Intelligent Device Management System Based on Cloud Platform

Abstract

The present invention relates to the field of intelligent device management technology, and more specifically to a cloud-based intelligent device management system comprising a data acquisition module, a data processing module, a feature extraction module, a prediction and analysis module, and an optimization and decision-making module. Specifically, the data acquisition module is used to collect operating data from each intelligent device; the data processing module is used to preprocess the operating data; the feature extraction module is used to extract relevant features from the preprocessed operating data and generate a feature vector for the device status; the prediction and analysis module is used to classify the operating status of the device; and the optimization and decision-making module is used to formulate an optimized operating strategy for the device. Through real-time data acquisition and intelligent analysis, the present invention achieves accurate device status assessment and dynamically optimizes operating strategies, effectively improving device management efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent device management, and in particular to an intelligent device management system based on a cloud platform. Background Art

[0002] The continuous development of intelligent technology and the widespread application of various intelligent devices, especially in the fields of industry, energy, and construction, have driven the upgrading of the demand for equipment management and maintenance. Traditional equipment management methods often rely on manual operation and lack efficient monitoring and intelligent decision-making capabilities, resulting in poor equipment operating efficiency and maintenance effects, and even unexpected downtime, losses, and safety hazards caused by unreasonable operating parameters or equipment failures. In order to improve equipment operating efficiency, reduce energy consumption, and timely predict and prevent failures, cloud-based equipment management systems have gradually become a mainstream solution. However, existing intelligent equipment management systems usually rely only on traditional rule-based control and early warning mechanisms, lacking intelligent strategies for real-time optimization and adjustment based on the equipment's operating status.

[0003] Most existing intelligent equipment management systems have certain limitations in terms of equipment failure prediction, status assessment, and maintenance decision-making. First, many systems use models based on historical data for prediction, ignoring the sudden changes in equipment behavior and real-time data fluctuations that may occur during equipment operation, resulting in inaccurate prediction results. Second, although some systems can provide fault warnings, they lack sufficient processing power to effectively distinguish between normal equipment states and potential fault states, and then propose accurate optimized operation strategies. The adjustment of equipment's optimized operating parameters, operating schedules, and preventive maintenance plans lacks systematic, intelligent decision-making based on real-time changes in equipment status, and often relies on human judgment, resulting in low management efficiency. Summary of the Invention

[0004] Based on the above objectives, the present invention provides a cloud platform-based intelligent device management system.

[0005] The cloud-based intelligent device management system includes a data acquisition module, a data processing module, a feature extraction module, a prediction and analysis module, and an optimization and decision-making module.

[0006] Data acquisition module: used to collect the operating data of each smart device in real time through various sensors and network interfaces, including the device's working status, environmental parameters, energy consumption data and operation records, and upload the collected operating data in a structured format to the cloud platform;

[0007] Data processing module: used to pre-process the operating data in the cloud platform, including data cleaning, missing value filling and data normalization;

[0008] Feature extraction module: Utilizes statistical analysis and signal processing algorithms to extract relevant features from pre-processed operating data, including trend features, periodic features, and mutation features of equipment operation, and generates a feature vector of the equipment status.

[0009] Prediction and analysis module: Based on the feature vectors generated by the feature extraction module, the K-means clustering algorithm is used to classify the operating status of the equipment and identify whether the equipment is in a normal state or a potential fault state;

[0010] Optimization decision module: Based on the operating status identified by the predictive analysis module, it formulates an optimized operation strategy for the equipment, including adjusting the equipment's operating parameters, optimizing the equipment's operating schedule, and formulating a preventive maintenance plan.

[0011] Optionally, the data acquisition module includes a sensor acquisition unit, an environmental data acquisition unit, an energy consumption monitoring unit, an operation record acquisition unit, and a data upload unit; wherein:

[0012] Sensor acquisition unit: including temperature sensor, humidity sensor, pressure sensor, current sensor and vibration sensor, used to collect working status data of smart devices in real time, including temperature, humidity, pressure, current and vibration parameters of the device;

[0013] Environmental data acquisition unit: includes environmental sensors for real-time acquisition of parameter data of the environment where the device is located, including temperature, humidity, and light intensity;

[0014] Energy consumption monitoring unit: includes energy consumption sensors, which are used to monitor the energy consumption data of the equipment in real time, including the power usage, working hours and load changes of the equipment;

[0015] Operation record collection unit: used to record the operation information of the equipment in real time, including equipment startup, shutdown and operation mode switching;

[0016] Data upload unit: used to upload all operating data collected by each unit, including the equipment's working status, environmental parameters, energy consumption data and operation records, to the cloud platform for storage in a structured data format through the network interface.

[0017] Optionally, the data processing module includes a data cleaning unit, a missing value filling unit, and a data normalization unit; wherein:

[0018] Data cleaning unit: used to remove abnormal and noisy data from the operational data received from the cloud platform. Specifically, it uses the upper and lower limit checking method to set the upper and lower limit ranges and check whether the data points exceed the range. If the data points exceed the range, they are regarded as abnormal data and are eliminated.

