Modular sensor adaptive integration and data processing method and system in complex terrain environment

Abstract

The invention discloses a modular sensor adaptive integration and data processing method and system in a complex terrain environment, and the method comprises the steps: constructing an extensible sensor sensing layer through a standardized interface module, and achieving the plug and play of various environment monitoring sensors; performing an inspection task in a complex terrain area based on a mobile device, and performing fusion processing on real-time data acquired by multiple sensors through an edge calculation layer; according to the environment data characteristics, the terrain complexity and the running state of the mobile device, the inspection behavior strategy of the mobile device is dynamically adjusted, and follow-up data monitoring and processing are carried out; the processed real-time data is uploaded to a cloud collaboration layer for collaborative analysis in a self-adaptive selection communication mode, and fundamental conversion from a static and fixed monitoring mode to a dynamic and adaptive mode is realized through modular sensor integration, mobile inspection, edge intelligent processing and self-adaptive communication; and the coverage rate, the efficiency and the intelligent level of environment monitoring are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a modular sensor adaptive integration and data processing method and system for complex terrain environments. Background Technology

[0002] Currently, with the rapid development of smart cities, ecological monitoring, and precision agriculture, the demand for efficient and accurate environmental monitoring in complex terrain environments (such as mountains, forests, riverbanks, and irregular areas of large industrial parks) is becoming increasingly urgent. However, existing environmental monitoring technologies have revealed many limitations when dealing with such complex scenarios. The current environmental monitoring field faces three core pain points: First, existing monitoring equipment has fixed functions, and sensor modules are mostly dedicated integrated designs, which cannot be flexibly expanded or replaced according to actual needs, resulting in poor equipment versatility and high upgrade costs. Second, in complex terrain environments (such as mountains, forests, and riverbanks), traditional fixed monitoring stations have limited coverage, while manual inspections suffer from low efficiency (average daily inspection distance of less than 5 kilometers), data lag (average transmission delay of more than 30 minutes), and inability to reach dangerous areas. Third, the multi-sensor data fusion processing capability is insufficient. Existing systems mostly adopt a single threshold judgment mode and lack correlation analysis of multimodal environmental data, resulting in an abnormal event false negative rate of over 23%, which reduces overall work efficiency and stability. Summary of the Invention

[0003] To address the aforementioned problems, this invention provides a modular sensor adaptive integration and data processing method and system for complex terrain environments. This addresses the issues mentioned in the background art, such as the poor versatility of dedicated integrated sensor modules, their inability to achieve efficient inspection in complex terrain, and the lack of correlation analysis of multimodal environmental data, leading to a high rate of missed abnormal events and thus reducing overall work efficiency and stability.

[0004] A modular sensor adaptive integration and data processing method for complex terrain environments includes the following steps: An extensible sensor perception layer is built through standardized interface modules, enabling plug-and-play functionality for various environmental monitoring sensors. Based on mobile devices performing inspection tasks in complex terrain areas, the real-time data collected by multiple sensors is fused and processed through the edge computing layer; Based on environmental data characteristics, terrain complexity, and mobile device operating status, the inspection behavior strategy of the mobile device is dynamically adjusted and subsequent data monitoring and processing are carried out. The processed real-time data is uploaded to the cloud collaboration layer for collaborative analysis via an adaptively selected communication method.

[0005] Preferably, the step of constructing an extensible sensor sensing layer through standardized interface modules to achieve plug-and-play functionality for various environmental monitoring sensors includes: The interface base adopts a unified mechanical size and electrical definition, supporting hot-swapping of sensors such as temperature and humidity, gas concentration, particulate matter concentration, and flame detection. The system automatically identifies and registers various types of sensors when they are connected, using a metadata protocol that includes device identifiers, measurement ranges, and calibration parameters.

[0006] Preferably, the inspection task performed by the mobile device in complex terrain areas involves fusing real-time data collected by multiple sensors through an edge computing layer, including: The mobile device initiates its self-test program and loads a preset inspection path. It then obtains the current terrain parameters along the preset inspection path, controls the mobile device to move steadily based on the current terrain parameters, and periodically collects geographic location parameters and environmental data through multiple sensors. The edge computing layer loads the target terrain parameters on the travel path based on the first main visual image. It then determines whether the travel path has reached complex terrain based on the target terrain parameters. If so, it reduces the travel speed of the mobile device and adjusts the sampling frequency of the multiple sensors to 1.5 times the preset sampling frequency. The edge computing layer employs a sliding filter algorithm with a variable window size to denoise the real-time data, and uses an improved Z-score algorithm for dynamic threshold anomaly detection to detect outliers in the processed real-time data. Correlation analysis is performed on abnormal data values ​​to identify abnormal environmental conditions and trigger corresponding audible and visual alarms.

