficiency of the entire robot cluster. Ultimately, all robots will evenly distribute and optimize tasks based on structural stability and risk assessment results, ensuring that areas most in need receive timely and adequate rescue, while minimizing the risks associated with structural instability.

[0031] Vital sign sampling information is introduced to optimize the balanced distribution of detection partitions and rescue tasks of the micro-robot cluster, and a parallel scheduling strategy is output.

[0032] In one embodiment, when conducting rescue operations within earthquake debris, task allocation within a robot swarm must consider not only the structural stability of the debris and the risk of aftershocks, but also the vital signs of trapped individuals. Introducing vital sign sampling information can significantly improve rescue efficiency and prioritize priorities. Vital sign information, including important parameters such as body temperature and micro-vibrations, can be collected in real time by sensors carried by the robots. During mission execution, each robot detects and collects vital sign data from trapped individuals within its coverage area. When vital sign data in a particular area indicates the presence of trapped individuals, and if the vital signs indicate they are in a critical condition, this area is assigned a higher priority, and robots are promptly dispatched to the rescue. This real-time vital sign information allows for dynamic optimization of all detection areas, adjusting task allocation strategies based on the trapped individuals' vital conditions, the hazardousness of the environment, and the available resources of the robot swarm. For example, if vital sign data indicates that trapped individuals in certain areas may be in a critical condition due to hypoxia or severe injury, tasks can be automatically adjusted, prioritizing robots to those areas for emergency rescue. On this basis, balanced scheduling will be carried out based on the collaboration between robots. Taking into account the scale of the robot cluster and the working capabilities of different robots (for example, some robots may need more time for material replenishment or maintenance), the task load of each robot will be dynamically adjusted to ensure the efficient operation of the robot cluster and that some robots will not be overloaded or waste resources. The parallel scheduling strategy will enable multiple tasks to be carried out simultaneously, minimizing the rescue time and reducing secondary injuries caused by structural instability or vibration. In short, by introducing vital signs information and combining it with real-time data analysis, detection zoning and rescue tasks can be comprehensively optimized to ensure that rescue activities are carried out efficiently, scientifically, and accurately, while ensuring that the most critical life-saving tasks are given priority.

[0033] Furthermore, the present application provides that the vital signs sampling information includes body temperature marker data, micro-vibration marker data and sweat marker data; based on the micro-vibration marker data, the micro-vibration signal of the trapped person's body surface is extracted, and the respiratory rate characteristics and heartbeat waveform characteristics are separated; the state criticality level is evaluated through the respiratory rate characteristics and heartbeat waveform characteristics, and the state criticality level is used to prioritize the rescue tasks.

[0034] Preferably, during earthquake rescue operations, robot swarms use a variety of sensors to collect vital sign information from trapped individuals, providing real-time insights into their health status. This is especially true in rubble, where many trapped individuals may face life-threatening situations due to prolonged time and harsh conditions. To quickly determine the criticality of each individual, rescue priorities are determined by analyzing vital sign markers. The collected vital sign information primarily includes temperature marker data, micro-vibration marker data, and sweat marker data. Temperature markers can help determine whether a trapped individual is experiencing low or high temperatures, thereby inferring physical exhaustion and shock. Sweat markers can help identify stress responses and determine whether the individual's body is in an abnormal state due to dehydration, overexertion, or excessive stress. Micro-vibration marker data is particularly critical. Using micro-vibration sensors, the robots can capture tiny vibration signals from the trapped individual's body surface, which contain information such as the individual's breathing rate and heartbeat waveform. By extracting and analyzing these microvibration signals, respiratory rate and heartbeat waveform features can be precisely isolated. Respiratory rate reflects the trapped person's breathing status; excessively fast or slow breathing may indicate unstable vital signs. The heartbeat waveform, on the other hand, reveals the heart's functioning; abnormal waveforms may indicate internal injuries or the risk of other life-threatening illnesses. Based on these characteristics, the criticality of the trapped person can be further assessed. Abnormal respiratory rate, erratic heartbeat waveforms, or hypothermia indicate a high-risk condition, leading to priority allocation of rescue resources. This assessment allows the robot swarm to intelligently prioritize rescue tasks, ensuring the most critically ill receive the fastest rescue. Therefore, by comprehensively analyzing vital sign data, particularly microvibration markers, the health status of the trapped person can be efficiently assessed. Based on this assessment, the robot swarm's rescue missions can be rationally scheduled to maximize the utilization of rescue resources and the safety of the trapped.