[0019] Missing value filling unit: used to fill in data with missing values. Specifically, it uses interpolation method to fill in missing data values based on the existing adjacent data point values through linear interpolation or polynomial interpolation method to ensure data integrity;

[0020] Data normalization unit: used to normalize the collected data, convert data of different dimensions into a unified standard range, and scale the data to the range of [0,1] or [-1,1] by normalizing the standard deviation of each data item or the minimum-maximum normalization method to avoid comparison deviations between different dimensions.

[0021] Optionally, the feature extraction module includes a trend feature extraction unit, a periodic feature extraction unit, a mutation feature extraction unit, and a feature vector generation unit; wherein:

[0022] Trend feature extraction unit: used to identify and extract the long-term trend features of equipment operation from pre-processed operation data, determine the overall trend of equipment performance changes over time by fitting time series data, and generate trend parameters reflecting changes in equipment operation status;

[0023] Periodic feature extraction unit: This unit uses the Fourier transform algorithm to perform frequency domain analysis on pre-processed operating data. By converting time series data into a frequency spectrum, it determines the periodic frequency and amplitude of the equipment operation and generates frequency parameters that reflect the periodic behavior of the equipment.

[0024] Mutation feature extraction unit: used to detect and extract mutation features in equipment operation from pre-processed operation data. By calculating the moving average and its difference of the data series, it identifies the mutation points in the data and generates mutation parameters that reflect abnormal fluctuations of the equipment.

[0025] Feature vector generation unit: used to integrate various feature parameters extracted by the trend feature extraction unit, the periodic feature extraction unit and the mutation feature extraction unit, and form a feature vector of the device state by arranging the various feature parameters in a predetermined order.

[0026] Optionally, the mutation feature extraction unit includes:

[0027] Calculate the moving average: Select a fixed window size w from the preprocessed running data sequence and calculate the moving average of the data in the window. The calculation formula for the moving average is:

[0028] in, is the moving average at time t, x(i) is the original data at the i-th moment in the time series, and w is the size of the sliding window;

[0029] Calculate the difference: By calculating the difference of the data series, the severity of the data change, especially the mutation point, is detected. The difference formula is: Among them, Δx(t) is the difference value at time t, is the moving average at time t;

[0030] Identify mutation points: Based on the calculated difference value, set a threshold δ. When the difference value is greater than the threshold δ, it is considered that the data has mutated, and the moment is marked as the mutation point.

[0031] Generate mutation parameters: Generate mutation parameters that reflect abnormal fluctuations of the equipment by comparing the difference between the detected mutation points and the original operating data and the moving average. The expression is: p(t) = |Δx(t)|, where p(t) is the mutation parameter at the mutation point t.

[0032] Optionally, the feature vector generating unit includes:

[0033] Feature parameter sorting: Let the trend parameter generated by the trend feature extraction unit be T, the frequency parameter generated by the periodic feature extraction unit be F, and the mutation parameter generated by the mutation feature extraction unit be P; and sort the feature parameters in the following order: first arrange the trend feature parameter T, then arrange the periodic feature parameter F, and finally arrange the mutation feature parameter P, to form an ordered feature sequence {T, F, P};

[0034] Feature parameter normalization: normalize the sorted feature parameters and use the minimum-maximum normalization method to scale each feature parameter to the range of [0, 1].

[0035] Feature vector construction: Arrange the normalized feature parameters in a predetermined order to construct the feature vector V of the device state.

[0036] Optionally, the prediction analysis module includes a data classification unit, a state identification unit, and a result generation unit; wherein:

[0037] Data classification unit: Based on the feature vectors generated by the feature extraction module, the K-means clustering algorithm is used to perform unsupervised classification of the equipment's operating status data to form several categories, each of which represents a different mode of the equipment's operating status;

[0038] State recognition unit: Based on the clustering results generated by the data classification unit, it analyzes the feature centers of each category and determines whether the device is in a normal state or a potential fault state by comparing the distance between the device's real-time feature vector and the feature centers of each category.

[0039] Result generation unit: used to output the classification results in the form of structured data for reference by the optimization decision module.

[0040] Optionally, the data classification unit includes:

[0041] Initialize cluster centers: randomly select k eigenvectors as initial cluster centers, denoted as μ1, μ2, …, μ j ;

[0042] Iteratively assign eigenvectors: For each eigenvector V i , calculate its relationship with each cluster center μ j The Euclidean distance is: Among them, V i,m is the eigenvector V i The mth feature, μ j,m is the cluster center μ j The mth feature of V, M is the dimension of the feature vector, i Assigned to the cluster with the nearest cluster center;

[0043] Update cluster center: For each cluster j, calculate the average value of all feature vectors assigned to the cluster and update the cluster center μ j , the formula is: Among them, N j is the number of eigenvectors in cluster j, V i is the i-th eigenvector assigned to the corresponding cluster j;

[0044] Convergence determination: Repeat the above steps of iteratively assigning feature vectors and updating cluster centers until the cluster centers no longer change or the predetermined number of iterations is reached, forming the final k cluster categories.

