A multi-sensor fusion life detection positioning system and positioning method

Abstract

This application relates to the fields of emergency rescue, disaster search and rescue, and life detection and positioning. It discloses a multi-sensor fusion life detection and positioning system, including a sensor array module composed of multiple different types of sensors used to collect environmental data within a target area. It also discloses a multi-sensor fusion life detection and positioning method, including the following steps: S1, deploying a multi-type sensor array in the target area to collect environmental data; S2, performing noise filtering and signal enhancement processing on the raw data collected by the sensors to extract key features of living organisms. This invention solves problems such as single sensor failure, low signal-to-noise ratio misjudgment, positioning errors in complex environments, and static path planning lag by using adaptive multimodal fusion, deep learning feature recognition, nonlinear TDOA optimization, and reinforcement learning path planning. This improves the accuracy of life detection and the efficiency of rescue response, making it suitable for extreme disaster environments.

Description

Technical Field

[0001] This invention relates to the fields of emergency rescue, disaster search and rescue, and life detection and positioning, specifically to a multi-sensor fusion life detection and positioning system and positioning method. Background Technology

[0002] In the fields of emergency rescue and disaster search and rescue, the ability to quickly and accurately detect and locate life is one of the key technologies in rescue operations. After sudden disasters such as earthquakes, mine accidents, and fires, trapped people are often in extremely complex environments, possibly buried under rubble, exposed to high temperatures or dense smoke, or even with limited mobility due to lack of oxygen or injuries. In such situations, traditional manual search methods are inefficient and struggle to locate life in a timely manner, making the use of advanced detection technologies urgently needed to improve rescue efficiency.

[0003] Current life detection technologies typically rely on a single type of sensor, such as infrared thermal imaging sensors, radar detectors, or acoustic sensors. These sensors can detect life signals under certain environmental conditions, but in complex rescue environments, they are often affected by various factors, leading to a decrease in detection accuracy.

[0004] Infrared detection technology relies on the temperature difference between living organisms and the environment, determining the target's location by detecting thermal radiation. However, in fires or high-temperature environments, background heat sources can cause infrared detectors to misinterpret signals, making it difficult to distinguish genuine life signals. Furthermore, infrared detection is easily affected by obstacles, making it difficult to obtain effective life signals when trapped individuals are completely buried.

[0005] Acoustic detection technology uses microphones or ultrasonic sensors to collect sounds from a living being, such as breathing, heartbeat, or knocking. This technology can effectively detect weak life signals in quiet environments, but in noisy rescue scenes (such as those with operating rescue machinery, wind noise, or collapsed buildings), environmental noise can severely interfere with acoustic detection signals. Furthermore, trapped individuals may be unable to actively emit sound signals due to unconsciousness or exhaustion, causing acoustic detection methods to fail in certain situations.

[0006] Radar detection technology identifies the presence of life by detecting electromagnetic waves penetrating obstacles and analyzing signal reflection characteristics. This method is suitable for walls or rubble coverings of a certain thickness, but in complex multipath environments, radar signal reflection and attenuation can lead to false targets or positioning errors. Furthermore, metallic materials can block radar signals, affecting detection accuracy.

[0007] Based on life detection, existing life location methods mainly rely on techniques such as Time Difference of Occurrence (TDOA), Received Signal Strength Indication (RSSI), and triangulation. While these methods can provide relatively accurate location information in open environments, in complex scenarios such as building ruins and mines, signal propagation is affected by obstacles and multipath reflections, leading to significantly increased ranging errors. Furthermore, signal propagation speed can be affected by changes in the environmental medium, such as high humidity, smoke, and air pressure variations, further reducing the stability and reliability of existing positioning technologies.

[0008] Rescue route planning is a crucial component of life detection and positioning systems. Existing route planning methods are mostly based on static maps, such as Dijkstra's algorithm and A* algorithm, which can calculate the theoretical shortest path. However, in real disaster environments, the routes of rescue teams can be affected by environmental changes at any time, such as new collapses, fire spread, or the diffusion of toxic gases, rendering existing routes unusable. Because traditional route planning methods lack dynamic adjustment capabilities, in the event of an emergency, manual replanning of the route is often required, leading to delays and impacting rescue efficiency.

[0009] In summary, existing life detection and positioning technologies still have significant limitations in practical rescue applications. Single-sensor technologies are easily affected by specific environmental factors, impacting the stability of detection; existing positioning methods experience a decrease in accuracy in complex environments, failing to meet the demands of high-precision rescue; and path planning lacks dynamic adjustment capabilities, making it difficult to adapt to rapidly changing rescue site environments. These issues restrict the practicality of life detection and positioning systems in disaster relief, necessitating the development of more adaptable and efficient technical solutions tailored to the specific challenges of complex environments, thereby improving the overall capabilities of detection, positioning, and rescue planning. Summary of the Invention

[0010] To address the shortcomings of existing technologies, this invention provides a multi-sensor fusion life detection and positioning system and method, which solves the problems of existing life detection and positioning systems in complex environments, such as the limitations of single sensors, fixed feature fusion methods, positioning accuracy affected by occlusion, and lack of real-time optimization of path planning.

[0011] To achieve the above objectives, the present invention provides the following technical solution: a multi-sensor fusion life detection and positioning system, comprising:

[0012] The sensor array module consists of multiple sensors of different types, which are used to collect environmental data within the target area;

[0013] The data preprocessing module is used to filter out noise and enhance the signal of the data collected by the sensor, and extract key features of vital signs from it.