[0007] Preferably, the real-time data is denoised using a sliding filter algorithm with a variable window size at the edge computing layer, including: The standard deviation of each sensor data stream is determined by the edge computing layer based on real-time data, and the volatility of environmental data is assessed based on the standard deviation. When the data fluctuations are stable, the sliding window size is automatically increased and the sampling frequency is reduced. When the data fluctuates violently, the high-precision mode is immediately triggered and the filtering window is dynamically reduced, the sampling frequency is increased to twice the original reference, and multi-sensor data correlation verification is initiated. The improved Z-score algorithm uses a non-fixed dynamic threshold, optimized online based on the data distribution characteristics of historical data windows. Larger values ​​are automatically applied to parameters with concentrated and stable distributions, while smaller values ​​are applied to parameters with significant fluctuations.

[0008] Preferably, during the movement of the mobile device, it also includes: Terrain-aware maneuver control integrates inertial measurement unit (IMU) data with wheel speed and odometer information to estimate the current terrain slope in real time. When the current terrain slope is detected to be continuously increasing and approaching the design limit, the motor control parameters are dynamically adjusted to reduce the travel speed by 50% and increase the output torque; Infrared sensors are used for obstacle detection. When the infrared sensors continuously detect obstacles of a specific pattern, the fuzzy control algorithm determines the passage characteristics based on the current terrain slope and generates an obstacle detour strategy based on the passage characteristics. The mobile device is controlled to perform real-time obstacle avoidance operations through an obstacle avoidance strategy.

[0009] Preferably, the strategy for dynamically adjusting the inspection behavior of the mobile device includes path planning adjustment, sensor sampling frequency adjustment, and data transmission mode adjustment. Specifically, path planning adjustment includes: Generate the globally optimal path based on terrain elevation information, obstacle distribution, and energy consumption constraints; Real-time obstacle avoidance in localized areas is achieved through multi-sensor fusion perception. When an environmental anomaly is detected, a detailed exploration of the area centered on the anomaly point is triggered; The sampling frequency is adjusted as follows: When environmental data fluctuations exceed preset thresholds or when encountering complex terrain, the sampling frequency of key sensors is automatically increased. When the device battery level is lower than a preset level or the environmental data is stable, the sampling frequency of non-core sensors is automatically reduced. The data transmission mode has been adjusted as follows: Real-time assessment of communication link quality, and automatic switching between Wi-Fi and LoRa; Dynamically adjust data transmission strategies based on remaining battery power and data priority. Establish a high-priority transmission channel for urgent and abnormal data.

[0010] Preferably, when an environmental anomaly is detected, a detailed regional exploration centered on the anomaly point is triggered, specifically as follows: When the edge computing layer identifies a local anomaly based on the detected anomaly value, the control mobile device immediately suspends the execution of the preset inspection path; High-density spiral exploration sub-paths are automatically generated with local anomaly points as the center, and dense data sampling and verification of multiple points and parameters are carried out to accurately define the pollution range and diffusion trend. The regional anomaly attributes of local anomalies are determined based on the pollution range and diffusion trend, and the areas where the local anomalies are located are marked based on the regional anomaly attributes. In subsequent inspection tasks, the inspection priority and sampling frequency of the area where the local anomaly point is located are automatically increased, and detailed data is transmitted back in real time through a high-priority link.

[0011] Preferably, the step of uploading the processed real-time data to the cloud collaboration layer for collaborative analysis via an adaptively selected communication method includes: Encapsulate multi-sensor data in JSON format and determine the data upload threshold; Based on the data upload volume threshold, an incremental transmission mechanism is adopted, and an adaptive communication method is selected based on the transmission bandwidth requirements to upload the encapsulated data to the cloud collaboration layer. The cloud-based collaborative layer platform employs a multi-parameter association model built on the random forest algorithm to perform association analysis on the encapsulated data in order to identify and analyze complex environmental anomalies. Early warning work will be carried out based on the identification and analysis results.

[0012] Preferably, the real-time evaluation of communication link quality and automatic switching between Wi-Fi and LoRa includes: The regional field of view attributes of the mobile device in its current moving state are determined based on the first main visual image of the mobile device. The regional field of view attributes include: open area field of view and occluded area field of view. The terrain complexity is determined based on the area's visibility attributes, and the communication link quality is determined based on the terrain complexity and the communication feedback parameters of the mobile device. If the communication link quality is high, switch the communication mode to Wi-Fi Direct; if the communication link quality is low, switch the communication mode to LoRa networking.