[0045] Optionally, the state identification unit includes:

[0046] Feature vector acquisition: Get the real-time feature vector V of the device real , the feature vector is generated by a feature vector generating unit;

[0047] Distance calculation: For each cluster center, calculate the real-time feature vector V real The Euclidean distance D from the cluster center j , the formula is: Among them, D j is the real-time feature vector V real and cluster center μ j The Euclidean distance between them; M is the dimension of the feature vector; V real,m is the real-time feature vector V real The mth feature of j,m is the cluster center μ j The mth feature of

[0048] Shortest distance recognition: Determine the feature vector V in real time real The cluster center with the shortest distance μ min , its expression is: μ min =argmin j D j ;

[0049] State judgment: According to the cluster center μ min The category label to which the device belongs is used to determine the current operating status of the device. The category label is pre-defined by the data classification unit in the K-means clustering process and includes a normal category and a fault category, which correspond to the normal state and the potential fault state respectively.

[0050] Optionally, the optimization decision module includes an operation parameter adjustment unit, an operation schedule optimization unit, a preventive maintenance plan formulation unit, and a strategy integration implementation unit; wherein:

[0051] Operation parameter adjustment unit: used to adjust the equipment's operating parameters based on the equipment's operating status identified by the predictive analysis module. When the equipment is in a normal state, the unit optimizes the equipment's power and speed to improve its energy efficiency and work efficiency. When the equipment is in a potential fault state, the unit limits the equipment's operating load or temporarily reduces its operating parameters to ensure its safe operation.

[0052] Operation schedule optimization unit: used to optimize the operation schedule of the equipment according to the operation status and usage requirements of the equipment. When the equipment is in a normal state, the operation schedule of the equipment is arranged by maximizing the utilization rate of the equipment. When the equipment is in a potential failure state, the operation schedule of the equipment is adjusted to reduce the continuous operation load of the equipment.

[0053] Preventive Maintenance Plan Development Unit: Develops preventive maintenance plans based on the identification of potential equipment failures. When equipment is in a potential failure state, it specifies specific maintenance content, including arranging equipment inspection, maintenance, and parts replacement. When the equipment is in a normal state, it develops a regular maintenance plan based on the equipment's maintenance history.

[0054] Strategy integration and implementation unit: used to integrate various optimization strategies formulated by the operating parameter adjustment unit, operating schedule optimization unit and preventive maintenance plan formulation unit, and implement specific optimization operations through the cloud platform.

[0055] Beneficial effects of the present invention:

[0056] The present invention can accurately assess the health status of equipment through real-time collection, preprocessing and feature extraction of equipment operation data, and effectively distinguish between the normal state and potential fault state of equipment using the K-means clustering algorithm; the system not only improves the accuracy of equipment status judgment, but also can dynamically adjust the equipment's operating parameters, optimize the equipment's operating schedule, and formulate a more scientific preventive maintenance plan, thereby minimizing the occurrence of equipment failures and improving equipment management efficiency.

[0057] This invention, by combining real-time equipment data with intelligent analysis technology, breaks through the limitations of traditional equipment management systems that rely solely on static rules and manual judgment; its intelligent prediction and optimization strategies can respond to changes in equipment operating status in real time, ensuring that the equipment always maintains the optimal working state under different working conditions, reducing unnecessary energy consumption and downtime. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0059] Figure 1 This is a schematic diagram of an intelligent device management system according to an embodiment of the present invention;

[0060] Figure 2 Schematic diagram of a prediction analysis module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. It is also noted that, to provide a more detailed description, the following embodiments are best and preferred embodiments, and those skilled in the art may employ alternative methods for implementing certain known technologies. Furthermore, the accompanying drawings are intended only to provide a more detailed description of the embodiments and are not intended to limit the present invention.

[0062] It should be noted that references in the specification to "one embodiment," "an embodiment," "an exemplary embodiment," "some embodiments," etc. indicate that the described embodiments may include specific features, structures, or characteristics, but not every embodiment necessarily includes such specific features, structures, or characteristics. In addition, when specific features, structures, or characteristics are described in conjunction with an embodiment, it is within the knowledge of persons skilled in the relevant art to implement such features, structures, or characteristics in conjunction with other embodiments (whether or not explicitly described).

