Remote monitoring system for multi-sensor data fusion of unmanned aerial vehicle

Abstract

The invention discloses a remote monitoring system for multi-sensor data fusion of an unmanned aerial vehicle, and relates to the technical field of unmanned aerial vehicles, and the system comprises the following components: a multi-sensor data fusion technology, a cross-platform cooperative monitoring mechanism, data security and authority management, and intelligent analysis and decision support. By integrating a multi-sensor data fusion technology, automatic identification and deep fusion of data from different sensors are realized, and the perception capability and decision support capability of the monitoring system are remarkably improved, so that the system can more quickly and accurately capture the state change of a monitored object, and meanwhile, the accuracy of the state change of the monitored object is improved. The intelligent analysis and decision support module analyzes newly generated data in real time, automatically identifies an abnormal event according to a preset anomaly detection rule, and triggers alarm to record related information, so that the response time of the abnormal event is greatly shortened, and the real-time performance of the monitoring system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, and in particular to a remote monitoring system for unmanned aerial vehicle multi-sensor data fusion. Background Art

[0002] With the rapid development of drone technology, drones are increasingly used in various fields, including environmental monitoring, target tracking, and disaster assessment. The multi-sensor systems carried by drones can collect large amounts of multi-dimensional data, which are of great significance for improving the accuracy and real-time performance of monitoring.

[0003] Traditional drone monitoring systems have many shortcomings in intelligent analysis and decision support. They are unable to define abnormal standards, perform real-time analysis on newly generated data according to anomaly detection rules, cannot automatically identify anomalies, and cannot trigger alarms and record relevant information when anomalies are detected. Therefore, a remote monitoring system based on drone multi-sensor data fusion was developed. Summary of the invention

[0004] The purpose of the present invention is to make up for the shortcomings of the prior art and provide a remote monitoring system for unmanned aerial vehicle multi-sensor data fusion. The system can automatically identify potential risks and provide decision makers with scientific basis and data analysis supported suggestions and options by defining abnormal standards, analyzing new data in real time, triggering alarms and recording relevant information, deeply analyzing abnormal events, establishing an abnormal classification system, analyzing risk-related factors and their causal correlations, selecting and optimizing prediction model steps.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a remote monitoring system for multi-sensor data fusion of unmanned aerial vehicles, which includes multi-sensor data fusion technology, cross-platform collaborative monitoring mechanism, data security and authority management, and intelligent analysis and decision support; The multi-sensor data fusion technology: has a built-in efficient data fusion engine that can automatically identify and match data from different sensors, deeply process and complement the data through algorithm optimization, and improve the perception and decision support capabilities of the monitoring system; The cross-platform collaborative monitoring mechanism: adopts data interface and communication protocol, through which each platform can share monitoring data, task instructions and analysis results, forming a widely covered, efficient and collaborative monitoring network; Data security and rights management: Built-in encryption technology to encrypt and protect important data during transmission. At the same time, a strict rights management mechanism is implemented to ensure data security and privacy; The intelligent analysis and decision support described herein: obtain relevant data from multiple data sources, pre-process missing and erroneous data, convert and unify data of different formats and structures, establish associations between data, remove duplicate data, and process noisy data. Statistical methods or judgments based on domain knowledge may be used to select algorithms based on data characteristics and analysis objectives, identify attributes that can reflect the essential characteristics of the data, and principal component analysis methods may be used to reduce data dimensions while retaining important information. Part of the data may be used for model training, model parameters may be adjusted to optimize performance, and a validation set may be used for model evaluation. Based on historical data and business knowledge, define the standards for anomalies, perform real-time analysis on newly generated data according to anomaly detection rules, and when an anomaly is detected, trigger an alarm and record relevant information. Conduct in-depth analysis on the identified abnormal events to determine the cause and type of the anomaly. Establish an anomaly classification system, extract risk-related factors from the data, analyze the causal correlation between these factors and risks, select a prediction model, evaluate the prediction model using historical data, and output the prediction results. Based on the model's prediction results, automatically identify potential risks, adjust and optimize the model based on the evaluation results, and improve prediction accuracy.