[0013] A modular sensor adaptive integration and data processing system for complex terrain environments, the system comprising: The building blocks are used to construct an extensible sensor perception layer through standardized interface modules, enabling plug-and-play functionality for various environmental monitoring sensors; The processing module is used to perform inspection tasks in complex terrain areas based on mobile devices, and to fuse and process real-time data collected by multiple sensors through the edge computing layer. The adjustment module is used to dynamically adjust the inspection behavior strategy of the mobile device and perform subsequent data monitoring and processing based on environmental data characteristics, terrain complexity and mobile device operating status. The collaborative analysis module is used to upload processed real-time data to the cloud collaborative layer for collaborative analysis via an adaptively selected communication method.

[0014] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0015] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0016] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0017] Figure 1 A flowchart illustrating the workflow of a modular sensor adaptive integration and data processing method for complex terrain environments provided by this invention. Figure 2 Another flowchart of the modular sensor adaptive integration and data processing method for complex terrain environments provided by the present invention; Figure 3 This is another flowchart illustrating the modular sensor adaptive integration and data processing method for complex terrain environments provided by the present invention. Figure 4 This is a schematic diagram of the structure of a modular sensor adaptive integration and data processing system for complex terrain environments provided by the present invention. Detailed Implementation

[0018] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0019] A modular sensor adaptive integration and data processing method for complex terrain environments, such as Figure 1 As shown, it includes the following steps: Step S101: Construct an extensible sensor perception layer through standardized interface modules to enable plug-and-play functionality for various environmental monitoring sensors; Step S102: Based on the mobile device, perform inspection tasks in complex terrain areas, and fuse real-time data collected by multiple sensors through the edge computing layer; Step S103: Based on environmental data characteristics, terrain complexity, and mobile device operating status, dynamically adjust the mobile device's inspection behavior strategy and conduct subsequent data monitoring and processing. Step S104: Upload the processed real-time data to the cloud collaboration layer for collaborative analysis via an adaptively selected communication method.

[0020] In this embodiment, a complex terrain area is defined as a natural or artificial environment that meets at least two of the following characteristics: (1) Drastic elevation changes: There are areas with a slope ≥15° along the planned path of the mobile device. (2) Unstructured surface: There are unpaved roads, gravel, or exposed rock layers, etc. (3) High obstacle density, (4) Poor communication conditions; The working principle of the above technical solution is as follows: an scalable sensor perception layer is built through standardized interface modules to enable plug-and-play use of various environmental monitoring sensors; based on the mobile device performing inspection tasks in complex terrain areas, the real-time data collected by multiple sensors is fused and processed through the edge computing layer; the inspection behavior strategy of the mobile device is dynamically adjusted according to environmental data characteristics, terrain complexity, and the operating status of the mobile device, and subsequent data monitoring and processing are carried out; the processed real-time data is uploaded to the cloud collaboration layer for collaborative analysis through an adaptively selected communication method.

[0021] The beneficial effects of the above technical solution are as follows: through modular sensor integration, mobile inspection, edge intelligent processing, and adaptive communication, a fundamental transformation from a static and fixed monitoring mode to a dynamic and adaptive mode is achieved, significantly improving the coverage, efficiency, and intelligence level of environmental monitoring. At the same time, combined with cloud-based remote analysis, abnormal data events are fundamentally analyzed and timely responses are made, improving inspection efficiency, stability, and reliability. This solves the problems mentioned in existing technologies, such as the poor versatility of dedicated integrated sensor modules, their inability to achieve efficient inspection in complex terrain, and the lack of correlation analysis of multimodal environmental data, which leads to a high rate of missed abnormal events and reduces overall work efficiency and stability.

[0022] In one embodiment, the construction of an extensible sensor sensing layer through standardized interface modules to achieve plug-and-play functionality for various environmental monitoring sensors includes: The interface base adopts a unified mechanical size and electrical definition, supporting hot-swapping of sensors such as temperature and humidity, gas concentration, particulate matter concentration, and flame detection. The system automatically identifies and registers various types of sensors when they are connected, using a metadata protocol that includes device identifiers, measurement ranges, and calibration parameters.

[0023] In this embodiment, the standardized interface module follows the following specific specifications for hardware and software interfaces: Mechanical Interface: An 8-pin M12 standard aviation connector is used as the electrical connector, equipped with a dual fixing structure of strong magnetic adsorption and mechanical locking. Electrical Interface: A unified pin assignment is defined, including Pins 1-2: 3.3VDC power supply, Pins 3-4: GND, Pins 5-6: I2C data line, Pin 7: interrupt signal, and Pin 8: device identifier. Software Protocol: Based on I2C communication, a metadata protocol based on JSONLite format is defined. After the sensor is powered on, the main controller reads the device identifier, range, calibration parameters, and data format from its built-in EEPROM via I2C address 0x50.