[0063] In general, terms can be understood, at least in part, from their use in context. For example, depending at least in part on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending at least in part on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0064] like Figure 1-Figure 2 As shown in FIG, the intelligent device management system based on the cloud platform includes a data acquisition module, a data processing module, a feature extraction module, a prediction analysis module, and an optimization decision module; among which:

[0065] Data acquisition module: used to collect the operating data of each smart device in real time through various sensors and network interfaces, including the device's working status, environmental parameters, energy consumption data and operation records, and upload the collected operating data in a structured format to the cloud platform;

[0066] Data processing module: used to pre-process the operating data in the cloud platform, including data cleaning, missing value filling and data normalization. The processed data will serve as the basis for subsequent analysis;

[0067] Feature extraction module: Utilizes statistical analysis and signal processing algorithms to extract relevant features from pre-processed operating data, including trend features, periodic features, and mutation features of equipment operation, and generates a feature vector of the equipment status.

[0068] Prediction and analysis module: Based on the feature vectors generated by the feature extraction module, the K-means clustering algorithm is used to classify the operating status of the equipment and identify whether the equipment is in a normal state or a potential fault state;

[0069] Optimization decision module: Based on the operating status identified by the predictive analysis module, it formulates an optimized operation strategy for the equipment, including adjusting the equipment's operating parameters (such as power, speed, etc.), optimizing the equipment's operating schedule, and formulating a preventive maintenance plan to improve the equipment's operating efficiency and extend its service life.

[0070] The data acquisition module includes a sensor acquisition unit, an environmental data acquisition unit, an energy consumption monitoring unit, an operation record acquisition unit, and a data upload unit; wherein:

[0071] Sensor acquisition unit: including temperature sensor, humidity sensor, pressure sensor, current sensor and vibration sensor, used to collect working status data of smart devices in real time, including temperature, humidity, pressure, current and vibration parameters of the device;

[0072] Environmental data acquisition unit: includes environmental sensors for real-time acquisition of parameter data of the environment where the device is located, including temperature, humidity, and light intensity;

[0073] Energy consumption monitoring unit: includes energy consumption sensors, which are used to monitor the energy consumption data of the equipment in real time, including the power usage, working hours and load changes of the equipment;

[0074] Operation record collection unit: used to record the operation information of the equipment in real time, including equipment startup, shutdown and operation mode switching;

[0075] Data upload unit: used to upload all operating data collected by each unit, including the equipment's working status, environmental parameters, energy consumption data and operation records, to the cloud platform for storage in a structured data format through the network interface to ensure the real-time and accuracy of data transmission; through the detailed configuration of the above-mentioned data acquisition module, it is possible to accurately obtain multi-dimensional data during the operation of various types of equipment, ensuring the comprehensiveness and real-time nature of data input to the equipment management system. Different types of sensors obtain various data related to the equipment and environment through special acquisition units, and structure these data through the data upload unit and transmit them to the cloud platform in real time, thereby providing high-quality basic data for subsequent data processing and analysis.

[0076] The data processing module includes a data cleaning unit, a missing value filling unit, and a data normalization unit; wherein:

[0077] Data cleaning unit: used to remove abnormal and noisy data from the operational data received from the cloud platform. Specifically, it uses the upper and lower limit checking method to set the upper and lower limit ranges and check whether the data points exceed the range. If the data points exceed the range, they are regarded as abnormal data and removed to ensure data quality.

[0078] Missing value filling unit: used to fill in data with missing values. Specifically, it uses interpolation method to fill in missing data values based on the existing adjacent data point values through linear interpolation or polynomial interpolation method to ensure data integrity;

[0079] Data normalization unit: used to normalize the collected data, convert data of different dimensions into a unified standard range, and scale the data to the range of [0,1] or [-1,1] by normalizing the standard deviation of each data item or using the minimum-maximum normalization method to avoid comparison deviations between different dimensions and ensure data consistency. The design of the above data processing module can efficiently clean and supplement the original data in the cloud platform, making the data more accurate and complete. These preprocessing measures provide a high-quality data foundation for subsequent feature extraction and predictive analysis.

[0080] The feature extraction module includes a trend feature extraction unit, a periodic feature extraction unit, a mutation feature extraction unit, and a feature vector generation unit; wherein:

[0081] Trend feature extraction unit: used to identify and extract the long-term trend features of equipment operation from preprocessed operating data, determine the overall trend of equipment performance changes over time by fitting time series data, and generate trend parameters reflecting changes in equipment operating status; trend features can be fitted through a linear regression model, the formula is as follows: y(t) = β0 + β1t + (t), where y(t) is the equipment performance value at time t, β0 is the intercept, β1 is the trend coefficient, and (t) is the error term. The β1 parameter obtained by fitting reflects the long-term trend changes in equipment operation;

[0082] Periodic feature extraction unit: Uses the Fourier transform algorithm to perform frequency domain analysis on pre-processed operating data. By converting time series data into a spectrum, the periodic frequency and amplitude of the equipment operation are determined, and frequency parameters reflecting the periodic behavior of the equipment are generated. The periodic characteristics can be expressed by the Fourier transform formula: Where x(t) is the time series data, X(f) is the spectrum in the frequency domain, f is the frequency, and N is the length of the time series data. By analyzing the significant frequency components in the spectrum, the periodic characteristics of the device can be obtained.