[0006] Furthermore, the multi-sensor data fusion technology automatically identifies and matches data from different sensors.

[0007] Furthermore, the multi-sensor data fusion technology algorithm optimizes the in-depth processing and complementation of data.

[0008] Furthermore, each platform of the cross-platform collaborative monitoring mechanism can share monitoring data, task instructions and analysis results to form a monitoring network with wide coverage and efficient collaboration.

[0009] Furthermore, the data security and rights management uses encryption technology to encrypt and protect important data during transmission.

[0010] Furthermore, the intelligent analysis and decision support selects algorithms based on data characteristics and analysis objectives to identify attributes that can reflect the essential characteristics of the data: It is a data point and In Properties The distance metric on is a tuning parameter used to control the influence of distance. It is a data point The comprehensive feature score of Define the abnormal standards, analyze the newly generated data in real time according to the abnormality detection rules, and trigger an alarm and record relevant information when an abnormality is detected: ,in, The newly generated data points we want to detect anomalies for, The mean of historical data, The standard deviation of historical data, This part represents the ratio of the new data point's deviation from the mean to the dispersion of the historical data. Risk-related factors Median, IQR Yes Factor The interquartile range, when Threshold, judgment data For abnormal data, an alarm is triggered and relevant information is recorded.

[0011] Furthermore, the intelligent analysis and decision support uses a prediction model to predict risk relationships: ,in, is a positive number, It indicates the absolute difference between the predicted value and the threshold. The larger the difference, the higher the potential risk. Normalize the differences so that they are scaled relative to the threshold, making it easier to compare predictions of different dimensions. is an optional weighting factor used to adjust the impact of the difference on risk. is the risk factor, which is used to adjust the sensitivity of the risk level according to business needs. The VaR is always non-negative because risk is not usually expressed as a negative value.

[0012] Furthermore, the intelligent analysis and decision support organizes data analysis results, abnormal events and potential risk information into clear and intuitive reports. The reports include charts, data tables and text descriptions. Based on the analysis results and risk predictions, specific suggestions and options are provided to decision makers. The suggestions should be based on scientific basis and data analysis, while taking into account the actual business situation, and building a friendly interactive interface so that decision makers can easily view analysis results, adjust parameters and obtain more information. Visualization tools are provided to enhance decision makers' understanding and insight into data.

[0013] Compared with the existing technology, the remote monitoring system of UAV multi-sensor data fusion has the following beneficial effects: 1. The system integrates multi-sensor data fusion technology to realize automatic recognition and deep fusion of data from different sensors. This feature significantly improves the perception and decision-making support capabilities of the monitoring system, enabling the system to capture the state changes of the monitored objects more quickly and accurately. At the same time, the intelligent analysis and decision support module automatically identifies abnormal events according to preset anomaly detection rules through real-time analysis of newly generated data, and triggers alarms to record relevant information, which greatly shortens the response time of abnormal events and improves the real-time performance of the monitoring system.