[0024] The beneficial effects of the above technical solution are as follows: it solves the problem of poor compatibility between sensors from different manufacturers and of different types, realizes plug-and-play hardware and flexible expansion, and significantly reduces system deployment and maintenance costs; furthermore, it automatically identifies sensors through metadata protocols, eliminates the cumbersome manual configuration process, and improves the system's automation level and response speed.

[0025] In one embodiment, the inspection task performed by the mobile device in complex terrain areas involves fusing real-time data collected by multiple sensors through an edge computing layer, including: The mobile device initiates its self-test program and loads a preset inspection path. It then obtains the current terrain parameters along the preset inspection path, controls the mobile device to move steadily based on the current terrain parameters, and periodically collects geographic location parameters and environmental data through multiple sensors. The edge computing layer loads the target terrain parameters on the travel path based on the first main visual image. It then determines whether the travel path has reached complex terrain based on the target terrain parameters. If so, it reduces the travel speed of the mobile device and adjusts the sampling frequency of the multiple sensors to 1.5 times the preset sampling frequency. The edge computing layer employs a sliding filter algorithm with a variable window size to denoise the real-time data, and uses an improved Z-score algorithm for dynamic threshold anomaly detection to detect outliers in the processed real-time data. Correlation analysis is performed on abnormal data values ​​to identify abnormal environmental conditions and trigger corresponding audible and visual alarms.

[0026] The beneficial effects of the above technical solution are as follows: by adaptively adjusting the travel speed and sampling frequency according to the terrain complexity, the energy consumption and motion stability of the equipment are optimized while ensuring data quality; furthermore, by adopting variable window filtering and dynamic threshold anomaly detection, the data processing strategy can be dynamically optimized with environmental fluctuations, which significantly improves the robustness of data denoising and the accuracy of anomaly identification in complex and changing environments, and reduces false alarms and missed alarms.

[0027] In one embodiment, a sliding filter algorithm with a variable window size is used to denoise real-time data through an edge computing layer, including: The standard deviation of each sensor data stream is determined by the edge computing layer based on real-time data, and the volatility of environmental data is assessed based on the standard deviation. When the data fluctuations are stable, the sliding window size is automatically increased and the sampling frequency is reduced. When the data fluctuates violently, the high-precision mode is immediately triggered and the filtering window is dynamically reduced, the sampling frequency is increased to twice the original reference, and multi-sensor data correlation verification is initiated. The improved Z-score algorithm uses a non-fixed dynamic threshold, optimized online based on the data distribution characteristics of historical data windows. Larger values ​​are automatically applied to parameters with concentrated and stable distributions, while smaller values ​​are applied to parameters with significant fluctuations.

[0028] The beneficial effects of the above technical solution are as follows: by dynamically adjusting the filtering window and sampling frequency based on data volatility, an intelligent balance between computing resources and data quality is achieved, saving computing power during stable periods and capturing details during volatile periods; furthermore, the improved Z-score algorithm adopts an online optimized dynamic threshold, overcoming the shortcomings of fixed thresholds in adapting to changing environments, making anomaly detection more accurate, especially suitable for environmental parameters that have normal fluctuations, ensuring the accuracy and high quality of the collected data.

[0029] In one embodiment, such as Figure 2 As shown, during the movement of the mobile device, the following are also included: Step S201: Based on terrain perception, maneuver control integrates inertial measurement unit (IMU) data and wheel speed odometer information to estimate the current terrain slope in real time. Step S202: When it is detected that the current terrain slope continues to increase and approaches the design limit, dynamically adjust the motor control parameters to reduce the travel speed to 50% and increase the output torque; Step S203: Use an infrared sensor to detect obstacles. When the infrared sensor continuously detects obstacles of a specific pattern, the fuzzy control algorithm determines the passage characteristics based on the current terrain slope and generates an obstacle detour strategy based on the passage characteristics. Step S204: Control the mobile device to perform real-time obstacle avoidance operation through obstacle avoidance strategy.

[0030] The beneficial effects of the above technical solution are as follows: by integrating IMU and odometer data to estimate the slope in real time and adjust the motor parameters, the passability, stability and safety of the mobile device on complex terrain are greatly enhanced, preventing overturning or slipping; furthermore, by combining infrared sensors and fuzzy control algorithms to achieve intelligent obstacle avoidance, the mobile device has the ability to navigate autonomously and smoothly in unknown dynamic environments, reducing dependence on preset paths and human intervention.