[0083] Mutation feature extraction unit: used to detect and extract mutation features in equipment operation from pre-processed operation data. By calculating the moving average and its difference of the data series, it identifies the mutation points in the data and generates mutation parameters that reflect abnormal fluctuations of the equipment.

[0084] Feature vector generation unit: used to integrate various feature parameters extracted by the trend feature extraction unit, the periodic feature extraction unit and the mutation feature extraction unit, and form a feature vector of the equipment status by arranging the various feature parameters in a predetermined order for use by the subsequent prediction and analysis module; through the design of the above feature extraction module, the key features of the equipment operation can be systematically and comprehensively extracted from the preprocessed operation data; the trend feature extraction unit accurately captures the long-term change trend of the equipment performance, the periodic feature extraction unit effectively identifies the periodic behavior in the equipment operation, and the mutation feature extraction unit timely detects abnormal fluctuations in the equipment operation; the comprehensive generation of these features into a feature vector provides a high-dimensional, comprehensive and representative description of the equipment status for the subsequent prediction and analysis module.

[0085] The mutation feature extraction unit includes:

[0086] Calculate the moving average: Select a fixed window size w from the preprocessed running data sequence and calculate the moving average of the data in the window. The calculation formula for the moving average is:

[0087] in, is the moving average at time t, x(i) is the original data at the i-th moment in the time series, w is the size of the sliding window, and the moving average is used to smooth the original data in order to better identify potential mutation points;

[0088] Calculate the difference: By calculating the difference of the data series, the severity of the data change, especially the mutation point, is detected. The difference formula is: Among them, Δx(t) is the difference value at time t, is the moving average at time t; a significant change in the differential value may indicate a sudden change in the operating status of the equipment;

[0089] Identify mutation points: Based on the calculated difference value, set a threshold δ. When the difference value is greater than the threshold δ, it is considered that the data has mutated, and the moment is marked as the mutation point.

[0090] Generate mutation parameters: By comparing the detected mutation points with the difference between the original operating data and the moving average, a mutation parameter reflecting abnormal fluctuations in the equipment is generated. The expression is: p(t) = |Δx(t)|, where p(t) is the mutation parameter at the mutation point t, reflecting the magnitude of the change in the equipment state. The design of the mutation feature extraction unit can accurately identify mutation points in the equipment operating data and timely capture drastic changes in the equipment state.

[0091] The feature vector generation unit includes:

[0092] Feature parameter sorting: Let the trend parameter generated by the trend feature extraction unit be T, the frequency parameter generated by the periodic feature extraction unit be F, and the mutation parameter generated by the mutation feature extraction unit be P; and sort the feature parameters in the following order: first arrange the trend feature parameter T, then arrange the periodic feature parameter F, and finally arrange the mutation feature parameter P, to form an ordered feature sequence {T, F, P};

[0093] Normalization of feature parameters: Normalize the sorted feature parameters using the minimum-maximum normalization method to scale each feature parameter to the range of [0, 1]. The normalization formula is as follows:

[0094] Among them, T norm ,F norm and P norm are the normalized trend parameter, frequency parameter and mutation parameter, T min ,T max ,F min ,F max ,Pmin and P max are the minimum and maximum values of each characteristic parameter respectively;

[0095] Feature vector construction: Arrange the normalized feature parameters in a predetermined order to construct the feature vector V of the device state. The expression is: V = [T norm , F norm , P norm ], where V is the characteristic vector of the equipment status, which includes the trend characteristics, periodic characteristics and mutation characteristics of the equipment operation; through the design of the above-mentioned characteristic vector generation unit, the multi-dimensional characteristic parameters can be systematically integrated into a unified characteristic vector, ensuring the comprehensiveness and consistency of the equipment status description.

[0096] The prediction and analysis module includes a data classification unit, a state recognition unit, and a result generation unit; wherein:

[0097] Data classification unit: Based on the feature vectors generated by the feature extraction module, the K-means clustering algorithm is used to perform unsupervised classification of the equipment's operating status data to form several categories, each of which represents a different mode of the equipment's operating status;

[0098] State recognition unit: Based on the clustering results generated by the data classification unit, it analyzes the feature centers of each category and determines whether the device is in a normal state or a potential fault state by comparing the distance between the device's real-time feature vector and the feature centers of each category.

[0099] Result generation unit: used to output the classification results in the form of structured data for reference by the optimization decision module; through the design of the above-mentioned prediction and analysis module, the K-means clustering algorithm can be effectively used to accurately classify the equipment operating status, distinguish between normal operation and potential fault status, enhance the fault prediction capability and operational reliability of the intelligent equipment management system, and reduce equipment maintenance costs.