[0014] Second, the system uses built-in encryption technology to encrypt and protect important data during transmission, effectively preventing the risk of data leakage and illegal access. At the same time, the system also implements a strict permission management mechanism to authenticate and assign permissions to users who access data, ensuring that only users with corresponding permissions can access relevant data. This dual protection mechanism greatly improves the security and privacy of data and provides users with more reliable monitoring services. Today, when data is becoming an increasingly important asset, this improvement in data security and privacy protection has undoubtedly brought greater value to users. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0016] Figure 1 Operational flow chart of the remote monitoring system based on multi-sensor data fusion of UAV. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Embodiment 1 The multi-sensor system carried by the drone starts working, collecting data from different sensors. The built-in efficient data fusion engine starts, automatically detecting and identifying the source and type of these data. Through predefined rules or algorithms, the data fusion engine matches relevant data from different sensors to ensure the relevance and consistency of the data, and conducts a preliminary check on the identified data, including processing missing values, error values ​​and outliers, converting and unifying data of different formats and structures to facilitate subsequent data processing and analysis, and using algorithms to deeply fuse the pre-processed data. These algorithms may include statistical methods, machine learning algorithms, or reasoning methods based on domain knowledge. Through algorithm optimization, deep fusion and complementarity of data are achieved, and the integrity and accuracy of data are improved. During the fusion process, the algorithm parameters may be dynamically adjusted according to the characteristics of the data to achieve the best fusion effect and establish relationships between data. It can also be used to identify the attributes that can reflect the essential characteristics of the data, provide a basis for subsequent data analysis and decision support, and use part of the data to train the model of the fused data, evaluate the quality of the data and the performance of the model, and adjust and optimize the data fusion algorithm and model parameters according to the evaluation results to improve the effect of data fusion and the overall performance of the monitoring system, and enhance the perception and decision support capabilities of the monitoring system. The data that has been deeply fused and optimized can more accurately reflect the status and changes of the monitored objects. These data provide a richer source of information for the monitoring system and enhance the perception capabilities of the monitoring system. At the same time, based on these data, the monitoring system can make more scientific and accurate decision support and improve monitoring efficiency and accuracy.

[0019] A set of standard data interfaces and communication protocols suitable for multi-platform interaction is adopted. Each platform uploads its own monitoring data to the shared platform according to the standardized data format and interface specifications. The shared platform is responsible for receiving and integrating data from various platforms to form a unified monitoring data warehouse. The monitoring center or the main control platform generates task instructions according to monitoring needs and strategies, and sends the instructions to each relevant platform through a standardized communication protocol. After receiving the task instructions, each platform parses and executes the instruction content. After each platform analyzes and processes the collected data, it uploads the analysis results to the shared platform in a standardized format. The shared platform summarizes and organizes the analysis results of each platform to form a comprehensive monitoring analysis report and displays it to decision makers or users through a friendly interface. During the whole process, the platforms maintain real-time communication and collaboration. By establishing an effective feedback mechanism, the monitoring strategies and instructions are adjusted and optimized in time to improve the efficiency and accuracy of the entire monitoring network. With the increase of monitoring tasks and the expansion of monitoring needs, new platforms can be continuously added to the monitoring network. The newly added platforms only need to follow the existing data interface and communication protocol specifications to quickly integrate into and participate in the collaborative work of the entire monitoring network.

[0020] The system needs to identify which data is sensitive. These sensitive data need special protection during transmission. According to the importance and security requirements of the data, the selection of encryption algorithms should take into account encryption strength, computing efficiency and compatibility factors. Before data transmission, sensitive data is encrypted using the selected encryption algorithm. The encryption process can ensure that even if the data is intercepted during transmission, it cannot be easily decrypted by unauthorized personnel. The encrypted data is transmitted through the network or other communication channels. During the transmission process, even if the data is intercepted, the original information cannot be directly obtained because it has been encrypted. The system needs to authenticate the user who accesses the data. The user identity is verified by multiple methods such as username, password, and biometrics. According to the user's role and responsibilities, the corresponding data access rights are assigned. During the data access process, the system needs to monitor the user's operation behavior in real time and perform access control according to the user's authority. If the user's operation exceeds the scope of his authority, the system will immediately block the operation and record relevant logs for audit. The system needs to record the operation logs of all users. These logs can be used for subsequent audits and tracking to ensure that data access and use comply with regulations.