[0014] The feature fusion analysis module is used to weight and fuse features from various sensors, use a deep learning model to evaluate whether the signal source is a living organism, the deep learning model including but not limited to convolutional neural networks and recurrent neural networks, and automatically adjust the weights of the fused features;

[0015] The positioning algorithm module is used to calculate the initial position of the living organism based on the time difference ranging data of the signal source and triangulation.

[0016] The result verification module is used to evaluate the reliability of the location results;

[0017] The emergency command center module is used to receive the final location results and update the rescue route plan in real time, guiding the rescue team to quickly locate and carry out rescue operations.

[0018] Preferably, the sensors include, but are not limited to, infrared sensors, acoustic sensors, gas sensors, and vibration sensors.

[0019] Preferably, the data preprocessing module includes:

[0020] A noise filtering unit is used to remove noise from environmental factors, including but not limited to wind speed and electromagnetic interference.

[0021] Signal enhancement unit, which is used to increase the intensity of weak vital signs signals;

[0022] The feature extraction unit is used to extract key features of a living organism from the signal, such as body temperature, respiratory rate, gas concentration, and vibration frequency.

[0023] Preferably, the feature fusion analysis module includes:

[0024] The feature fusion unit is used to perform weighted fusion of features obtained from different sensors, enabling the system to automatically adjust the fusion weights based on the reliability and information quality of different sensors.

[0025] The signal recognition unit is used to compare the fused features with a database of living organisms using deep learning algorithms to determine whether the signal source is a living organism.

[0026] Preferably, the positioning algorithm module includes:

[0027] The TDOA calculation unit is used to estimate the position of the target by using the time difference between sensors;

[0028] A triangulation unit is used to combine the known positions of multiple sensors with the time difference of signal propagation received to optimize positioning accuracy through triangulation.

[0029] Preferably, the TDOA calculation unit uses a time difference ranging formula to estimate the target's position, and the formula is as follows:

[0030]

[0031] in, For the signal to reach the sensor Time difference, The actual location of the target life form. For the first The known locations of the sensors, For signal propagation speed, For sensors The receiving time.

[0032] Preferably, the result verification module includes:

[0033] The credibility assessment unit is used to assess the credibility of the positioning results based on the signal-to-noise ratio and positioning error factors.

[0034] The credibility judgment unit is used to determine whether the credibility is higher than the threshold.

[0035] Preferably, the credibility assessment unit uses a credibility calculation formula to assess the credibility of the positioning result, and the formula is as follows:

[0036]

[0037] in, The signal-to-noise ratio (SNR) of the signal. The actual location of the target life form. The estimated location is obtained through algorithm calculation. It is an adjustment coefficient that controls the impact of position error on reliability.

[0038] Preferably, the emergency command center module includes:

[0039] The path planning unit is used to calculate and optimize the rescue path in real time based on the positioning results;

[0040] The real-time feedback unit is used to adjust the route planning in real time based on the on-site rescue progress data.

[0041] This invention also provides a multi-sensor fusion life detection and localization method, comprising the following steps:

[0042] S1. Deploy multiple types of sensor arrays in the target area to collect environmental data of the target area;

[0043] S2. Perform noise filtering and signal enhancement processing on the raw data collected by the sensor to extract key features of the living organism;

[0044] S3. Weighted fusion of the feature information extracted by each sensor, and analysis of whether the signal source is a living organism using a deep learning model;

[0045] S4. Calculate the initial position of the target life form based on the time difference ranging and triangulation of the signal source;

[0046] S5. Conduct a credibility assessment on the preliminary location calculation results and determine the validity of the results;

[0047] S6. If the credibility is higher than the preset threshold, the location result is confirmed to be valid and sent to the emergency command center to guide the rescue team to carry out the task; if the credibility is low, the monitoring will continue.

[0048] This invention provides a multi-sensor fusion life detection and positioning system and method. It has the following beneficial effects:

[0049] 1. This invention proposes a multimodal sensor data adaptive weight adjustment mechanism. Combining the differentiable feature weight optimization method of deep learning model, it can dynamically adjust the weight of different sensor features in the fusion calculation, so that the system can adapt to signal changes under different disaster environments (such as high temperature, strong noise, low visibility). Compared with the existing fixed weight fusion or simple feature splicing methods, the differentiable feature weight self-optimization strategy of this invention can effectively solve the problems of single sensor signal failure and high false recognition rate under specific environment, thereby improving the robustness and generalization ability of life recognition, and is suitable for large-scale unstructured disaster sites.

[0050] 2. This invention proposes a feature clustering neural network that combines time series and spatial distribution characteristics. By optimizing the clustering distribution of target features through a spatiotemporal constraint loss function, it effectively enhances the ability to identify living organism signals in low signal-to-noise ratio environments. Compared with existing single-frame feature extraction and simple time series matching methods, the cross-time frame feature joint learning mechanism of this invention solves the problem of misidentification caused by short-term signal fluctuations and noise pollution, ensuring that living organisms can still be reliably distinguished from environmental noise even in extremely low signal-to-noise ratio environments (such as buried, mine disasters, and high background noise environments).