[0100] Data classification units include:

[0101] Initialize cluster centers: randomly select k eigenvectors as initial cluster centers, denoted as μ1, μ2, …, μ j ;

[0102] Iteratively assign eigenvectors: For each eigenvector V i , calculate its relationship with each cluster center μ j The Euclidean distance is: Among them, V i,m is the eigenvector V i The mth feature, μ j,m is the cluster center μ jThe mth feature of V, M is the dimension of the feature vector, i Assigned to the cluster with the nearest cluster center;

[0103] Update cluster center: For each cluster j, calculate the average value of all feature vectors assigned to the cluster and update the cluster center μ j , the formula is: Among them, N j is the number of eigenvectors in cluster j, V i is the i-th eigenvector assigned to the corresponding cluster j;

[0104] Convergence determination: Repeat the above steps of iteratively assigning feature vectors and updating cluster centers until the cluster centers no longer change or the predetermined number of iterations is reached, thereby forming the final k cluster categories. Through the detailed design of the above data classification unit, the K-means clustering algorithm can be used to efficiently and accurately classify the equipment operating status.

[0105] The state recognition unit includes:

[0106] Feature vector acquisition: Get the real-time feature vector V of the device real ,The feature vector is generated by the feature vector generation unit, and ,includes the trend feature, periodic feature and mutation feature of the ,device;

[0107] Distance calculation: For each cluster center, calculate the real-time feature vector V real The Euclidean distance D from the cluster center j , the formula is: Among them, D j is the real-time feature vector V real and cluster center μ j The Euclidean distance between them; M is the dimension of the feature vector; V real,m is the real-time feature vector V real The mth feature of j,m is the cluster center μ j The mth feature of

[0108] Shortest distance recognition: Determine the feature vector V in real time real The cluster center with the shortest distance μ min , its expression is: μ min =argmin j D j , where μ min is the real-time feature vector V real The cluster center with the shortest distance;

[0109] State judgment: According to the cluster center μ minThe category label to which the equipment belongs is used to judge the current operating status of the equipment. The category label is pre-defined by the data classification unit in the K-means clustering process and is used to clearly distinguish normal operation from potential fault status. The category label includes normal category and fault category, which correspond to normal status and potential fault status respectively. Through the design of the above-mentioned status recognition unit, the real-time operating status of the equipment can be accurately matched with the pre-defined normal and fault categories to ensure accurate judgment of the equipment status.

[0110] The optimization decision module includes an operation parameter adjustment unit, an operation schedule optimization unit, a preventive maintenance plan formulation unit, and a strategy integration implementation unit; among which:

[0111] Operation parameter adjustment unit: used to adjust the equipment's operating parameters based on the equipment's operating status identified by the predictive analysis module. When the equipment is in a normal state, the unit optimizes the equipment's power and speed to improve its energy efficiency and work efficiency. When the equipment is in a potential fault state, the unit limits the equipment's operating load or temporarily reduces its operating parameters to ensure its safe operation.

[0112] Operation schedule optimization unit: This unit is used to optimize the operation schedule of the equipment according to the equipment's operating status and usage requirements. When the equipment is in a normal state, the operation schedule is arranged by maximizing the equipment's utilization rate. When the equipment is in a potential failure state, the operation schedule is adjusted to reduce the equipment's continuous operating load, avoid aggravating failures caused by excessive use, and reserve time for maintenance work.

[0113] Preventive Maintenance Plan Development Unit: Based on the identification results of equipment in potential failure states, a preventive maintenance plan is developed to extend the equipment's service life and reduce the occurrence of failures. When the equipment is in a potential failure state, specific maintenance content is specified, including arranging equipment inspection, maintenance, and parts replacement. When the equipment is in normal condition, a regular maintenance plan is developed based on the equipment's maintenance history to ensure long-term stable operation of the equipment.

[0114] Strategy integration and implementation unit: used to integrate various optimization strategies formulated by the operating parameter adjustment unit, the operating schedule optimization unit and the preventive maintenance plan formulation unit, and implement specific optimization operations through the cloud platform; specifically, when the equipment is in a normal state, the optimized operating parameters and operating schedule are integrated, and adjustment instructions are sent to the equipment through the cloud platform to achieve efficient operation of the equipment; when the equipment is in a potential fault state, the adjusted operating parameters, optimized operating schedule and preventive maintenance plan are integrated, and corresponding instructions and notifications are sent to the equipment and maintenance personnel through the cloud platform to ensure safe operation and timely maintenance of the equipment; through the detailed design of the above-mentioned optimization decision-making module, the system can formulate scientific and effective optimization operation strategies according to the real-time operating status of the equipment.

[0115] The present invention encompasses any alternatives, modifications, equivalents, and solutions that fall within the spirit and scope of the present invention. To provide a thorough understanding of the present invention, specific details are described in detail below in connection with the preferred embodiments of the present invention, but those skilled in the art will be able to fully understand the present invention without these detailed descriptions. Furthermore, to avoid unnecessary confusion regarding the essence of the present invention, well-known methods, processes, procedures, components, and circuits have not been described in detail.