[0035] Furthermore, the state criticality level is used to prioritize the rescue missions, and the method further includes:

[0036] A digital twin engine is deployed in the adaptive network to simulate the interaction force between the motion trajectory of the micro-robot cluster and the ruin structure in real time; the risk of secondary collapse caused by the operation of the robotic arm is predicted based on the simulation data of the interaction force between the motion trajectory of the micro-robot cluster and the ruin structure; and based on the secondary collapse risk, a structural support reinforcement task is generated and added to the rescue task queue.

[0037] Optionally, a digital twin engine is deployed within the adaptive network to simulate in real time the interactions between the motion trajectory of the microrobot swarm and the ruined structure. This virtual model, a highly accurate reconstruction of the ruin's three-dimensional environment and the dynamic behavior of the robot swarm, reflects the interactions between the robots and the structure as they move through the ruins. Each time the robot swarm performs a task, the digital twin engine interacts with sensor data, robot motion data, and a three-dimensional model of the ruin's interior to update the relative positions between the robots and the ruins, contact forces, and the pressure exerted by the robots' motion on the structure. This real-time simulation focuses on the operation of the robotic arm, as these often require intensive physical manipulation, such as lifting heavy objects or clearing obstacles. These movements can potentially impact structural components within the ruins, particularly when the structure is partially damaged or unstable. Through digital twin simulations and finite element analysis, the risk of secondary collapse, which may be triggered by the robot arm during its mission, can be anticipated. Secondary collapse refers to the further collapse of a structure after initial damage when subjected to external forces, posing a significant threat to rescue efforts. Based on the secondary collapse risk data obtained through the digital twin engine simulation, it will automatically assess which areas are at high risk, and generate structural support reinforcement tasks based on this information. These reinforcement tasks will be added to the rescue task queue as a priority. The purpose of the reinforcement task is to reduce the possibility of secondary collapse by providing additional support or repairing structural weaknesses before conducting other rescue operations, thereby ensuring the safety of the robot cluster and trapped personnel. Reinforcement tasks may include reinforcing the structure, stabilizing key areas with temporary supports, or marking and isolating areas in the ruins that may cause collapse, ensuring that subsequent rescue missions can proceed smoothly without causing more serious damage. This process can dynamically adjust the rescue plan through real-time feedback from the digital twin engine, ensuring that not only can trapped people be found quickly during the rescue operation, but also that potential secondary collapse risks can be effectively avoided, thereby improving overall rescue efficiency and safety.

[0038] Furthermore, the present application provides a method of triggering a federated learning mechanism to integrate the operation data of each robot corresponding to the micro-robot cluster when the rescue mission queue is updated; through the federated learning mechanism, combined with the state criticality level, the detection partition of the micro-robot cluster is updated online.

[0039] Optionally, when the rescue task queue is updated, a federated learning mechanism is triggered. The purpose is to integrate the operation data of each robot in the micro-robot cluster and dynamically adjust the detection tasks based on this data. Specifically, as the on-site rescue situation changes, such as newly discovered high-risk areas, changes in the status of trapped people, or changes in the robot's work progress, the rescue task queue will be updated in real time. The queue update is usually sorted and adjusted based on multiple factors such as the priority of the task, the vital signs of the rescuers, and the structural stability of the ruins. Once the task queue changes, the federated learning mechanism will be triggered. Federated learning is a distributed machine learning method. Under this method, each robot in the robot cluster will process and learn its own operation data (such as path information, task completion status, environmental changes, etc.) locally without sending the data to the system. This mechanism can ensure the privacy and real-time nature of the data, while avoiding the system response lag caused by data transmission delays. Each robot analyzes local data and updates its own learning model. For example, it optimizes itself based on its own task performance (such as obstacles encountered and path selection efficiency). This data is not directly transmitted to the system. Instead, it is aggregated into a higher-level global model through the federated learning mechanism. Robots collaborate to share information and optimize their individual models, thereby improving the efficiency and rescue effectiveness of the entire robot cluster. Based on federated learning, further task optimization is performed based on the criticality level of trapped individuals. Specifically, each trapped individual's criticality is determined based on their vital signs (such as breathing rate and heart rate waveform). Areas with higher criticality are prioritized for more robot resources. This priority adjustment is updated in real time by the robot cluster's control system. Subsequently, the global model integrated by federated learning is used to update the detection zones in real time. The robot cluster replans its work areas and mission objectives based on each robot's latest learning model. For example, if the criticality level of a particular area increases or new vital sign information is provided, task allocation will be automatically adjusted, increasing robot detection efforts in that area or shifting the focus of tasks to areas with greater resource needs. Each robot redistributes its work within the updated detection zone to ensure maximum efficiency while avoiding over-concentration in certain areas. Through the federated learning mechanism, each member of the robot cluster can continuously learn and adapt to the task environment locally and dynamically collaborate with other robots. When the task queue is updated or the on-site conditions change, the robot cluster can respond quickly and optimize resource scheduling to ensure real-time and efficient execution of tasks. The collaboration between robots enables the entire cluster to efficiently complete search and rescue missions in complex and dynamic debris environments while avoiding waste of resources and duplication of tasks. This not only improves rescue efficiency, but also enhances safety and flexibility during the rescue process.

