An internet cloud-based remote monitoring method

Abstract

The application discloses a kind of remote monitoring methods based on internet cloud, it is related to remote monitoring technical field, by data collection module, from the multiple data sources of internet cloud platform, remote and real-time collection monitoring data, by data preprocessing module, monitoring data is preprocessed and outputs fusion preprocessing index RC, based on fusion preprocessing index RC, and by comprehensive evaluation adjustment module, the risk level of monitoring system is comprehensively evaluated and monitoring strategy is automatically adjusted, feedback and optimization module feeds back monitoring result to data preprocessing module and comprehensive evaluation adjustment module, reach the continuous optimization of system, realize the improvement of data preprocessing quality, the enhancement of abnormal detection and risk assessment ability, the adjustment of self-adapting monitoring strategy and the overall beneficial effect of system continuous optimization and cyclic improvement, and adapt to the changing environment, realize long-term stable monitoring effect.

Description

Technical Field

[0001] This invention relates to the field of remote monitoring technology, specifically to a remote monitoring method based on the Internet cloud. Background Technology

[0002] In the field of remote monitoring in the cloud, with the rapid development of IoT, big data, and artificial intelligence technologies, the complexity and data volume of monitoring systems have increased dramatically. Among them, the popularization of IoT technology provides a wealth of monitoring data sources for remote monitoring systems. IoT connects the physical world with the digital world through sensors, RFID tags, and camera devices, enabling real-time data collection and transmission. In remote monitoring systems, these devices act as monitoring terminals, responsible for collecting environmental parameters, images, and video information, and transmitting them to the cloud platform via wireless or wired networks. The application of IoT technology enables remote monitoring systems to achieve comprehensive, 24 / 7 monitoring of target objects. Furthermore, by mining and analyzing massive amounts of monitoring data, patterns and rules can be discovered, providing strong support for decision-making. In remote monitoring systems, big data analytics technology is widely used in fault early warning, anomaly detection, and behavior analysis.

[0003] However, traditional monitoring systems typically rely on only a few data sources and do not fully consider the importance and noise intensity of these data sources, resulting in low-quality data fusion that affects the accuracy of subsequent analysis and decision-making. When processing large datasets, traditional methods often struggle to effectively identify outliers and lack risk assessment models that comprehensively consider data distribution characteristics and the number of outliers, leading to inaccurate risk assessment results. Furthermore, after detecting anomalies, traditional monitoring systems often only adopt fixed response strategies and cannot automatically adjust monitoring strategies based on the current risk level and system stability, resulting in slow response times, resource waste, and over-adjustment. Summary of the Invention

[0004] The purpose of this invention is to provide a remote monitoring method based on the Internet cloud, which solves the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a remote monitoring method based on the Internet cloud and including electronic digital data processing, computer and monitoring aspects;

[0006] It includes a data collection module, a data preprocessing module, a comprehensive evaluation and adjustment module, and a feedback and optimization module;

[0007] The specific implementation steps are as follows:

[0008] The data collection module remotely and in real time collects monitoring data from multiple data sources on the Internet cloud platform.

[0009] The data preprocessing module preprocesses the monitoring data and outputs the fusion preprocessing index RC. The data preprocessing module includes a multi-source data fusion and preprocessing unit.

[0010] Based on the fusion preprocessing index RC, and through the comprehensive evaluation and adjustment module, the risk level of the monitoring system is comprehensively evaluated and the monitoring strategy is automatically adjusted. The comprehensive evaluation and adjustment module includes an advanced anomaly detection and risk assessment unit and an adaptive monitoring strategy adjustment unit.

[0011] The feedback and optimization module feeds the monitoring results back to the data preprocessing module and the comprehensive evaluation and adjustment module to achieve continuous system optimization.

[0012] The data collection module uses devices including sensors, cameras, and web crawlers;

[0013] The data preprocessing module, the comprehensive evaluation and adjustment module, and the feedback and optimization module all use equipment including servers and cloud computing platforms.