[0051] 3. This invention innovatively proposes a TDOA positioning correction algorithm that integrates nonlinear multi-objective optimization. By combining the dynamic correction parameters of the signal propagation model, the system can optimize positioning calculations in complex obstructed environments (such as collapsed buildings and underground spaces). Compared with existing static TDOA or single-path propagation models, the nonlinear multi-objective adaptive optimization algorithm of this invention solves the positioning deviation caused by signal reflection, attenuation, and environmental noise. This enables high-precision positioning of living beings even under non-line-of-sight (NLOS) propagation conditions, providing more reliable location information for disaster relief.

[0052] 4. This invention proposes a dynamic optimization strategy for disaster relief paths based on reinforcement learning (RL). By constructing an environmental feedback reward function, the system can adaptively adjust the rescue path in different disaster sites (such as earthquake ruins, fire spread areas, and mine collapses). Compared with existing fixed path planning or preset weight path optimization methods, the reinforcement learning-based dynamic path optimization mechanism of this invention solves the problem of rescue failure or delay caused by sudden environmental changes (such as increased obstacles or path blockages), ensuring that rescue operations can always be executed along the optimal path under dynamically changing disaster conditions, thereby improving the response speed and success rate of life rescue. Attached Figure Description

[0053] Figure 1 This is a system framework diagram of the present invention;

[0054] Figure 2 This is a flowchart of the system of the present invention;

[0055] Figure 3 This is a framework diagram of the data preprocessing module of the present invention;

[0056] Figure 4 This is a framework diagram of the feature fusion analysis module of the present invention;

[0057] Figure 5 This is a framework diagram of the positioning algorithm module of the present invention;

[0058] Figure 6 This is a framework diagram of the result verification module of the present invention;

[0059] Figure 7 This is a schematic diagram of the emergency command center module of the present invention;

[0060] Figure 8 This is a flowchart of the positioning method of the present invention. Detailed Implementation

[0061] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Please see the appendix Figure 1 -Appendix Figure 2 This invention provides a multi-sensor fusion life detection and positioning system, comprising:

[0063] The sensor array module consists of multiple sensors of different types, which are used to collect environmental data within the target area;

[0064] Sensors include, but are not limited to, infrared sensors, acoustic sensors, gas sensors, and vibration sensors.

[0065] Specifically, the sensor array module of this invention consists of multiple different types of sensors, with the aim of comprehensively collecting environmental data within the target area. Each sensor has different working principles and characteristics, and through the complementarity of multiple sensors, more accurate and comprehensive life detection and location results can be provided. In this embodiment, the sensor array module includes, but is not limited to, infrared sensors, acoustic sensors, gas sensors, and vibration sensors. These sensors can work effectively under different environmental conditions and complement each other, improving the system's detection accuracy and anti-interference capability.

[0066] In this embodiment, the structural design and operation of the sensor array module are as follows:

[0067] Infrared sensors: Infrared sensors are used to detect changes in the body temperature of living organisms. They effectively capture temperature differences, especially in scenarios such as ruins and fires, where the temperature difference between a living organism and its surrounding environment is significant. The signal strength of infrared sensors is affected by distance and environmental interference; therefore, compensation and signal enhancement processing are necessary in complex environments.

[0068] Acoustic sensors: Acoustic sensors are used to monitor physiological signals such as breathing and heartbeats of living organisms. They employ ultrasonic or acoustic wave technology, detecting the presence of living organisms through the reflection and transmission time difference of sound waves between the sensor and the target. Changes in the intensity and frequency of the acoustic signal can provide information about the target's location and activity status. Acoustic signals are often interfered with by environmental noise (such as wind noise and mechanical noise); therefore, noise filtering and signal enhancement techniques are needed to improve signal recognition accuracy.

[0069] Gas sensors: Gas sensors are used to detect the gaseous components emitted by living organisms, such as carbon dioxide and ammonia. Living organisms release specific gaseous components during respiration, and gas sensors can capture these gas molecules in the air and identify the activity of living organisms. Gas sensors determine the presence of living organisms by measuring changes in gas concentration. Gas sensors are easily affected by environmental factors such as airflow and humidity; therefore, data filtering and compensation are required within the system.

[0070] Vibration sensors: Vibration sensors are used to capture subtle vibrations on the ground or target surface caused by the activity of living organisms. In environments such as ruins and mines, the activity of living organisms generates minute vibration signals. Vibration sensors can accurately capture these signals in a sensor array and further analyze them. The signals from vibration sensors are often affected by the surrounding environment (such as wind, vibrations from mechanical equipment, etc.), therefore, signal filtering techniques must be used to reduce false detections.

[0071] In this embodiment, the sensor array module, through the collaborative work between sensors, collects environmental data and then processes different types of sensor signals through the data preprocessing module to remove noise, enhance signals, and extract key features of living organisms.

[0072] Data from each sensor is analyzed through weighted fusion. Deep learning models (such as convolutional neural networks (CNNs) or recurrent neural networks (RNNs) are used to evaluate whether the signal source is a living organism. The weights of the fused features are automatically adjusted to provide high-precision life detection judgment.

[0073] The advantage of sensor array modules lies in the fact that data from multiple types of sensors complement each other, enabling more accurate identification of the presence of living organisms in complex environments, thereby improving the accuracy and reliability of the system.

[0074] Alternatively, the sensor array module in this invention can flexibly select sensor configurations according to the needs of different environmental scenarios. For example, in scenarios such as mines or ruins, a combination of vibration sensors and acoustic sensors can be prioritized, while in high-temperature environments such as fires, the configuration of infrared sensors will be more important. In addition, gas sensors can provide supplementary information based on changes in on-site gas concentration, further optimizing the system's positioning capabilities.