[0040] Furthermore, the present application provides a method of freezing operation instructions in high-risk areas when the arrival of a P wave is detected, and activating a preset obstacle avoidance path to guide the micro-robot cluster to withdraw to a safe area.

[0041] Optionally, when an earthquake occurs, particularly when P waves (first-arrival waves) arrive, sensors can detect initial vibration signals in real time. P waves are the fastest-arriving waves in an earthquake and are typically detected within seconds of the quake. Therefore, they can provide early warning of subsequent strong seismic waves (S waves and surface waves). Within the ruins area, a micro-robot swarm uses its sensors (such as accelerometers and vibration sensors) to monitor ground vibrations in real time. When these vibration sensors detect P-wave vibration signals, they check all currently active tasks, particularly those located in high-risk areas. They then freeze all tasks within these high-risk areas (such as secondary collapses), halting the current tasks and preventing the robots from continuing to operate within these dangerous areas. The robot's obstacle avoidance mechanism is then immediately activated, guiding the robots away from these potentially dangerous areas using pre-set obstacle avoidance paths. Each robot adjusts its trajectory based on a path map, structural stability data, and real-time environmental changes. Through dynamic path planning and obstacle avoidance guidance, the robot swarm follows a safe evacuation route, quickly evacuating high-risk areas and returning to a safe area. The evacuation path takes into account obstacles, damage, and the dynamic conditions of other robot swarms within the ruins, ensuring the robots' safe evacuation and avoiding collisions or entrapment. Although the robot swarm is guided to a safe area after the P wave arrives, the risk level of each area is assessed based on actual conditions. Once aftershocks and other hazards have passed, the robots re-enter the ruins based on the latest structural assessment data and continue the rescue mission. The entire evacuation and recovery process is highly automated, and the robots continuously adjust their behavior and paths based on real-time data to ensure the mission is carried out safely and efficiently. This mechanism not only effectively prevents secondary disasters during an earthquake, but also ensures that the robot swarm is always in the safest working state, reducing unnecessary losses.

[0042] Furthermore, the present application provides a method for loading a pre-earthquake building BIM model in a digital twin environment, performing non-rigid registration evaluation with post-earthquake point cloud data, and identifying the residual bearing capacity attenuation coefficient of key load-bearing components; introducing micro-crack topology data, combined with the residual bearing capacity attenuation coefficient of key load-bearing components, to identify the safe area of ​​the three-dimensional model inside the ruins environment.