[0031] In one embodiment, the strategy for dynamically adjusting the inspection behavior of the mobile device includes path planning adjustment, sensor sampling frequency adjustment, and data transmission mode adjustment, wherein path planning adjustment specifically includes: Generate the globally optimal path based on terrain elevation information, obstacle distribution, and energy consumption constraints; Real-time obstacle avoidance in localized areas is achieved through multi-sensor fusion perception. When an environmental anomaly is detected, a detailed exploration of the area centered on the anomaly point is triggered; The sampling frequency is adjusted as follows: When environmental data fluctuations exceed preset thresholds or when encountering complex terrain, the sampling frequency of key sensors is automatically increased. When the device battery level is lower than a preset level or the environmental data is stable, the sampling frequency of non-core sensors is automatically reduced. The data transmission mode has been adjusted as follows: Real-time assessment of communication link quality, and automatic switching between Wi-Fi and LoRa; Dynamically adjust data transmission strategies based on remaining battery power and data priority. Establish a high-priority transmission channel for urgent and abnormal data.

[0032] The beneficial effects of the above technical solution are as follows: it achieves resource optimization and intelligent task scheduling across the entire chain. Path planning adjustment ensures both global efficiency and local security; sampling frequency adjustment dynamically allocates sensing resources based on environmental needs and equipment status, extending equipment battery life; data transmission mode adjustment makes intelligent decisions based on link quality and data value, ensuring the reliable and timely uploading of critical data, and achieving an optimal balance between energy consumption, data value, and communication efficiency overall.

[0033] In this embodiment, a globally optimal path is generated based on terrain elevation information, obstacle distribution, and energy consumption constraints, including: The elevation layer is constructed by outputting the elevation value of each grid based on the terrain elevation information using a digital elevation model. Gradient calculations are performed on the elevation layer to determine the slope value of each grid cell. Multiple movement paths are generated based on the slope value of each grid cell and the preset terrain traversal cost. Determine the static and dynamic obstacle distributions on each movement path based on the obstacle distributions, and determine the current obstacle avoidance cost for each movement path based on the static and dynamic obstacle distributions. Multiple first movement paths with current obstacle avoidance costs greater than or equal to preset avoidance costs are selected, and the strong correlation maintenance parameter of motion kinetic energy and the strong correlation obstacle parameter of motion resistance are determined when the mobile device moves on each first movement path; Based on the strong correlation between kinetic energy and the strong correlation between motion resistance and the obstacle parameter, an energy consumption cost function is constructed for the mobile device when moving along each first movement path. The energy consumption cost index of the mobile device when it moves along each first mobile path is determined based on the energy consumption cost function. Select multiple second movement paths whose energy consumption cost index is less than or equal to a preset index, and obtain the status information of multiple turning nodes on each second movement path; Determine the travel state transition matrix of the mobile device when passing through each turning node based on the turning node status information; Based on the travel state transition matrix, the mobility efficiency impairment factor of the mobile device is determined, and the target second movement path with the smallest mobility efficiency impairment factor is selected as the global optimal path.

[0034] In this embodiment, the preset terrain passage cost is represented as a dimensionless relative weight value, which is calculated as slope value (degree) × slope weight coefficient + surface roughness coefficient. The slope weight coefficient and surface roughness coefficient are pre-calibrated according to the passage capability of the mobile device. In this embodiment, static obstacle distribution refers to a set of fixed-position obstacles (such as rocks and buildings) obtained through a pre-loaded digital map, while dynamic obstacle distribution refers to a set of variable-position obstacles (such as other vehicles and pedestrians) detected in real time by a LiDAR or camera mounted on a mobile device. In this embodiment, the current obstacle avoidance cost is calculated as follows: for each point on the path, calculate its Euclidean distance to the nearest static and dynamic obstacles, and then map it to a cost value through an inverse proportional function. The total avoidance cost of the path is the integral of the cost values ​​of all points on the path. In this embodiment, the motion kinetic energy strong correlation maintenance parameter is represented as the inertial tensor that maintains a normal velocity kinetic energy when the mobile device moves along each first movement path under the influence of terrain factors; In this embodiment, the motion resistance strongly correlated with the obstacle parameter is represented as the directional obstacle parameter of the motion resistance when the mobile device moves along each first movement path under the combined resistance including rolling resistance coefficient, air resistance coefficient and ramp resistance. In this embodiment, the energy consumption cost function takes the following specific form: ; in, Total energy consumption, Let be the mass of the mobile device, and 'a' be the acceleration of the mobile device on each of the first movement paths. For rolling resistance, This is the component of gravity on the slope (related to the slope). Let v be the air resistance and v be the average speed of the mobile device on each first movement path. This function accurately estimates the total energy consumption by integrating the work done by the mobile device to overcome various resistances as it moves along the path. In this embodiment, the energy consumption cost index is expressed as the calculation result of the energy consumption cost function E mentioned above, and its unit is joule (J), which directly represents the energy consumed to travel on each first movement path; In this embodiment, the turning node status information refers to the curvature κ, turning angle Δθ, and velocity v at the point (turning node) where the direction changes significantly in the path; In this embodiment, the travel state transition matrix is ​​a 2x2 matrix used to describe the coupling relationship between the speed and heading angle of the mobile device as it moves from the current path segment to the next path segment. In this embodiment, the mobility efficiency impairment factor is a scalar value, which is calculated by summing the Frobenius norms of all travel state transition matrices on each first travel path. The larger this sum is, the more speed and direction adjustments are required when traveling on the path, and the lower the overall mobility efficiency.