[0014] Optionally, the calculation formula for the multi-source data fusion and preprocessing unit is as follows:

[0015] ;

[0016] in:

[0017] RC stands for Fusion Preprocessing Index;

[0018] m represents the total number of data sources;

[0019] SZ i Score the importance of the i-th data source;

[0020] SY i Let the noise intensity be the i-th data source.

[0021] a i The weighting factor for the i-th data source;

[0022] b is the natural base;

[0023] CYF is the redundancy detection score, which measures the degree of duplication and redundancy in a dataset.

[0024] c is the attenuation factor, which controls the rate at which the redundancy detection score CYF affects the fusion preprocessing index RC.

[0025] Optionally, the calculation formula for the advanced anomaly detection and risk assessment unit is as follows:

[0026] GF=RC*[(ZP / SF)+e*log(1+YZ)];

[0027] in:

[0028] GF stands for Advanced Risk Index;

[0029] ZP stands for Median Absolute Deviation, which measures the dispersion of the data distribution.

[0030] SF stands for interquartile range, which reflects the dispersion of the data in the middle 50%.

[0031] YZ represents the number of outliers;

[0032] e is the anomaly sensitivity factor, which controls the impact of the number of outliers YZ on the advanced risk index GF.

[0033] Optionally, the calculation formula for the adaptive monitoring strategy adjustment unit is as follows:

[0034] ZS=(GF*XY) / [GF+(XY*WY)];

[0035] in:

[0036] ZS is the adaptive strategy index;

[0037] XY is the response time factor, which represents the reciprocal of the time required for the system to execute the adjustment strategy after detecting an anomaly, i.e., the response speed.

[0038] WY is the stability factor. WY assesses the stability of the system based on its historical performance and external conditions. The higher the stability, the lower the probability of adjustment.

[0039] Optionally, when the adaptive policy index ZS indicates that the system needs a faster response, i.e., the advanced risk index GF is high and the response time factor XY is small, the system adjusts the response time factor XY, including optimizing the algorithm and increasing computing resources, to shorten the response time. Simultaneously, the system reassesses the importance of each data source based on the result of the adaptive policy index ZS, i.e., the importance score SZ of the i-th data source. i Since the calculation of the fusion preprocessing index RC depends on the importance score SZ of the i-th data source. i The noise intensity SY of the i-th data source i The redundancy detection score CYF and the above values ​​are both affected by system adjustments, which will indirectly affect the calculation of the next fusion preprocessing index RC.

[0040] Optionally, the adaptive strategy index ZS indicates that the system needs a faster response. That is, when the advanced risk index GF is high and the response time factor XY is small, the system adjusts the response time factor XY to shorten the response time. Other processing methods are as follows:

[0041] S1. Increase the frequency of monitoring data collection to promptly capture changes in system status;

[0042] S2. Adjust and optimize the parameters and thresholds of the anomaly detection algorithm based on the current risk situation;

[0043] S3. When a high-risk event is detected, the early warning mechanism is triggered in advance to notify relevant personnel to prepare for the response;

[0044] S4. Based on the results of the adaptive strategy index ZS, reassess the importance and reliability of each data source;

[0045] S5. If the system architecture itself has bottlenecks and deficiencies, resulting in limited response speed, the system architecture will be optimized and upgraded.

[0046] S6. Increase the detail and auditing of system logs to quickly locate the cause of the problem and take corresponding remedial measures after a risk event occurs;

[0047] S7. Develop detailed emergency plans and operating procedures in advance for potential high-risk events.

[0048] Optionally, the formula for calculating the absolute deviation of the median ZP is as follows:

[0049] ;

[0050] ZWS is the median, X i For each value in the dataset, ZWS(X) is the median of the dataset.

[0051] Optionally, the interquartile range (SF) is calculated using the following formula:

[0052] SF = F3 - F1;

[0053] F3 is the third quartile, or 75th quartile, located at the 75th percentile of the sorted dataset;

[0054] F1 is the first quartile, or 25th percentile, located at the 25th percentile of the sorted dataset;

[0055] Using the data preprocessing module, the dataset is sorted, and the third quartile F3 and the first quartile F1 are determined based on the position of the data points.