[0043] Optionally, a pre-earthquake BIM (Building Information Model) can be loaded. This is a precise 3D digital model of the building, showcasing its structure, materials, layout, and other detailed information. The pre-earthquake BIM model provides the building's structural state under normal conditions, including the design and expected performance of each load-bearing component. After the earthquake, a detailed 3D scan of the post-earthquake ruins is performed using point cloud data. This point cloud data represents the actual ruin structure and can reveal deformation, damage, and collapse of the building after the earthquake. This post-earthquake point cloud data is then compared and analyzed with the pre-earthquake BIM model. This process is called non-rigid registration, which spatially aligns the pre-earthquake BIM model with the post-earthquake point cloud data. Unlike rigid registration, which only considers rotation and translation, non-rigid registration allows for local deformations in the model, such as cracks, looseness, or deformation. These can be captured through non-rigid registration. Registration algorithms are typically based on deformation models, allowing the two different datasets (BIM model and point cloud data) to be matched and compared. During the registration process, the post-earthquake point cloud data is cleaned and denoised to remove irrelevant data or erroneous points. Common features between the point cloud data and the BIM model, such as key structural nodes or geometric features, are then selected for matching. The matching results are optimized using a deformation model to find an optimal match. This process allows the point cloud data to deform in certain areas to align with the pre-earthquake BIM model. After registration, the BIM model is used to identify the building's main load-bearing components (such as load-bearing walls, beams, and columns), which are critical to the stability of the entire structure. Subsequently, the post-earthquake point cloud data and the BIM model are combined to compare and analyze deformation, cracks, and fractures in each load-bearing component to assess its damage level. For example, cracks or deformations in beams and columns can indicate a reduction in their load-bearing capacity. Finite element analysis is then used to calculate the residual bearing capacity of each key load-bearing component. By comparing the post-earthquake damage with the pre-earthquake design bearing capacity, a bearing capacity reduction coefficient is derived for each component. This coefficient represents the proportion of load that the component can still bear after the earthquake. Then, micro-crack topology data was introduced. Micro-crack topology data refers to the detection of cracks in buildings through high-precision sensors (such as laser scanning, digital image processing, or ultrasonic testing), which obtains information such as the distribution, morphology, width, and depth of the cracks. This data provides detailed damage information on the building after the earthquake, especially the microscopic morphology of the cracks, which helps to analyze the degree of damage and safety of the structure in different parts. By combining crack data with the residual bearing capacity attenuation coefficient of key load-bearing components, it is analyzed whether the cracks have penetrated these key parts and affected their bearing capacity. For example, if the cracks in the load-bearing wall are large, it may mean that the bearing capacity of the area has dropped significantly, which in turn affects the stability of the overall structure. Then, using information such as the depth, length, and width of the cracks, the attenuation coefficient of each load-bearing component is corrected. For areas with more serious cracks, the attenuation coefficient will be lower, and vice versa.Then, by combining microscopic crack topology data and residual bearing capacity attenuation coefficients, we can generate safety indicators for each region in the 3D ruins model. The safety of these regions is determined by the following factors: First, areas with dense cracks or large crack extensions indicate that they may have lost their original bearing capacity and are at risk of collapse, thus reducing safety. Second, combined with crack data, the bearing capacity of each load-bearing component is checked to see if it is below a safety threshold. If the bearing capacity of a load-bearing component in a certain area is significantly reduced due to cracks, the safety level is low. After conducting a risk assessment, the crack topology data and bearing capacity analysis results are used to identify safe areas within the ruins. These safe areas are defined as areas where the structure remains stable and has sufficient bearing capacity, based on crack distribution and structural attenuation coefficients. These areas are typically characterized by fewer cracks and no significant reduction in the bearing capacity of key load-bearing components. The identification of safe areas is crucial for rescue operations, as it allows rescuers or robot swarms to avoid dangerous areas and focus their efforts on safer areas. In summary, this process combines the BIM model of the pre-earthquake building with the post-earthquake point cloud data for non-rigid registration and analysis, which can accurately assess the structural damage and bearing capacity changes of the building, and identify safe areas in the ruins, providing a scientific basis for subsequent rescue operations.

[0044] In summary, the embodiments of the present application have at least the following technical effects:

[0045] The embodiment of the present application first obtains point cloud data corresponding to the millimeter-wave radar and thermal imaging data corresponding to the infrared thermal imager based on the target area; then, based on the point cloud data corresponding to the millimeter-wave radar and the thermal imaging data corresponding to the infrared thermal imager, a three-dimensional model of the interior of the ruins environment corresponding to the narrow earthquake space is drawn up, and a passage path map is generated; then, an adaptive network is deployed for the micro-robot cluster, and structural stability assessment and residual vibration risk prediction are performed in combination with the passage path map; then, based on the structural stability assessment results and the residual vibration risk prediction results, the adaptive network is used to allocate the detection partitions and rescue tasks of the micro-robot cluster in real time; finally, vital sign sampling information is introduced to optimize the balanced allocation of the detection partitions and rescue tasks of the micro-robot cluster, and a parallel scheduling strategy is output. These technical effects jointly solve the technical problem of low efficiency and accuracy of rescue in narrow spaces due to complex environments and incomplete perception data, and realize the technical effect of using micro-robot clusters to perform life search and rescue and task allocation in narrow spaces, thereby improving the accuracy and efficiency of rescue in narrow earthquake spaces.