[0075] Specifically, the sensor array module is designed with multi-sensor fusion in mind. The signal characteristics of each sensor are weighted during the data fusion stage, ensuring that the contribution of each sensor is dynamically adjusted based on its data quality. This approach improves the accuracy of life detection while effectively addressing the challenges of environmental interference and complex background noise.

[0076] In one possible implementation, the operating mode and parameter configuration of each sensor can be adjusted in real time through the sensor array module to ensure that the system can maintain high detection accuracy and response speed under different environmental conditions.

[0077] Please see the appendix Figure 3 The data preprocessing module is used to filter out noise and enhance signals from the data collected by the sensors, and to extract key features of vital signs from them.

[0078] The data preprocessing module includes:

[0079] A noise filtering unit is used to remove noise from environmental factors, including but not limited to wind speed and electromagnetic interference.

[0080] Signal enhancement unit, which is used to increase the intensity of weak vital signs signals;

[0081] The feature extraction unit is used to extract key features of a living organism from the signal, such as body temperature, respiratory rate, gas concentration, and vibration frequency.

[0082] Specifically, the data preprocessing module is a key component of this invention, responsible for the preliminary processing of environmental data collected by sensors. Its main tasks include removing environmental noise, enhancing weak vital signs, and extracting key features related to living organisms from the signals. By processing and optimizing data from various sensors, the data preprocessing module ensures that downstream modules (such as the feature fusion analysis module and the localization algorithm module) can obtain clearer and more reliable signal information, thereby improving the accuracy and precision of life detection.

[0083] In this embodiment, the specific composition and operation of the data preprocessing module are as follows:

[0084] The primary task of the noise filtering unit is to remove noise caused by environmental factors. Various environmental interferences, such as changes in wind speed, electromagnetic interference, and weather fluctuations, can affect the sensor's signal, thus impacting the accuracy of vital sign detection. Therefore, the noise filtering unit employs multiple filtering techniques to eliminate these irrelevant signals.

[0085] Typically, noise filtering units use Kalman filtering, a technique that smooths sensor data based on minimizing estimation errors.

[0086] As an alternative, the noise filtering unit can also use wavelet transform, which is particularly suitable for situations where there are multiple frequency components in the signal and can effectively remove high-frequency noise.

[0087] Specifically, the Kalman filter uses a recursive calculation method to continuously refine the estimate of the signal, as shown in the following formula:

[0088]

[0089] in, For the first Predicted value at time, For Kalman gain, For the first The observed value at time, For the observation matrix, This is the updated estimate.

[0090] The primary function of the signal enhancement unit is to amplify the intensity of weak vital sign signals, ensuring that signals with low signal-to-noise ratios can be fully identified. In complex environments, the signals generated by living organisms are typically weak, thus requiring signal enhancement techniques to increase their intensity and ensure accurate analysis by downstream modules.

[0091] In some embodiments, the signal enhancement unit enhances the signal using adaptive filtering technology. Adaptive filtering can dynamically adjust filtering parameters based on changes in real-time acquired data, optimizing the signal enhancement effect.

[0092] As an alternative, the signal enhancement unit can utilize local maximum detection methods to extract sudden changes in biological signals (such as heartbeat and respiration) and further enhance the signal strength of these important features.

[0093] Specifically, the signal enhancement process can be described by the following formula:

[0094]

[0095] in, For the enhanced signal, The original signal, This is the background noise signal. The enhancement factor determines the strength of the signal enhancement.

[0096] The feature extraction unit is used to extract key features of vital signs from the signal after noise filtering and enhancement, including body temperature, respiratory rate, gas concentration, and vibration frequency. These features form the basis for subsequent feature fusion analysis modules to identify and locate living organisms.

[0097] Specifically, body temperature signals usually come from infrared sensors. When extracting body temperature features, the temperature change rate analysis method is used. By calculating the temperature change rate, the heat source changes of living organisms can be accurately identified.

[0098] Respiratory and heart rate characteristics are typically acquired using acoustic sensors, and the signals are analyzed to extract specific frequency bands of heart and respiratory signals. The respiratory and heart rates of a living organism are extracted through Fast Fourier Transform (FFT) processing of the signal spectrum.

[0099] Gas concentration characteristics are obtained through gas sensors, and the extracted features are typically the rate of change of gas concentration. This can be expressed by the following formula:

[0100]

[0101] in, For gas concentration, For gas volume, Let be the detection area of ​​the sensor. For time. By calculating the rate of change of concentration over time, the breathing patterns of living organisms can be extracted.

[0102] Vibration frequency is acquired using a vibration sensor, and feature extraction utilizes frequency domain analysis to identify the main frequency components of the vibration signal. Vibration signal It can be described by the following formula:

[0103]

[0104] in, It is a vibration signal. The initial amplitude of the signal. Let be the angular frequency of the vibration. For phase shift, As the attenuation factor, It is a positive attenuation coefficient. It is a sine function.

[0105] Alternatively, features from multiple sensors can be combined and weighted during feature extraction. Weighted averaging or principal component analysis (PCA) can further reduce signal redundancy between different sensors and improve the accuracy of feature extraction.

[0106] In one possible implementation, during the data preprocessing stage, signals from all sensors can be acquired synchronously and processed in real time by an integrated preprocessing system. This system adaptively adjusts filtering and enhancement parameters based on the quality of the data acquired by the sensors, thereby ensuring that data from each sensor can be processed under optimal conditions.