[0116] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. The intelligent device management system based on cloud platform is characterized by: It includes data acquisition module, data processing module, feature extraction module, prediction analysis module and optimization decision module; among which: Data acquisition module: used to collect the operating data of each smart device in real time through various sensors and network interfaces, including the device's working status, environmental parameters, energy consumption data and operation records, and upload the collected operating data in a structured format to the cloud platform; Data processing module: used to pre-process the operating data in the cloud platform, including data cleaning, missing value filling and data normalization; Feature extraction module: Utilizes statistical analysis and signal processing algorithms to extract relevant features from pre-processed operating data, including trend features, periodic features, and mutation features of equipment operation, and generates a feature vector of the equipment status. Prediction and analysis module: Based on the feature vectors generated by the feature extraction module, the K-means clustering algorithm is used to classify the operating status of the equipment and identify whether the equipment is in a normal state or a potential fault state; Optimization decision module: Based on the operating status identified by the predictive analysis module, it formulates an optimized operation strategy for the equipment, including adjusting the equipment's operating parameters, optimizing the equipment's operating schedule, and formulating a preventive maintenance plan; The feature extraction module includes a trend feature extraction unit, a periodic feature extraction unit, a mutation feature extraction unit, and a feature vector generation unit; wherein: Trend feature extraction unit: used to identify and extract the long-term trend features of equipment operation from pre-processed operation data, determine the overall trend of equipment performance changes over time by fitting time series data, and generate trend parameters reflecting changes in equipment operation status; Periodic feature extraction unit: This unit uses the Fourier transform algorithm to perform frequency domain analysis on pre-processed operating data. By converting time series data into a frequency spectrum, it determines the periodic frequency and amplitude of the equipment operation and generates frequency parameters that reflect the periodic behavior of the equipment. Mutation feature extraction unit: used to detect and extract mutation features in equipment operation from pre-processed operation data. By calculating the moving average and its difference of the data series, it identifies the mutation points in the data and generates mutation parameters that reflect abnormal fluctuations of the equipment. Feature vector generation unit: used to integrate various feature parameters extracted by the trend feature extraction unit, the periodic feature extraction unit and the mutation feature extraction unit, and form a feature vector of the device state by arranging the various feature parameters in a predetermined order.

2. The cloud platform-based intelligent device management system according to claim 1, characterized in that: The data acquisition module includes a sensor acquisition unit, an environmental data acquisition unit, an energy consumption monitoring unit, an operation record acquisition unit, and a data upload unit; wherein: Sensor acquisition unit: including temperature sensor, humidity sensor, pressure sensor, current sensor and vibration sensor, used to collect working status data of smart devices in real time, including temperature, humidity, pressure, current and vibration parameters of the device; Environmental data acquisition unit: includes environmental sensors for real-time acquisition of parameter data of the environment where the device is located, including temperature, humidity, and light intensity; Energy consumption monitoring unit: includes energy consumption sensors, which are used to monitor the energy consumption data of the equipment in real time, including the power usage, working hours and load changes of the equipment; Operation record collection unit: used to record the operation information of the equipment in real time, including equipment startup, shutdown and operation mode switching; Data upload unit: used to upload all operating data collected by each unit, including the equipment's working status, environmental parameters, energy consumption data and operation records, to the cloud platform for storage in a structured data format through the network interface.

3. The cloud platform-based intelligent device management system according to claim 1, characterized in that: The data processing module includes a data cleaning unit, a missing value filling unit and a data normalization unit; wherein: Data cleaning unit: used to remove abnormal and noisy data from the operational data received from the cloud platform. Specifically, it uses the upper and lower limit checking method to set the upper and lower limit ranges and check whether the data points exceed the range. If the data points exceed the range, they are regarded as abnormal data and are eliminated. Missing value filling unit: used to fill in data with missing values. Specifically, it uses interpolation method to fill in missing data values based on the existing adjacent data point values through linear interpolation or polynomial interpolation method to ensure data integrity; Data normalization unit: used to normalize the collected data, convert data of different dimensions into a unified standard range, and scale the data to the range of [0,1] or [-1,1] by normalizing the standard deviation of each data item or the minimum-maximum normalization method to avoid comparison deviations between different dimensions.

4. The cloud platform-based intelligent device management system according to claim 1, characterized in that: The mutation feature extraction unit includes: Calculate the moving average: Select a fixed window size from the preprocessed running data series , and calculate the moving average of the data in the window. The calculation formula of the moving average is: ,in, For in time The moving average of time, For the time series The original data at each moment, is the size of the sliding window; Calculate the difference: By calculating the difference of the data series, the severity of the data change, especially the mutation point, is detected. The difference formula is: ,in, For in time The difference value at time, For in time Moving average of time; Identify mutation points: set thresholds based on calculated differential values , when the difference value is greater than the threshold When , it is considered that the data has mutated, and this moment is marked as the mutation point; Generate mutation parameters: Generate mutation parameters that reflect abnormal fluctuations of the equipment by comparing the difference between the detected mutation points and the moving average. The expression is: ,in, Mutation point The mutation parameters at .