[0021] Obtain relevant data from multiple data sources, pre-process missing and erroneous data, convert and unify data of different formats and structures, establish associations between data, remove duplicate data, and process noisy data. Statistical methods or judgments based on domain knowledge can be used to select algorithms based on data characteristics and analysis goals to identify attributes that can reflect the essential characteristics of the data. It is a data point and In Properties The distance metric on is a tuning parameter used to control the influence of distance. It is a data point The comprehensive feature score can be reduced by principal component analysis while retaining important information, using part of the data for model training, adjusting model parameters to optimize performance, using the validation set for model evaluation, defining the standard for anomalies based on historical data and business knowledge, and performing real-time analysis on newly generated data according to anomaly detection rules. When an anomaly is detected, an alarm is triggered and relevant information is recorded. An in-depth analysis is performed on the identified abnormal events to determine the cause and type of the anomaly. An anomaly classification system can be established to extract risk-related factors from the data, analyze the causal correlation between these factors and risks, select a prediction model to predict risk relationships, use historical data to evaluate the prediction model, calculate the prediction error index, adjust and optimize the model based on the evaluation results, and improve the prediction accuracy. ,in, is a positive number, It indicates the absolute difference between the predicted value and the threshold. The larger the difference, the higher the potential risk. Normalize the differences so that they are scaled relative to the threshold, making it easier to compare predictions of different dimensions. is an optional weighting factor used to adjust the impact of the difference on risk. is the risk factor, which is used to adjust the sensitivity of the risk level according to business needs. The risk value is always non-negative because risk is usually not expressed as a negative value. The model is adjusted and optimized based on the evaluation results to improve the prediction accuracy.

[0022] Embodiment 2 Soil moisture sensors, meteorological sensors, and visible light and infrared cameras carried by drones are evenly deployed in the farmland. These sensors can collect multi-dimensional data on the farmland environment in real time. The system has a built-in efficient data fusion engine that can automatically identify and match data from different sensors, and achieve deep data fusion and complementarity through advanced algorithm optimization.

[0023] By establishing a unified data interface and communication protocol with meteorological stations, irrigation systems, water resource management centers, and farmers' smartphones or tablet platforms, each platform can share monitoring data, irrigation needs, real-time weather information, and irrigation instructions in real time, forming a widely covered, efficient and coordinated monitoring network. Farmers can view the real-time status of farmland and receive irrigation reminders through mobile phone APP.

[0024] Encrypt sensitive data during transmission to ensure the security and privacy of data during transmission, implement a strict permission management mechanism, set different access rights according to user roles, and ensure safe access and use of data.

[0025] The collected data is preprocessed to ensure the accuracy and availability of the data. An irrigation decision model is constructed based on soil moisture, meteorological data, and crop growth models. The model is trained using historical data and model parameters are adjusted to optimize performance. The system can automatically calculate the optimal irrigation time and amount. When it detects that the soil moisture is lower than the crop water requirement threshold, the irrigation instruction is automatically triggered. At the same time, the system can generate an irrigation effect evaluation report to provide data support for future irrigation strategies. The system can monitor farmland environmental data in real time. When an abnormality is found, ,in, is a positive number, It indicates the absolute difference between the predicted value and the threshold. The larger the difference, the higher the potential risk. Normalize the differences so that they are scaled relative to the threshold, making it easier to compare predictions of different dimensions. is an optional weighting factor used to adjust the impact of the difference on risk. is the risk factor, which is used to adjust the sensitivity of the risk level according to business needs. The risk value is always non-negative because risk is usually not expressed as a negative value. Alarms are triggered in a timely manner and relevant information is recorded. By deeply analyzing the causes and types of abnormal events, farmers can be provided with response suggestions.