[0107] Please see the appendix Figure 4The feature fusion analysis module is used to weight and fuse features from various sensors, and use a deep learning model to evaluate whether the signal source is a living organism. The deep learning model includes, but is not limited to, convolutional neural networks and recurrent neural networks, and automatically adjusts the weights of the fused features.

[0108] The feature fusion analysis module includes:

[0109] The feature fusion unit is used to perform weighted fusion of features obtained from different sensors, enabling the system to automatically adjust the fusion weights based on the reliability and information quality of different sensors.

[0110] The signal recognition unit is used to compare the fused features with a database of living organisms using deep learning algorithms to determine whether the signal source is a living organism.

[0111] Specifically, the feature fusion analysis module is one of the core components of this invention. Its main function is to weightedly fuse feature information from different types of sensors and analyze the fused features using a deep learning model to determine whether the signal source is a living organism. The key to this module lies in how to automatically adjust the fusion weights of the features from each sensor, ensuring that the system can adaptively adjust the fusion strategy according to the reliability and information quality of different sensors, thereby improving the accuracy and robustness of life detection.

[0112] In this embodiment, the structure and working principle of the feature fusion analysis module are as follows:

[0113] The task of the feature fusion unit is to perform weighted fusion of features obtained from various sensors. Different types of sensors (such as infrared, acoustic, gas, vibration, etc.) have their own advantages and limitations. The feature fusion unit, through weighting, comprehensively considers the data quality of each sensor, enabling the system to adaptively adjust the fusion weights according to the actual situation.

[0114] Typically, feature fusion employs a weighted average method, calculating a weighted sum of features from each sensor. This weighted sum reflects the importance of information from different sensors, and the weights can be adjusted through the system's real-time evaluation and feedback mechanism.

[0115] As an alternative, the feature fusion unit can employ principal component analysis (PCA) to reduce the dimensionality of the high-dimensional feature space, extract the most important features, reduce the impact of redundant information, and further improve the fusion accuracy.

[0116] Specifically, the following weighted calculation formula can be used in the fusion process:

[0117]

[0118] in, The fused feature vector For the first Features extracted by each sensor For the first The weight of each sensor, Number of sensors. Weights The system can be dynamically adjusted based on the signal quality (such as signal-to-noise ratio) of each sensor to ensure that higher-quality sensors contribute more to the fusion results.

[0119] The task of the signal recognition unit is to analyze the fused features using deep learning algorithms to assess whether the signal source is a living organism. Deep learning models can be trained on large databases of vital signs, enabling the system to learn potential patterns in living organism signals, thus more accurately determining whether the signal source belongs to a living organism.

[0120] In one possible implementation, the signal recognition unit uses a convolutional neural network (CNN) or a recurrent neural network (RNN) to analyze the fused features. CNNs can efficiently process image-type data and are suitable for analyzing spatial features; while RNNs are suitable for processing time-series data and can capture the patterns of signal changes over time.

[0121] Specifically, in a convolutional neural network, after several convolutional and pooling layers, the feature map is gradually reduced and abstracted, and finally, a fully connected layer maps the high-dimensional features into a classification result. The output of the CNN uses the softmax activation function to determine whether the signal source is a living organism.

[0122] As an alternative, the signal recognition unit can also use a Long Short-Term Memory (LSTM) network to process temporal features. LSTM networks are able to process long-distance dependent information in time series and are very effective at capturing changes in signals such as breathing and heartbeat of living organisms.

[0123] In this implementation, the feature fusion analysis module continuously optimizes the weights and feature selection, so that the fused features can reflect the characteristics of living organism signals to the greatest extent, reduce the interference of environmental noise and non-living organism signals, and improve the accuracy of the system's judgment.

[0124] In some embodiments, the feature fusion analysis module, through incremental learning and adaptive adjustment mechanisms, can dynamically adjust the weighting of features based on the real-time environment and sensor operating conditions. This flexible feature fusion method can be optimized for specific needs in different application scenarios, improving the system's adaptability in various complex environments.

[0125] In this embodiment, the module's output not only provides information on the presence or absence of a living organism, but also the contribution of each sensor to the final judgment. This feature allows the system to further improve accuracy by analyzing the contributions of each sensor and optimizing the system's parameter configuration when misjudgments occur.

[0126] As an extension, to further enhance the system's robustness, the feature fusion analysis module can also combine traditional rule matching methods (such as threshold detection) with deep learning methods for hybrid pattern decision-making. That is, based on the deep learning judgment results, threshold determination and rule verification are added to further reduce false positives and false negatives.

[0127] Please see the appendix Figure 5 The positioning algorithm module is used to calculate the initial position of the living organism based on the time difference ranging data of the signal source and the triangulation method;

[0128] The positioning algorithm module includes:

[0129] The TDOA calculation unit is used to estimate the position of the target by using the time difference between sensors;

[0130] A triangulation unit is used to combine the known positions of multiple sensors with the time difference of signal propagation received to optimize positioning accuracy through triangulation.

[0131] The TDOA calculation unit uses a time difference ranging formula to estimate the target's position. The formula is as follows:

[0132]

[0133] in, For the signal to reach the sensor Time difference, The actual location of the target life form. For the first The known locations of the sensors, For signal propagation speed, For sensors The receiving time.