[0056] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0057] I. This invention integrates the preprocessing index RC and comprehensively considers the importance score SZ of the i-th data source. i The noise intensity SY of the i-th data source i The weighting factor a of the i-th data sourcei In addition, the Redundancy Detection Score (CYF) enables efficient fusion and preprocessing of multi-source data. This method not only improves the quality of data fusion but also reduces the impact of redundant information on subsequent analysis, providing a high-quality data foundation for subsequent anomaly detection and risk assessment.

[0058] Second, the advanced anomaly detection and risk assessment unit in this invention combines the fusion preprocessing index RC, median absolute deviation ZP, interquartile range SF, and outlier count YZ to construct a comprehensive risk assessment model. This model can more accurately identify abnormal events and assess their risk level, providing strong support for the decision-making of the monitoring system.

[0059] Third, the adaptive monitoring strategy adjustment unit in this invention achieves adaptive adjustment of the monitoring strategy by balancing the advanced risk index GF, response time factor XY, and stability factor WY. When the risk level is high and the system can respond quickly, the system tends to adopt a more aggressive adjustment strategy, while when the system stability is low, a more conservative adjustment scheme is adopted. This adaptive adjustment mechanism ensures that the monitoring system can flexibly cope with various complex situations, thereby improving monitoring efficiency and accuracy.

[0060] IV. This invention feeds back the result of the adaptive strategy index ZS to the importance score SZ of the i-th data source. i The adjustment of the response time factor XY and the impact of the number of outliers YZ in the advanced risk index GF on the redundancy detection score CYF form a cyclical influence mechanism. This mechanism enables the monitoring system to continuously learn, optimize and adapt to the ever-changing environment, thereby improving the system's intelligence level and long-term stability. Attached Figure Description

[0061] Figure 1 This is a flowchart of the remote monitoring method based on the Internet cloud.

[0062] Figure 2 This is a schematic diagram of the structure based on the data preprocessing module;

[0063] Figure 3 This is a schematic diagram of the structure of the comprehensive evaluation and adjustment module of the present invention;

[0064] Figure 4 This is a closed-loop diagram illustrating the cyclical influence of the three algorithm units in this invention. Detailed Implementation

[0065] The technical solutions of 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.

[0066] This cloud-based remote monitoring method differs from traditional monitoring systems. Traditional systems often rely on a single data source and have relatively simple data processing and anomaly detection methods, making it difficult to handle the fusion and efficient processing of multi-source heterogeneous data. Furthermore, traditional systems lack adaptive adjustment capabilities and cannot dynamically adjust monitoring strategies according to real-time risk levels, resulting in slow response speeds, high false alarm rates, and low resource utilization. In contrast, this algorithm unit achieves the overall beneficial effects of improved data preprocessing quality, enhanced anomaly detection and risk assessment capabilities, adaptive monitoring strategy adjustment, and continuous system optimization and iterative improvement. It also adapts to constantly changing environments and achieves long-term stable monitoring results.

[0067] Example 1, please refer to Figures 1 to 4 This embodiment provides a remote monitoring method based on the Internet cloud, which includes electronic digital data processing, computer and monitoring aspects of remote monitoring.

[0068] It includes a data collection module, a data preprocessing module, a comprehensive evaluation and adjustment module, and a feedback and optimization module;

[0069] The specific implementation steps are as follows:

[0070] The data collection module remotely and in real time collects monitoring data from multiple data sources on the Internet cloud platform.

[0071] The data preprocessing module preprocesses the monitoring data and outputs the fusion preprocessing index RC. The data preprocessing module includes a multi-source data fusion and preprocessing unit.

[0072] Based on the fusion preprocessing index RC, and through the comprehensive evaluation and adjustment module, the risk level of the monitoring system is comprehensively evaluated and the monitoring strategy is automatically adjusted. The comprehensive evaluation and adjustment module includes an advanced anomaly detection and risk assessment unit and an adaptive monitoring strategy adjustment unit.