[0035] The beneficial effects of the above technical solution are as follows: By introducing a strong correlation parameter for maintaining kinetic energy and a strong correlation parameter for hindering motion resistance, and constructing an energy consumption cost function, path planning can accurately predict the actual energy consumption on different paths and directly select the path with the lowest energy consumption cost index, which can significantly extend the endurance of the mobile device in a single operation and improve its practicality. Furthermore, by analyzing the state information of the turning nodes on the path and using the travel state transition matrix to quantify the impact of turning behavior on the motion state, the mobility efficiency impairment factor is finally calculated, and the path with the smallest mobility efficiency impairment factor is selected as the global optimal path, which can ensure that the mobile device operates more smoothly, has less mechanical wear, and performs tasks with higher quality and efficiency, further improving stability and practicality.

[0036] In one embodiment, such as Figure 3 As shown, when an environmental anomaly is detected, a detailed exploration of the area centered on the anomaly point is triggered, specifically: Step S301: When the edge computing layer identifies a local anomaly based on the detected anomaly value, the control mobile device immediately suspends the execution of the preset inspection path; Step S302: Automatically generate high-density spiral exploration sub-paths with local anomaly points as the center, and perform intensive data sampling and verification of multiple points and parameters to accurately define the pollution range and diffusion trend; Step S303: Determine the regional anomaly attributes of local anomalies based on the pollution range and diffusion trend, and mark the areas where local anomalies are located based on the regional anomaly attributes. Step S304: In subsequent inspection tasks, automatically increase the inspection priority and sampling frequency of the area where the local anomaly point is located, and transmit detailed data back in real time through the high-priority link.

[0037] In this embodiment, the high-density spiral exploration sub-path is a path that gradually decreases in sampling point density from the inside out, and its generation algorithm is as follows: Centered on the outlier point P(x,y), the path consists of a sequence of Archimedean spiral points. The parametric equations are as follows: ; ; Where a is the initial radius and b is the spiral expansion coefficient, with a value of 0.2. The mobile device sequentially traverses these points from 0 to 4π at fixed angles to generate a high-density spiral exploration subpath.

[0038] The beneficial effects of the above technical solution are as follows: by pausing the preset path and performing spiral exploration, high-precision positioning of abnormal areas is achieved, which can quickly and accurately characterize the pollution range and trend; furthermore, by marking abnormal areas and increasing their subsequent priority, an intelligent memory mechanism of one-time discovery and long-term monitoring is established, which enables the monitoring system to have the ability to continuously learn and focus on key monitoring, and greatly improves the long-term regulatory effectiveness of potential risk sources.

[0039] In one embodiment, uploading the processed real-time data to the cloud collaboration layer for collaborative analysis via an adaptively selected communication method includes: Encapsulate multi-sensor data in JSON format and determine the data upload threshold; Based on the data upload volume threshold, an incremental transmission mechanism is adopted, and an adaptive communication method is selected based on the transmission bandwidth requirements to upload the encapsulated data to the cloud collaboration layer. The cloud-based collaborative layer platform employs a multi-parameter association model built on the random forest algorithm to perform association analysis on the encapsulated data in order to identify and analyze complex environmental anomalies. Early warning work will be carried out based on the identification and analysis results.

[0040] The beneficial effects of the above technical solution are as follows: First, by adopting JSON encapsulation and incremental transmission mechanism, data transmission efficiency is optimized and valuable wireless communication bandwidth is saved. Second, by using a multi-parameter correlation model based on random forest for cloud analysis, it is possible to discover complex and correlated environmental anomalies that cannot be identified by a single sensor, realizing the upgrade from single-point alarm to situational awareness and improving the comprehensive judgment capability of macro-environmental risks.

[0041] In one embodiment, the real-time assessment of communication link quality and automatic switching between Wi-Fi and LoRa includes: The regional field of view attributes of the mobile device in its current moving state are determined based on the first main visual image of the mobile device. The regional field of view attributes include: open area field of view and occluded area field of view. The terrain complexity is determined based on the area's visibility attributes, and the communication link quality is determined based on the terrain complexity and the communication feedback parameters of the mobile device. If the communication link quality is high, switch the communication mode to Wi-Fi Direct; if the communication link quality is low, switch the communication mode to LoRa networking.