[0026] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. The remote monitoring system of UAV multi-sensor data fusion is characterized by: The system includes multi-sensor data fusion technology, cross-platform collaborative monitoring mechanism, data security and authority management, intelligent analysis and decision support; The multi-sensor data fusion technology: has a built-in efficient data fusion engine that can automatically identify and match data from different sensors, deeply process and complement the data through algorithm optimization, and improve the perception and decision support capabilities of the monitoring system; The cross-platform collaborative monitoring mechanism: adopts data interface and communication protocol, through which each platform can share monitoring data, task instructions and analysis results, forming a widely covered, efficient and collaborative monitoring network; Data security and rights management: Built-in encryption technology to encrypt and protect important data during transmission. At the same time, a strict rights management mechanism is implemented to ensure data security and privacy; The intelligent analysis and decision support described herein: obtain relevant data from multiple data sources, pre-process missing and erroneous data, convert and unify data of different formats and structures, establish associations between data, remove duplicate data, and process noisy data. Statistical methods or judgments based on domain knowledge may be used to select algorithms based on data characteristics and analysis objectives, identify attributes that can reflect the essential characteristics of the data, and principal component analysis methods may be used to reduce data dimensions while retaining important information. Part of the data may be used for model training, model parameters may be adjusted to optimize performance, and a validation set may be used for model evaluation. Based on historical data and business knowledge, define the standards for anomalies, perform real-time analysis on newly generated data according to anomaly detection rules, and when an anomaly is detected, trigger an alarm and record relevant information. Conduct in-depth analysis on the identified abnormal events to determine the cause and type of the anomaly. Establish an anomaly classification system, extract risk-related factors from the data, analyze the causal correlation between these factors and risks, select predictive models to predict risk relationships, use historical data to evaluate the predictive models, calculate the prediction error index, adjust and optimize the model based on the evaluation results, and improve the prediction accuracy.

2. The remote monitoring system for multi-sensor data fusion of unmanned aerial vehicles according to claim 1 is characterized in that: The multi-sensor data fusion technology automatically identifies and matches data from different sensors.

3. The remote monitoring system for multi-sensor data fusion of unmanned aerial vehicles according to claim 2 is characterized in that: The multi-sensor data fusion technology algorithm optimizes the in-depth processing and complementation of data.

4. The remote monitoring system for multi-sensor data fusion of unmanned aerial vehicles according to claim 1 is characterized in that: Each platform of the cross-platform collaborative monitoring mechanism can share monitoring data, task instructions and analysis results, forming a monitoring network with wide coverage and efficient collaboration.

5. The remote monitoring system for multi-sensor data fusion of unmanned aerial vehicles according to claim 1 is characterized in that: The data security and rights management uses encryption technology to encrypt and protect important data during transmission.

6. The remote monitoring system for multi-sensor data fusion of unmanned aerial vehicles according to claim 1 is characterized in that: The intelligent analysis and decision support selects algorithms based on data characteristics and analysis objectives, and identifies attributes that can reflect the essential characteristics of the data: It is a data point and In Properties The distance metric on is a tuning parameter used to control the influence of distance. It is a data point The comprehensive feature score of Define the abnormal standards, analyze the newly generated data in real time according to the abnormality detection rules, and trigger an alarm and record relevant information when an abnormality is detected: ,in, The newly generated data points we want to detect anomalies for, The mean of historical data, The standard deviation of historical data, This part represents the ratio of the new data point's deviation from the mean to the dispersion of the historical data. Risk-related factors Median, IQR Yes Factor The interquartile range, when Threshold, judgment data For abnormal data, an alarm is triggered and relevant information is recorded.

7. The remote monitoring system for multi-sensor data fusion of unmanned aerial vehicles according to claim 1 is characterized in that: The intelligent analysis and decision support uses a prediction model to predict risk relationships: ,in, is a positive number, It indicates the absolute difference between the predicted value and the threshold. The larger the difference, the higher the potential risk. Normalize the differences so that they are scaled relative to the threshold, making it easier to compare predictions of different dimensions. is an optional weighting factor used to adjust the impact of the difference on risk. is the risk factor, which is used to adjust the sensitivity of the risk level according to business needs. The VaR is always non-negative because risk is not usually expressed as a negative value.

8. The remote monitoring system for multi-sensor data fusion of unmanned aerial vehicles according to claim 1 is characterized in that: The intelligent analysis and decision support organizes data analysis results, abnormal events and potential risk information into clear and intuitive reports. The reports include charts, data tables and text descriptions. Based on the analysis results and risk predictions, they provide decision makers with specific suggestions and options. The suggestions should be based on scientific basis and data analysis, while taking into account the actual business situation, and building a friendly interactive interface so that decision makers can easily view analysis results, adjust parameters and obtain more information. Visualization tools are provided to enhance decision makers' understanding and insight into data.

Images