[0073] The feedback and optimization module feeds the monitoring results back to the data preprocessing module and the comprehensive evaluation and adjustment module to achieve continuous system optimization.

[0074] The data collection module uses devices including sensors, cameras, and web crawlers;

[0075] The data preprocessing module, comprehensive evaluation and adjustment module, and feedback and optimization module all use equipment including servers and cloud computing platforms.

[0076] In this embodiment, the system plays a crucial role in remote monitoring via the internet cloud through the coordinated operation of three algorithm units. These units are responsible for evaluating and optimizing data preprocessing quality, assessing the risk level of abnormal events, and adaptively adjusting monitoring strategies, respectively. Combined with the results of the three calculations (RC, GF, and ZS), they constitute the core mechanism for the intelligent and efficient operation of the monitoring system. RC, the fusion preprocessing index, helps the system identify and reduce the impact of noise and redundant information during the data fusion stage, ensuring that the fused data is more accurate and reliable, providing a high-quality data foundation for subsequent anomaly detection and risk assessment. GF, the advanced risk index, provides important basis for the monitoring system's decision-making, helping the system quickly identify high-risk events and take corresponding countermeasures to reduce the impact of potential risks on the system. ZS, the adaptive strategy index... The data enables the monitoring system to flexibly respond to various complex situations, automatically adjusting monitoring strategies based on risk levels and system stability, such as adjusting monitoring frequency and changing alarm thresholds, thereby improving the intelligence level and response speed of the monitoring system. Furthermore, the calculation results of ZS can influence the calculations of RC and GF, resulting in a high degree of correlation and entanglement among the three algorithms in this system. This allows the overall algorithm system to automatically provide feedback and optimization based on actual conditions, making it more realistic. In addition, the three algorithm units of this system provide a basis for quantitative evaluation and adjustment of the remote monitoring system, enabling the system to more accurately identify and handle abnormal events, improving monitoring efficiency and accuracy. It is closely related to calculation, estimation, counting, electronic digital data processing, computers, and monitoring, and through complex algorithms and data processing technologies, it achieves the fusion, preprocessing, anomaly detection, and strategy adjustment of multi-source monitoring data.

[0077] Please see Figures 1 to 4 The calculation formula for the multi-source data fusion and preprocessing unit is as follows:

[0078] ;

[0079] in:

[0080] RC stands for Fusion Preprocessing Index;

[0081] m represents the total number of data sources;

[0082] SZ i Score the importance of the i-th data source;

[0083] SY i Let the noise intensity be the i-th data source.

[0084] a i The weighting factor for the i-th data source;

[0085] b is the natural base;

[0086] CYF is the redundancy detection score, which measures the degree of duplication and redundancy in a dataset.

[0087] c is the attenuation factor, which controls the rate at which the redundancy detection score CYF affects the fusion preprocessing index RC.

[0088] In this embodiment: First, in this algorithm unit, SZ i The importance score for the i-th data source is typically assessed based on factors such as the data source's historical performance, reliability, and accuracy. This score is derived from system administrator evaluation and algorithmic analysis. i Let be the noise intensity of the i-th data source. This noise intensity index is calculated using multiple methods, including the proportion of outliers in the statistical data and the reciprocal of the signal-to-noise ratio. The specific calculation formula depends on the type and characteristics of the data source. i The weighting factor for the i-th data source is usually set by the system administrator based on historical data and data source quality factors, and is determined based on experience and strategy. CYF is the redundancy detection score, which is calculated by calculating the correlation coefficient matrix between data sources, identifying highly correlated data source pairs, and then calculating the score based on the number of these pairs and the strength of their correlation.

[0089] This algorithm unit comprehensively considers the importance score SZ of the i-th data source. i The noise intensity SY of the i-th data source i and the weighting factor a of the i-th data source i This enables the multi-source data fusion and preprocessing unit to effectively filter out high-quality data sources and reduce the impact of noise on the preprocessing results, thereby improving the overall quality of the fused data.

[0090] The introduction of the correlation-based redundancy detection score CYF enables the multi-source data fusion and preprocessing unit to identify and reduce duplicate and redundant information in the dataset, further improving the effectiveness and processing efficiency of the data.