[0042] The beneficial effects of the above technical solution are as follows: by combining the analysis of terrain complexity based on visual perception with actual communication feedback to comprehensively evaluate the link quality, the selection of communication mode becomes more intelligent and closer to the actual scenario, avoiding communication interruptions caused by blindly trying Wi-Fi in areas with signal obstruction, and ensuring the overall reliability of the data backhaul link.

[0043] In one embodiment, this embodiment also discloses a modular sensor adaptive integration and data processing system for complex terrain environments, such as... Figure 4 As shown, the system includes: Module 401 is used to build an extensible sensor perception layer through standardized interface modules, enabling plug-and-play functionality for various environmental monitoring sensors; Processing module 402 is used to perform inspection tasks in complex terrain areas based on mobile devices, and to fuse and process real-time data collected by multiple sensors through the edge computing layer. The adjustment module 403 is used to dynamically adjust the inspection behavior strategy of the mobile device and perform subsequent data monitoring and processing based on environmental data characteristics, terrain complexity and mobile device operating status. The collaborative analysis module 404 is used to upload the processed real-time data to the cloud collaborative layer for collaborative analysis through an adaptively selected communication method.

[0044] The working principle and beneficial effects of the above technical solution have been explained in the method embodiments, and will not be repeated here.

[0045] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.

[0046] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A modular sensor adaptive integration and data processing method for complex terrain environments, characterized in that, Includes the following steps: An extensible sensor sensing layer is built through standardized interface modules, enabling plug-and-play functionality for various environmental monitoring sensors. Based on mobile devices performing inspection tasks in complex terrain areas, the real-time data collected by multiple sensors is fused and processed through the edge computing layer; Based on environmental data characteristics, terrain complexity, and mobile device operating status, the inspection behavior strategy of the mobile device is dynamically adjusted and subsequent data monitoring and processing are carried out. The processed real-time data is uploaded to the cloud collaboration layer for collaborative analysis via an adaptively selected communication method.

2. The modular sensor adaptive integration and data processing method for complex terrain environments according to claim 1, characterized in that, The aforementioned method constructs an extensible sensor sensing layer through standardized interface modules, enabling plug-and-play functionality for various environmental monitoring sensors, including: It adopts a standardized interface base with uniform mechanical dimensions and electrical definitions, supporting hot-swapping of sensors such as temperature and humidity, gas concentration, particulate matter concentration, and flame detection. The system automatically identifies and registers various types of sensors upon connection using a metadata protocol that includes device identifiers, measurement ranges, and calibration parameters.

3. The modular sensor adaptive integration and data processing method for complex terrain environments according to claim 1, characterized in that, The inspection task based on mobile devices in complex terrain areas involves fusing real-time data collected by multiple sensors through an edge computing layer, including: The mobile device initiates its self-test program and loads a preset inspection path. It then obtains the current terrain parameters along the preset inspection path, controls the mobile device to move steadily based on the current terrain parameters, and periodically collects geographic location parameters and environmental data through multiple sensors. The edge computing layer loads the target terrain parameters on the travel path based on the first main visual image. It then determines whether the travel path has reached complex terrain based on the target terrain parameters. If so, it reduces the travel speed of the mobile device and adjusts the sampling frequency of the multiple sensors to 1.5 times the preset sampling frequency. The edge computing layer employs a sliding filter algorithm with a variable window size to denoise the real-time data, and uses an improved Z-score algorithm for dynamic threshold anomaly detection to detect outliers in the processed real-time data. Correlation analysis is performed on abnormal data values ​​to identify abnormal environmental conditions and trigger corresponding audible and visual alarms.

4. The modular sensor adaptive integration and data processing method for complex terrain environments according to claim 3, characterized in that, The edge computing layer employs a sliding filter algorithm with a variable window size to denoise real-time data, including: The standard deviation of each sensor data stream is determined by the edge computing layer based on real-time data, and the environmental data volatility is assessed based on the standard deviation. When the data fluctuations are stable, the sliding window size is automatically increased and the sampling frequency is reduced. When the data fluctuates violently, the high-precision mode is immediately triggered and the filtering window is dynamically reduced, the sampling frequency is increased to twice the original reference, and multi-sensor data correlation verification is initiated. The improved Z-score algorithm uses a non-fixed dynamic threshold, optimized online based on the data distribution characteristics of historical data windows. Larger values ​​are automatically applied to parameters with concentrated and stable distributions, while smaller values ​​are applied to parameters with significant fluctuations.