[0134] Specifically, the positioning algorithm module is a key component of this invention, responsible for estimating the initial location of a living being based on Time Difference of Origin (TDOA) data from sensors and triangulation. This module combines multiple positioning technologies to provide high-precision positioning results, effectively identifying the location of living beings in complex environments and supporting subsequent rescue operations. In particular, the combination of TDOA calculation and triangulation enables the system to effectively overcome problems such as signal reflection and obstruction, further improving positioning accuracy and robustness.

[0135] In this embodiment, the specific composition and operation of the positioning algorithm module are as follows:

[0136] The primary task of the TDOA computing unit is to estimate the target's position by measuring the time difference between signals received from multiple sensors. TDOA technology calculates the target's position by measuring the time difference of signal propagation from the living organism to different sensors and combining this with the known positions of the sensors.

[0137] Under normal circumstances, the TDOA calculation unit calculates the initial position of the target life form based on the time difference ranging formula, using the known signal propagation speed and the time difference between the signals received by each sensor.

[0138] Specifically, the TDOA calculation formula is as follows:

[0139]

[0140] in, For the signal to reach the sensor Time difference, The actual location of the target life form is yet to be determined. For the first The known locations of the sensors, The speed of signal propagation can vary depending on the environment (such as air or other media), but is usually a known constant. For sensors The reception time is usually recorded by the sensor when it receives the signal.

[0141] The key to TDOA technology lies in calculating the target's position using time difference. Generally, the signal propagation speed... The location of the target is known in different media (such as air, rubble, etc.), so the time difference between different sensors can be used to deduce the location of the target.

[0142] The role of the triangulation unit is to combine the known positions of multiple sensors and the time difference of signal propagation received to further optimize positioning accuracy through triangulation. Triangulation uses the known positions of the sensors and the time difference of signal propagation between each sensor and the target to calculate the precise coordinates of the target's position.

[0143] Alternatively, if there are more than three sensors in the system, the triangulation unit can optimize the positioning results by combining the results of multiple TDOA calculations and utilizing geometric relationships.

[0144] Specifically, the constraint equations for the target position can be established by utilizing the signal propagation time differences between multiple sensors through the following geometric calculations:

[0145]

[0146] in, For target to sensor distance, For target to sensor Time difference, For signal propagation speed, For sensors The system can calculate the target's precise location using these equations, which determine the reception time.

[0147] As an alternative, triangulation can be combined with least squares to optimize positioning results. Least squares can further improve positioning accuracy by minimizing the sum of squared errors.

[0148] In practical applications, the speed of signal propagation It may be affected by environmental conditions, such as temperature, humidity, or obstruction. Therefore, during the positioning process, the TDOA computing unit needs to adjust and compensate for the signal propagation speed in real time.

[0149] Specifically, under conditions of low ambient temperature or high air density, the signal propagation speed Changes will occur. To compensate for this effect, the following adjustment formula can be used:

[0150]

[0151] in, The adjusted signal propagation speed, The speed of signal propagation under standard conditions. For temperature coefficient, This represents the difference between the ambient temperature and the standard temperature.

[0152] In one possible implementation, the TDOA calculation unit and the triangulation unit are used in combination to ensure high-precision positioning in complex environments. In a multi-sensor configuration, the triangulation unit can combine the calculation results of multiple signal propagation time differences to provide a more accurate target location.

[0153] Please see the appendix Figure 6 The result verification module is used to evaluate the reliability of the positioning results.

[0154] The result verification module includes:

[0155] The credibility assessment unit is used to assess the credibility of the positioning results based on the signal-to-noise ratio and positioning error factors.

[0156] The credibility judgment unit is used to determine whether the credibility is higher than the threshold.

[0157] The credibility assessment unit uses a credibility calculation formula to evaluate the credibility of the positioning results. The formula is as follows:

[0158]

[0159] in, The signal-to-noise ratio (SNR) of the signal. The actual location of the target life form. The estimated location is obtained through algorithm calculation. It is an adjustment coefficient that controls the impact of position error on reliability.

[0160] Specifically, the result verification module is a crucial component of this invention, primarily responsible for evaluating the reliability of the positioning results. This module rigorously assesses the reliability of the positioning results to ensure the accuracy and reliability of the positioning system in complex environments. By evaluating the signal-to-noise ratio (SNR) and positioning error, the result verification module effectively determines the authenticity and reliability of the positioning results and decides whether to accept the results or whether further monitoring is necessary.

[0161] In this embodiment, the specific composition and working principle of the result verification module are as follows:

[0162] The reliability assessment unit's task is to evaluate the reliability of the positioning results based on factors such as the signal-to-noise ratio (SNR) and positioning error. SNR is a crucial indicator for evaluating signal quality; a high SNR indicates a clear signal, small error, and high reliability; conversely, a low SNR indicates poor signal quality and potentially large errors, thus requiring careful verification of the positioning results.

[0163] Generally, the credibility assessment unit uses the following formula to calculate the credibility of the location results:

[0164]

[0165] in, The higher the calculated confidence level, the greater the confidence level. Signal-to-noise ratio (SNR) is a direct measure of signal quality; a higher SNR value indicates better signal quality. The actual location of the target life form. The estimated location is calculated using a positioning algorithm. This is an adjustment coefficient used to control the impact of position error on reliability. When the value is large, the positioning error has a greater impact on reliability.