[0091] By adjusting the weighting factor a of the i-th data source i The attenuation factor c allows the system to flexibly adjust the weight of the data source and the degree of influence of redundant detection according to the actual situation, so as to adapt to different monitoring needs and environmental changes.

[0092] Please see Figures 1 to 4 The calculation formula for the Advanced Anomaly Detection and Risk Assessment Unit is as follows:

[0093] GF=RC*[(ZP / SF)+e*log(1+YZ)];

[0094] in:

[0095] GF stands for Advanced Risk Index;

[0096] ZP stands for Median Absolute Deviation, which measures the dispersion of the data distribution.

[0097] SF stands for interquartile range, which reflects the dispersion of the data in the middle 50%.

[0098] YZ represents the number of outliers;

[0099] e is the anomaly sensitivity factor, which controls the impact of the number of outliers YZ on the advanced risk index GF;

[0100] The logarithm is base 10.

[0101] In this embodiment, firstly, the advanced anomaly detection and risk assessment unit combines the fusion preprocessing index RC, the statistical characteristics of the data, and the number of outliers to comprehensively assess the risk level of the monitoring system, namely the advanced risk index GF.

[0102] This algorithm unit combines the fusion preprocessing index RC, median absolute deviation ZP, interquartile range SF, and outlier count YZ to comprehensively and accurately assess the advanced risk index GF of abnormal events in the monitoring system, providing strong support for subsequent decision-making and response. By introducing the statistical characteristics of median absolute deviation ZP and interquartile range SF, the advanced anomaly detection and risk assessment unit can more accurately identify outliers in the data, reducing false alarm and false negative rates. The introduction of the anomaly sensitivity factor e allows the system to adjust the impact of the outlier count YZ on the advanced risk index GF according to the actual situation, thereby flexibly responding to different types of abnormal events.

[0103] Please see Figures 1 to 4 The calculation formula for the adaptive monitoring strategy adjustment unit is as follows:

[0104] ZS=(GF*XY) / [GF+(XY*WY)];

[0105] in:

[0106] ZS is the adaptive strategy index;

[0107] XY is the response time factor, which represents the reciprocal of the time required for the system to execute the adjustment strategy after detecting an anomaly, i.e., the response speed.

[0108] WY is the stability factor. WY assesses the stability of the system based on its historical performance and external conditions. The higher the stability, the lower the probability of adjustment.

[0109] In this embodiment, the algorithm unit first calculates an adaptive strategy index ZS based on the advanced risk index GF, response time factor XY, and stability factor WY. This index guides the system on how to adjust the monitoring strategy according to the real-time situation to achieve the best monitoring effect. The adaptive strategy index ZS enables the monitoring system to flexibly cope with various complex situations and automatically adjust the monitoring strategy according to the risk level and system stability, such as adjusting the monitoring frequency and changing the alarm threshold, thereby improving the intelligence level and response speed of the monitoring system.

[0110] This algorithm unit, by balancing the advanced risk index GF, response time factor XY, and stability factor WY, guides the monitoring system to respond quickly and take effective countermeasures when high-risk events are detected, reducing the impact of risks on the system. Furthermore, when the risk level is low and the system stability is high, the adaptive monitoring strategy adjustment unit can guide the monitoring system to adopt a more conservative adjustment strategy, avoiding unnecessary resource waste and over-adjustment. Subsequently, by adaptively adjusting the monitoring strategy, the adaptive monitoring strategy adjustment unit enables the monitoring system to continuously learn, optimize, and adapt to the ever-changing environment, improving the system's intelligence level and long-term stability.

[0111] Please see Figures 1 to 4 When the adaptive policy index ZS indicates that the system needs a faster response (i.e., a higher advanced risk index GF and a smaller response time factor XY), the system adjusts the response time factor XY, including optimizing the algorithm and increasing computational resources, to shorten the response time. Simultaneously, based on the result of the adaptive policy index ZS, the system reassesses the importance of each data source, i.e., the importance score SZ of the i-th data source. i Since the calculation of the fusion preprocessing index RC depends on the importance score SZ of the i-th data source. i The noise intensity SY of the i-th data source i The redundancy detection score CYF and the above values ​​are all affected by system adjustments, which will indirectly affect the calculation of the next fusion preprocessing index RC.