5. The modular sensor adaptive integration and data processing method for complex terrain environments according to claim 1, characterized in that, During the movement of the mobile device, it also includes: Terrain-aware maneuver control integrates inertial measurement unit (IMU) data with wheel speed and odometer information to estimate the current terrain slope in real time. When the current terrain slope is detected to be continuously increasing and approaching the design limit, the motor control parameters are dynamically adjusted to reduce the travel speed by 50% and increase the output torque; Infrared sensors are used for obstacle detection. When the infrared sensors continuously detect obstacles of a specific pattern, the fuzzy control algorithm determines the passage characteristics based on the current terrain slope and generates an obstacle detour strategy based on the passage characteristics. The mobile device is controlled to perform real-time obstacle avoidance operations through an obstacle avoidance strategy.

6. The modular sensor adaptive integration and data processing method for complex terrain environments according to claim 1, characterized in that, The strategy for dynamically adjusting the inspection behavior of mobile devices includes path planning adjustment, sensor sampling frequency adjustment, and data transmission mode adjustment. Specifically, path planning adjustment involves: Generate the globally optimal path based on terrain elevation information, obstacle distribution, and energy consumption constraints; Real-time obstacle avoidance in localized areas is achieved through multi-sensor fusion perception. When an environmental anomaly is detected, a detailed exploration of the area centered on the anomaly point is triggered; The sampling frequency is adjusted as follows: When environmental data fluctuations exceed preset thresholds or when encountering complex terrain, the sampling frequency of key sensors is automatically increased. When the device battery level is lower than a preset level or the environmental data is stable, the sampling frequency of non-core sensors is automatically reduced. The data transmission mode has been adjusted as follows: Real-time assessment of communication link quality, and automatic switching between Wi-Fi and LoRa; Dynamically adjust data transmission strategies based on remaining battery power and data priority. Establish a high-priority transmission channel for urgent and abnormal data.

7. The modular sensor adaptive integration and data processing method for complex terrain environments according to claim 6, characterized in that, When an environmental anomaly is detected, a detailed exploration of the area centered on the anomaly point is triggered, specifically: When the edge computing layer identifies a local anomaly based on the detected anomaly value, the control mobile device immediately suspends the execution of the preset inspection path; High-density spiral exploration sub-paths are automatically generated with local anomaly points as the center, and dense data sampling and verification of multiple points and parameters are carried out to accurately define the pollution range and diffusion trend. The regional anomaly attributes of local anomalies are determined based on the pollution range and diffusion trend, and the areas where the local anomalies are located are marked based on the regional anomaly attributes. In subsequent inspection tasks, the inspection priority and sampling frequency of the area where the local anomaly point is located are automatically increased, and detailed data is transmitted back in real time through a high-priority link.

8. The modular sensor adaptive integration and data processing method for complex terrain environments according to claim 1, characterized in that, The process of uploading the processed real-time data to the cloud collaboration layer for collaborative analysis via an adaptively selected communication method includes: Encapsulate multi-sensor data in JSON format and determine the data upload threshold; Based on the data upload volume threshold, an incremental transmission mechanism is adopted, and an adaptive communication method is selected based on the transmission bandwidth requirements to upload the encapsulated data to the cloud collaboration layer. The cloud-based collaborative layer platform employs a multi-parameter association model built on the random forest algorithm to perform association analysis on the encapsulated data in order to identify and analyze complex environmental anomalies. Early warning work will be carried out based on the identification and analysis results.

9. The modular sensor adaptive integration and data processing method for complex terrain environments according to claim 6, characterized in that, The real-time assessment of communication link quality and automatic switching between Wi-Fi and LoRa include: The regional field of view attributes of the mobile device in its current moving state are determined based on the first main visual image of the mobile device. The regional field of view attributes include: open area field of view and occluded area field of view. The terrain complexity is determined based on the area's visibility attributes, and the communication link quality is determined based on the terrain complexity and the communication feedback parameters of the mobile device. If the communication link quality is high, switch the communication mode to Wi-Fi Direct; if the communication link quality is low, switch the communication mode to LoRa networking.

10. A modular sensor adaptive integration and data processing system for complex terrain environments, characterized in that, The system includes: The building blocks are used to construct an extensible sensor perception layer through standardized interface modules, enabling plug-and-play functionality for various environmental monitoring sensors; The processing module is used to perform inspection tasks in complex terrain areas based on mobile devices, and to fuse and process real-time data collected by multiple sensors through the edge computing layer. The adjustment module is used to dynamically adjust the inspection behavior strategy of the mobile device and perform subsequent data monitoring and processing based on environmental data characteristics, terrain complexity and mobile device operating status. The collaborative analysis module is used to upload processed real-time data to the cloud collaborative layer for collaborative analysis via an adaptively selected communication method.

Images