[0166] Specifically, the reliability assessment unit calculates the relationship between the signal-to-noise ratio (SNR) and the estimated position error to derive a comprehensive reliability index. When the signal quality is good and the estimated position error is small, the reliability is high, indicating a relatively accurate positioning result. Conversely, if the signal quality is poor or the position error is large, the reliability is low, indicating that the positioning result may have a significant deviation and requires further confirmation or repositioning.

[0167] The credibility judgment unit determines the validity of the positioning result based on the credibility value calculated by the credibility assessment unit. This unit judges whether the positioning result is sufficiently reliable based on a preset threshold. If the credibility is higher than the set threshold, the positioning result is considered valid; if it is lower than the threshold, the process returns to the step of continuing monitoring and repositioning.

[0168] Generally, the credibility assessment unit sets a credibility threshold based on historical data or experimental results. ,like If so, the location result is confirmed to be valid; if If the target is detected and monitored, return to the first step.

[0169] Specifically, if the credibility is calculated after credibility assessment Exceeding the preset threshold If the location result is satisfactory, the system considers it usable for subsequent processing (e.g., sending it to the emergency command center for route planning). Otherwise, if the reliability does not meet the requirements, the system will perform further testing and optimize the location result.

[0170] As an alternative, in real-time applications, a credibility threshold The system can be dynamically adjusted through system learning or real-time feedback mechanisms, flexibly adjusting thresholds according to different environments, sensor states, and operating conditions to improve system adaptability.

[0171] In one possible implementation, the confidence assessment unit can also incorporate other auxiliary information (such as environmental changes, sensor status, etc.) to optimize the assessment process, making the assessment more accurate and efficient. For example, when the sensor is interfered with, the confidence assessment unit can appropriately relax the threshold to adapt to the needs of environmental changes.

[0172] Please see the appendix Figure 7 The emergency command center module is used to receive the final location results and update the rescue route plan in real time, guiding the rescue team to quickly locate and carry out rescue operations.

[0173] The emergency command center module includes:

[0174] The path planning unit is used to calculate and optimize the rescue path in real time based on the positioning results;

[0175] The real-time feedback unit is used to adjust the route planning in real time based on the on-site rescue progress data.

[0176] Specifically, the emergency command center module is one of the key modules in this invention. Its main function is to receive the final positioning results provided by the positioning system and update the rescue route planning in real time, ensuring that the rescue team can quickly reach the location of the living person and carry out effective rescue. This module provides dynamic rescue route planning through the cooperation of the route planning unit and the real-time feedback unit, and adjusts the rescue route according to the real-time rescue progress data on site to improve rescue efficiency and personnel safety.

[0177] In this embodiment, the specific composition and operation of the emergency command center module are as follows:

[0178] The primary task of the path planning unit is to calculate and optimize the path planning for the rescue team in real time based on the final positioning results provided by the positioning system, ensuring that the rescue team can reach the location of the target life in the shortest possible time. The path planning unit generates one or more optimal paths for the rescue team by combining various algorithms (such as graph theory algorithms, Dijkstra's algorithm, and A* algorithm).

[0179] In general, the path planning unit performs path calculations based on target location information and existing environmental maps (such as ruin structure diagrams, obstacle distribution at fire scenes, etc.).

[0180] As an alternative, the path planning unit uses the A* algorithm for path search. The A* algorithm combines the advantages of heuristic search and shortest path search, and can find the shortest path from the starting point to the target in complex environments.

[0181] Specifically, in rescue missions, the A* algorithm calculates the cost function of the path. ,in It is the actual cost from the starting point to the current point. It is the estimated cost from the current point to the target point. The total cost is used to select and optimize the path with the minimum cost. This method allows the path planning unit to avoid obstacles and dangerous areas while ensuring the shortest possible path.

[0182] The real-time feedback unit adjusts the route planning in real time based on on-site rescue progress data, ensuring that the rescue team can cope with constantly changing on-site conditions. As the rescue team advances, new obstacles, hazards, or environmental changes may arise. The real-time feedback unit dynamically optimizes the route planning by acquiring on-site rescue progress and real-time environmental data.

[0183] In general, the real-time feedback unit is connected to the on-site sensors and monitoring system to dynamically adjust the path planning by acquiring data such as the team's current location, task progress, and environmental changes in real time.

[0184] As an alternative, the real-time feedback unit can use particle filtering or Kalman filtering techniques to estimate and predict field data in real time in order to cope with rapid changes in the environment.

[0185] Specifically, suppose the path planning unit has generated a preliminary rescue path, but during the rescue operation, obstacles at the scene (such as collapsed walls or fire sources) render the original path infeasible. In this case, the real-time feedback unit adjusts the path planning objective and recalculates a path that the rescue team can safely traverse. For example, the formula for updating the path state using particle filtering is:

[0186]

[0187] in, It is the optimal path estimate at the current moment. This is the latest on-site feedback data. It is the observation matrix. It is a Kalman gain, and the particle filter update method can dynamically adjust the path planning according to changes in the field.

[0188] In this embodiment, the feedback mechanism and dynamic adjustment capability of the real-time feedback unit ensure the flexibility of the rescue operation. As the on-site environment constantly changes, the real-time feedback unit can update the path information in a timely manner, ensuring that the rescue team can bypass new obstacles or dangerous areas, thus guaranteeing the efficiency and safety of the rescue operation.