[0112] The adaptive policy index ZS indicates that the system needs a faster response. That is, when the advanced risk index GF is high and the response time factor XY is small, the system adjusts the response time factor XY to shorten the response time. Other processing methods are as follows:

[0113] S1. Increase the frequency of monitoring data collection to promptly capture changes in system status;

[0114] S2. Adjust and optimize the parameters and thresholds of the anomaly detection algorithm based on the current risk situation;

[0115] S3. When a high-risk event is detected, the early warning mechanism is triggered in advance to notify relevant personnel to prepare for the response;

[0116] S4. Based on the results of the adaptive strategy index ZS, reassess the importance and reliability of each data source;

[0117] S5. If the system architecture itself has bottlenecks and deficiencies, resulting in limited response speed, the system architecture will be optimized and upgraded.

[0118] S6. Increase the detail and auditing of system logs to quickly locate the cause of the problem and take corresponding remedial measures after a risk event occurs;

[0119] S7. Develop detailed emergency plans and operating procedures in advance for potential high-risk events.

[0120] In this embodiment, the algorithm unit is based on the cyclical influence of the adaptive monitoring strategy adjustment unit on the multi-source data fusion and preprocessing unit. Its feedback influence is mainly reflected in the adaptive strategy index ZS being fed back to the importance score SZ of the i-th data source. i Specifically, regarding the adjustment, when the adaptive strategy index ZS indicates that the system needs to optimize data fusion, the system adjusts the importance score SZ of the i-th data source from different data sources. i To optimize the data fusion process, this feedback mechanism enables the system to dynamically adjust the importance of data sources based on actual monitoring results and risk levels, thereby improving the quality and accuracy of data fusion. At the same time, this cyclical influence mechanism also promotes the continuous learning and optimization of the monitoring system, enabling the system to better adapt to complex and ever-changing monitoring environments.

[0121] The multi-source data fusion and preprocessing unit calculates the adaptive policy index ZS, comprehensively considering the advanced risk index GF, response time factor XY, and stability factor WY, providing a basis for adjusting the monitoring system's policy. When the adaptive policy index ZS indicates that the system needs to optimize the monitoring policy, this optimization often involves reassessing the importance of the data source. The importance score SZ of the i-th data source in the multi-source data fusion and preprocessing unit... i It is one of the key parameters for data fusion and preprocessing. Through feedback from the adaptive strategy index ZS, the system can dynamically adjust the importance score SZ of different i-th data sources. i This allows for greater focus on data sources that significantly impact risk assessment and anomaly detection during the data fusion process, while reducing the impact of noise and redundant information on system performance.

[0122] Dynamic optimization of data source weights directly improves the quality of data fusion. In remote monitoring systems, data fusion is the process of integrating information from multiple data sources into consistent, accurate, and useful information. Through the feedback mechanism of the adaptive monitoring strategy adjustment unit, the system can continuously learn and optimize the allocation of data source weights, making the fused data more accurate and comprehensive in reflecting the actual status of the monitored object. This improvement in data fusion quality further enhances the anomaly detection capability and risk assessment accuracy of the monitoring system, providing more reliable data support for subsequent decision-making and response.

[0123] The adaptive monitoring strategy adjustment unit's cyclical influence mechanism on the multi-source data fusion and preprocessing unit enables the monitoring system to continuously adapt to complex and ever-changing monitoring environments. When the monitoring environment changes, the system can respond through feedback from the adaptive strategy index ZS and the importance score SZ of the i-th data source. i The system can quickly adapt to these changes and maintain the stability and accuracy of the monitoring system. At the same time, this cyclical impact mechanism also enhances the robustness of the system. When faced with interference factors such as data noise and outliers, the system can reduce the impact of these interference factors on system performance by optimizing data source weights and preprocessing strategies, thus ensuring the stable operation of the monitoring system.