5. The cloud platform-based intelligent device management system according to claim 4, characterized in that: The feature vector generating unit includes: Feature parameter sorting: Assume that the trend parameter generated by the trend feature extraction unit is , the frequency parameter generated by the periodic feature extraction unit is And the mutation parameters generated by the mutation feature extraction unit are ; And sort the characteristic parameters, the specific order is to sort the trend characteristic parameters first , followed by the periodic characteristic parameters , and finally arrange the mutation feature parameters , forming an ordered feature sequence ; Normalization of feature parameters: Normalize the sorted feature parameters and use the minimum-maximum normalization method to scale each feature parameter to within the scope of Feature vector construction: Arrange the normalized feature parameters in a predetermined order to construct the feature vector of the device state .

6. The cloud platform-based intelligent device management system according to claim 1, characterized in that: The prediction and analysis module includes a data classification unit, a state identification unit, and a result generation unit; wherein: Data classification unit: Based on the feature vectors generated by the feature extraction module, the K-means clustering algorithm is used to perform unsupervised classification of the equipment's operating status data to form several categories, each of which represents a different mode of the equipment's operating status; State recognition unit: Based on the clustering results generated by the data classification unit, it analyzes the feature centers of each category and determines whether the device is in a normal state or a potential fault state by comparing the distance between the device's real-time feature vector and the feature centers of each category. Result generation unit: used to output the classification results in the form of structured data for reference by the optimization decision module.

7. The cloud platform-based intelligent device management system according to claim 6, characterized in that: The data classification unit includes: Initialize cluster centers: randomly select The eigenvectors are used as the initial cluster centers, expressed as ; Iteratively assign eigenvectors: For each eigenvector , calculate its relationship with each cluster center The Euclidean distance is: ,in, is the eigenvector No. Features, is the cluster center No. Features, is the dimension of the feature vector, Assigned to the cluster with the nearest cluster center; Update cluster centers: For each cluster , calculate the average value of all feature vectors assigned to the cluster and update the cluster center , the formula is: ,in, Cluster The number of eigenvectors in , Assigned to the corresponding cluster No. feature vectors; Convergence judgment: Repeat the above steps of iteratively assigning feature vectors and updating cluster centers until the cluster centers no longer change or the predetermined number of iterations is reached to form the final Cluster categories.

8. The cloud platform-based intelligent device management system according to claim 7, characterized in that: The state recognition unit includes: Feature vector acquisition: Get the real-time feature vector of the device , the feature vector is generated by a feature vector generating unit; Distance calculation: For each cluster center, calculate the real-time feature vector Euclidean distance from the cluster center , the formula is: ,in, is the real-time feature vector and cluster centers The Euclidean distance between is the dimension of the feature vector; is the real-time feature vector No. Features is the cluster center No. Features Shortest distance recognition: Determining and real-time feature vectors The cluster center with the shortest distance , whose expression is: ; Status judgment: based on cluster center The category label to which the device belongs is used to determine the current operating status of the device. The category label is pre-defined by the data classification unit in the K-means clustering process and includes a normal category and a fault category, which correspond to the normal state and the potential fault state respectively.

9. The cloud platform-based intelligent device management system according to claim 1, characterized in that: The optimization decision module includes an operation parameter adjustment unit, an operation schedule optimization unit, a preventive maintenance plan formulation unit, and a strategy integration implementation unit; wherein: Operation parameter adjustment unit: used to adjust the equipment's operating parameters based on the equipment's operating status identified by the predictive analysis module. When the equipment is in a normal state, the unit optimizes the equipment's power and speed to improve its energy efficiency and work efficiency. When the equipment is in a potential fault state, the unit limits the equipment's operating load or temporarily reduces its operating parameters to ensure its safe operation. Operation schedule optimization unit: used to optimize the operation schedule of the equipment according to the operation status and usage requirements of the equipment. When the equipment is in a normal state, the operation schedule of the equipment is arranged by maximizing the utilization rate of the equipment. When the equipment is in a potential failure state, the operation schedule of the equipment is adjusted to reduce the continuous operation load of the equipment. Preventive Maintenance Plan Development Unit: Develops preventive maintenance plans based on the identification of potential equipment failures. When equipment is in a potential failure state, it specifies specific maintenance content, including arranging equipment inspection, maintenance, and parts replacement. When the equipment is in a normal state, it develops a regular maintenance plan based on the equipment's maintenance history. Strategy integration and implementation unit: used to integrate various optimization strategies formulated by the operating parameter adjustment unit, operating schedule optimization unit and preventive maintenance plan formulation unit, and implement specific optimization operations through the cloud platform.

Images