[0189] As an extension, the collaborative work of the path planning unit and the real-time feedback unit can be achieved through multi-objective optimization. This involves considering not only the shortest path but also factors such as rescue efficiency, team member safety, and resource optimization, further enhancing the system's intelligence. In emergency situations, the system can also allocate tasks among different rescue teams and optimize resource scheduling, making the overall rescue operation more efficient.

[0190] Please see the appendix Figure 8 The multi-sensor fusion life detection and localization method described below can be referred to in correspondence with the multi-sensor fusion life detection and localization system described above.

[0191] A multi-sensor fusion life detection and localization method includes the following steps:

[0192] S1. Deploy multiple types of sensor arrays in the target area to collect environmental data of the target area;

[0193] S2. Perform noise filtering and signal enhancement processing on the raw data collected by the sensor to extract key features of the living organism;

[0194] S3. Weighted fusion of the feature information extracted by each sensor, and analysis of whether the signal source is a living organism using a deep learning model;

[0195] S4. Calculate the initial position of the target life form based on the time difference ranging and triangulation of the signal source;

[0196] S5. Conduct a credibility assessment on the preliminary location calculation results and determine the validity of the results;

[0197] S6. If the credibility is higher than the preset threshold, the location result is confirmed to be valid and sent to the emergency command center to guide the rescue team to carry out the task; if the credibility is low, the monitoring will continue.

[0198] The method in this embodiment can be used to execute the above system embodiment, and its principle and technical effect are similar, so it will not be described again here.

[0199] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-sensor fusion life detection and positioning system, characterized in that, include: The sensor array module consists of multiple sensors of different types, which are used to collect environmental data within the target area; The data preprocessing module is used to filter out noise and enhance the signal of the data collected by the sensor, and extract key features of vital signs from it. The feature fusion analysis module is used to weight and fuse features from various sensors, use a deep learning model to evaluate whether the signal source is a living organism, the deep learning model including but not limited to convolutional neural networks and recurrent neural networks, and automatically adjust the weights of the fused features; The positioning algorithm module is used to calculate the initial position of the living organism based on the time difference ranging data of the signal source and triangulation. The result verification module is used to evaluate the reliability of the location results; The emergency command center module is used to receive the final location results and update the rescue route plan in real time, guiding the rescue team to quickly locate and rescue. The positioning algorithm module includes: The TDOA calculation unit is used to estimate the position of the target by using the time difference between sensors; A triangulation unit is used to combine the known positions of multiple sensors with the propagation time difference of the received signals to optimize positioning accuracy through triangulation. The TDOA calculation unit uses a time difference ranging formula to estimate the target's position. The formula is as follows: ； in, The time difference between the arrival of the signal at the sensor. The actual location of the target life form. Let i be the known position of the i-th sensor. For signal propagation speed, This represents the sensor's reception time.

2. The multi-sensor fusion life detection and positioning system according to claim 1, characterized in that, The sensors include, but are not limited to, infrared sensors, acoustic sensors, gas sensors, and vibration sensors.

3. The multi-sensor fusion life detection and positioning system according to claim 1, characterized in that, The data preprocessing module includes: A noise filtering unit is used to remove noise from environmental factors, including but not limited to wind speed and electromagnetic interference. Signal enhancement unit, which is used to increase the intensity of weak vital signs signals; The feature extraction unit is used to extract key features of a living organism from the signal, such as body temperature, respiratory rate, gas concentration, and vibration frequency.

4. The multi-sensor fusion life detection and positioning system according to claim 1, characterized in that, The feature fusion analysis module includes: The feature fusion unit is used to perform weighted fusion of features obtained from different sensors, enabling the system to automatically adjust the fusion weights based on the reliability and information quality of different sensors. The signal recognition unit is used to compare the fused features with a database of living organisms using deep learning algorithms to determine whether the signal source is a living organism.

5. The multi-sensor fusion life detection and positioning system according to claim 1, characterized in that, The result verification module includes: The credibility assessment unit is used to assess the credibility of the positioning results based on the signal-to-noise ratio and positioning error factors. The credibility judgment unit is used to determine whether the credibility is higher than the threshold.

6. The multi-sensor fusion life detection and positioning system according to claim 5, characterized in that, The credibility assessment unit uses a credibility calculation formula to evaluate the credibility of the positioning results. The formula is as follows: ； in, The signal-to-noise ratio (SNR) of the signal. The actual location of the target life form. The adjustment coefficient controls the impact of position error on reliability.

7. The multi-sensor fusion life detection and positioning system according to claim 1, characterized in that, The emergency command center module includes: The path planning unit is used to calculate and optimize the rescue path in real time based on the positioning results; The real-time feedback unit is used to adjust the route planning in real time based on the on-site rescue progress data.

8. A multi-sensor fusion life detection and localization method, employing the multi-sensor fusion life detection and localization system described in any one of claims 1-7, characterized in that, Includes the following steps: S1. Deploy multiple types of sensor arrays in the target area to collect environmental data of the target area; S2. Perform noise filtering and signal enhancement processing on the raw data collected by the sensor to extract key features of the living organism; S3. Weighted fusion of the feature information extracted by each sensor, and analysis of whether the signal source is a living organism using a deep learning model; S4. Calculate the initial position of the target life form based on the time difference ranging and triangulation of the signal source; S5. Conduct a credibility assessment on the preliminary location calculation results and determine the validity of the results; S6. If the credibility is higher than the preset threshold, the location result is confirmed to be valid and sent to the emergency command center to guide the rescue team to carry out the task; if the credibility is low, the monitoring will continue.

Images