[0124] The adaptive monitoring strategy adjustment unit's cyclical influence mechanism on the multi-source data fusion and preprocessing unit actually constructs a closed-loop learning and optimization process. In this process, the system continuously learns and optimizes its own performance through multiple steps such as continuously collecting data, assessing risks, adjusting strategies, and optimizing data source weights. This continuous learning and optimization process enables the monitoring system to continuously adapt to new monitoring needs and environmental changes, improve its intelligence level and long-term stability, and at the same time, provide a solid foundation for the further upgrading and expansion of the system.

[0125] In summary, during the specific implementation process, the cyclical influence mechanism of the adaptive monitoring strategy adjustment unit on the multi-source data fusion and preprocessing unit has significant beneficial effects on the remote monitoring system. These effects include dynamically optimizing data source weights, improving data fusion quality, enhancing system adaptability and robustness, and promoting continuous system learning and optimization. These effects collectively improve the overall performance and reliability of the monitoring system, providing strong support for the widespread application of remote monitoring.

[0126] Example 2, please refer to Figures 1 to 4 The formula for calculating the absolute deviation of the median (ZP) is as follows:

[0127] ;

[0128] ZWS is the median, X iFor each value in the dataset, ZWS(X) is the median of the dataset;

[0129] The formula for calculating the interquartile range (SF) is as follows:

[0130] SF = F3 - F1;

[0131] F3 is the third quartile, or 75th quartile, located at the 75th percentile of the sorted dataset;

[0132] F1 is the first quartile, or 25th percentile, located at the 25th percentile of the sorted dataset;

[0133] The dataset is sorted using the data preprocessing module, and the third quartile F3 and the first quartile F1 are determined based on the position of the data points.

[0134] In this embodiment, the median absolute deviation ZP represents the median absolute deviation, which is a statistic that measures the average absolute distance of each point in the dataset from the median. It is used to assess the dispersion of the data distribution, that is, the magnitude of the deviation between the data points and the median. Substituting the median absolute deviation ZP into the advanced anomaly detection and risk assessment unit is to assess the impact of the dispersion of the data distribution on the advanced risk index GF. The greater the dispersion, the more significant the anomalies or fluctuations in the data, thereby increasing the risk level of the system. Therefore, the advanced risk index GF, as a quantitative indicator of the dispersion of the data distribution, helps to more accurately assess the risk status of the system.

[0135] The interquartile range (IFR) is the difference between the third quartile (F3) and the first quartile (F1). It measures the dispersion of the middle 50% of the data, i.e., the distribution range of most of the data in the dataset. In advanced anomaly detection and risk assessment units, the IFR is used as background information when calculating the median absolute deviation (ZP). The magnitude of the IFR reflects the stability and variability of the middle part of the data. In outlier detection, the IFR is often used to determine the threshold for outliers. The IFR helps to more comprehensively understand the distribution characteristics of the data and its impact on risk assessment.