[0046] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0047] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

[0048] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A precise earthquake rescue method in a narrow space using a micro-intelligent robot, characterized by: The method comprises: Based on the target area, obtain the point cloud data corresponding to the millimeter wave radar and the thermal imaging data corresponding to the infrared thermal imager; Based on the point cloud data corresponding to the millimeter wave radar and the thermal imaging data corresponding to the infrared thermal imager, a three-dimensional model of the interior of the ruins environment corresponding to the narrow earthquake space is drawn up, and a traffic path map is generated; Adaptive networking is deployed for micro-robot clusters, and structural stability assessment and residual vibration risk prediction are performed in combination with the aforementioned path map; Based on the structural stability assessment results and the residual vibration risk prediction results, the detection partitions and rescue tasks of the micro-robot cluster are allocated in real time using the adaptive networking; Introducing vital sign sampling information, optimizing the balanced distribution of detection partitions and rescue tasks of the micro-robot cluster, and outputting a parallel scheduling strategy; Wherein, based on the point cloud data corresponding to the millimeter wave radar and the thermal imaging data corresponding to the infrared thermal imager, a three-dimensional model of the interior of the ruins environment corresponding to the narrow earthquake space is drawn up, and the method further includes: Detecting the internal cavity structure of the ruins in the target area and generating a dielectric constant distribution map to identify potential living spaces in the three-dimensional model of the interior of the ruins environment; Based on the potential living space of the three-dimensional model inside the ruins environment, the structural deformation caused by aftershocks is monitored and the dangerous area markings are dynamically updated; Using the updated danger zone markers, precision calibration is performed on the three-dimensional model of the interior of the ruins environment; The method for generating a traffic path map includes: Define the robot's motion capability constraint matrix, including body size, joint mobility, and obstacle crossing quantitative parameters; Based on the robot's motion capability constraint matrix, an initial feasible path is calculated, and the path nodes are embedded in the potential living space of the three-dimensional model inside the ruins environment; The safety and efficiency of the initial feasible path are evaluated to determine the pass path map.

2. The precise earthquake rescue method in a narrow space using a micro-intelligent robot as claimed in claim 1, characterized in that: The vital sign sampling information includes body temperature marker data, micro-vibration marker data and sweat marker data; Based on the micro-vibration marker data, extracting the trapped person's body surface micro-vibration signal and separating the respiratory frequency feature and the heartbeat waveform feature; The state criticality level is evaluated by the respiratory rate characteristics and the heartbeat waveform characteristics, and the state criticality level is used to prioritize the rescue tasks.

3. The precise earthquake rescue method in a narrow space using a micro-sized intelligent robot as claimed in claim 2, characterized in that: The state criticality level is used to prioritize the rescue missions, and the method further includes: Deploy a digital twin engine in the adaptive network to simulate the interaction between the motion trajectory of the micro-robot cluster and the ruin structure in real time; The risk of secondary collapse caused by the operation of the robotic arm is predicted by simulating the interaction force between the motion trajectory of the micro-robot cluster and the ruin structure; Based on the secondary collapse risk, a structural support reinforcement task is generated and added to the rescue task queue.

4. The earthquake narrow space precision rescue method using a micro-sized intelligent robot as claimed in claim 3, characterized in that: When the rescue mission queue is updated, the federated learning mechanism is triggered to integrate the operation data of each robot corresponding to the micro-robot cluster; Through the federated learning mechanism, combined with the state criticality level, the detection partition of the micro-robot cluster is updated online.

5. The precise earthquake rescue method in a narrow space using a micro-sized intelligent robot as claimed in claim 4, characterized in that: When the arrival of the P wave is detected, the operation instructions in the high-risk area are frozen, and the preset obstacle avoidance path is activated to guide the micro-robot cluster to withdraw to the safe area.

6. The precise earthquake rescue method in a narrow space using a micro-sized intelligent robot as claimed in claim 5, characterized in that: Load the pre-earthquake building BIM model in the digital twin environment and perform non-rigid registration evaluation with the post-earthquake point cloud data to identify the residual bearing capacity attenuation coefficient of key load-bearing components; The micro-crack topology data is introduced and combined with the residual bearing capacity attenuation coefficient of key load-bearing components to identify the safe area of ​​the three-dimensional model inside the ruin environment.

Images