[0136] 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. An internet cloud-based remote monitoring method, characterized in that, A remote monitoring method based on internet cloud and including electronic digital data processing, computer and monitoring aspects; It comprises a data collection module, a data preprocessing module, a comprehensive evaluation adjustment module and a feedback and optimization module; The specific implementation steps are as follows: The data collection module is used to remotely and real-time collect monitoring data from the data source of the internet cloud platform; The data preprocessing module is used to preprocess the monitoring data and output a fusion preprocessing index RC, and the data preprocessing module comprises a multi-source data fusion and preprocessing unit; Based on the fusion preprocessing index RC, the comprehensive evaluation adjustment module is used to comprehensively evaluate the risk level of the monitoring system and automatically adjust the monitoring strategy, and the comprehensive evaluation adjustment module comprises a high-level anomaly detection and risk evaluation unit and a self-adaptive monitoring strategy adjustment unit; The feedback and optimization module feeds back the monitoring results to the data preprocessing module and the comprehensive evaluation adjustment module, so as to achieve continuous optimization of the system; The calculation formula of the multi-source data fusion and preprocessing unit is as follows: ； Wherein: RC is the fusion preprocessing index; m is the total number of data sources; SZ i Importance score for the i-th data source; SY i is the noise intensity for the i-th data source; a i weighting factor for the i-th data source; b is the natural base; CYF is the redundancy detection score, which measures the degree of repeated and redundant information in the data set; c is the attenuation factor, which controls the influence speed of the redundancy detection score CYF on the fusion preprocessing index RC; The calculation formula of the high-level anomaly detection and risk evaluation unit is as follows: GF=RC*[(ZP / SF)+e*log(1+YZ)]; Wherein: GF is the high-level risk index; ZP is the median absolute deviation, which measures the dispersion degree of data distribution; SF is the interquartile range, which reflects the dispersion degree of the middle 50% of data; YZ is the number of outliers; e is the anomaly sensitivity factor, which controls the influence of the number of outliers YZ on the high-level risk index GF; The calculation formula of the self-adaptive monitoring strategy adjustment unit is as follows: ZS=(GF*XY) / [GF+(XY*WY)]; Wherein: ZS is the adaptive strategy index; XY is the response time factor, which represents the inverse of the time required for the system to detect an anomaly and execute an adjustment strategy, i.e. response speed; WY is the stability factor, which evaluates the stability of the system according to its historical performance and external conditions. The higher the stability, the lower the adjustment probability.

2. The remote monitoring method based on internet cloud according to claim 1, wherein The devices used by the data collection module include sensors, cameras and web crawlers; The devices used by the data preprocessing module, the comprehensive evaluation adjustment module and the feedback and optimization module all include servers and cloud computing platforms. 3.The Internet cloud-based remote monitoring method according to claim 2, characterized in that: When the adaptive strategy index Zs indicates that the system needs a faster response, i.e. the high risk index GF is high and the response time factor XY is short, the system adjusts the response time factor XY, including optimizing the algorithm and increasing the computing resources, to shorten the response time, at the same time, the system re-evaluates the importance of each data source according to the result of the adaptive strategy index Zs, i.e. the importance score SZ of the i-th data source i Since the calculation of the fusion preprocessing index RC depends on the importance score SZ of the i-th data source i , the noise intensity SY of the i-th data source i and the redundancy detection score CYF, the above values are affected by the system adjustment, which in turn affects the calculation of the fusion preprocessing index RC next time.

4. The method of claim 3, wherein the method further comprises: The adaptive strategy index ZS indicates that the system needs faster response, i.e. the high-level risk index GF is high and the response time factor XY is low, so the system adjusts the response time factor XY to shorten the processing method as follows: S1, increase the collection frequency of monitoring data to capture the changes in system state in time; S2, adjust and optimize the parameters and thresholds of the anomaly detection algorithm according to the current risk situation; S3, when a high-risk event is detected, the early warning mechanism is triggered in advance to inform relevant personnel to prepare for response; S4, according to the result of the adaptive strategy index ZS, the importance and reliability of each data source are re-evaluated; S5, if the system architecture itself has bottlenecks and deficiencies, resulting in limited response speed, when the system architecture is optimized and upgraded; S6, increase the detail level and audit intensity of system logs, quickly locate the problem cause after the risk event occurs, and take corresponding remedial measures; S7, for possible high-risk events, detailed emergency plans and operation processes are prepared in advance.

5. The method of claim 2, wherein the method further comprises: The calculation formula of the median absolute deviation ZP is as follows: ； ZWS is the median, X i is each value in the data set, and ZWS(X) is the median in the data set.

6. The method of claim 2, wherein the method further comprises: The calculation formula of the interquartile range SF is as follows: SF=F3-F1; F3 is the third quartile, that is, the 75% quantile, located at the 75% position of the sorted data set; F1 is the first quartile, that is, the 25% quantile, located at the 25% position of the sorted data set; The data set is sorted by using the data preprocessing module, and the third quartile F3 and the first quartile F1 are determined according to the position of the data points.

Images