Electromechanical equipment self-adaptive intelligent early warning system based on multi-source sensing data

Abstract

The invention belongs to the technical field of electromechanical equipment operation and maintenance, and discloses an electromechanical equipment self-adaptive intelligent early warning system based on multi-source sensing data. The system is composed of a multi-source sensing module, an edge data acquisition and preprocessing module, a data cleaning and multi-dimensional feature extraction module, an equipment state dynamic modeling module, an intelligent fault prediction and trend analysis module, a self-adaptive early warning threshold generation and dynamic adjustment module, and an intelligent decision and remote cooperation module. The system is composed of a multi-source sensing module, an intelligent low-carbon operation and maintenance management and control module and a digital twin system integration and full-period mapping module, multiple sensors are deployed through the multi-source sensing module to acquire multi-dimensional data of equipment, cleaning and calibration are performed through the edge data acquisition and preprocessing module, deep processing is performed through the data cleaning and multi-dimensional feature extraction module, and multi-dimensional feature extraction is performed through the multi-source sensing module. The data integrity and accuracy are ensured; and data are quickly transmitted among the modules, so that the monitoring system can accurately present the running state of the equipment in real time.

Description

Technical Field

[0001] The present invention belongs to the technical field of electromechanical equipment operation and maintenance, and specifically relates to an adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data. Background Art

[0002] In modern industrial production, the stable operation of electromechanical equipment is crucial to production efficiency and economic benefits. However, its maintenance and monitoring face many challenges. At present, the maintenance of electromechanical equipment mainly relies on two modes: post-maintenance and regular maintenance. Post-maintenance is passive maintenance, and repair work is only carried out after the equipment fails and stops. The maintenance time is seriously delayed, which not only leads to interruption of production process, but also may cause a series of chain reactions such as order delays and customer loss due to excessive downtime. Although regular maintenance is proactive maintenance, it adopts a one-size-fits-all fixed-cycle maintenance method. Regardless of the actual working conditions of the equipment, the same maintenance operations are performed. This not only causes a waste of manpower, material and financial resources, but also due to the lack of accurate judgment of individual differences and real-time status of equipment, it is difficult to detect potential failures in advance and cannot effectively prevent equipment accidents.

[0003] In terms of monitoring technology, although some systems have introduced single sensor monitoring methods such as vibration and electrical parameters, the data collected by a single sensor can only reflect the operating information of a certain aspect of the equipment, and it is difficult to fully present the true operating status of the equipment, resulting in the inability to detect equipment abnormalities in a timely and accurate manner. More importantly, the existing monitoring systems generally ignore the complex nonlinear coupling characteristics between multi-source data and are unable to effectively separate fault signals from non-fault signals, resulting in a high false alarm rate and serious early warning delays, making it difficult to meet the stringent requirements for efficient and safe operation of electromechanical equipment. Summary of the Invention

[0004] The purpose of the present invention is to provide an adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data to solve the problems raised in the above background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solutions: an adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data, the system comprising a multi-source sensor perception module, an edge data acquisition and preprocessing module, a data cleaning and multi-dimensional feature extraction module, an equipment status dynamic modeling module, an intelligent fault prediction and trend analysis module, an adaptive early warning threshold generation and dynamic adjustment module, an intelligent decision-making and remote collaboration module, an intelligent low-carbon operation and maintenance management module, and a digital twin system integration and full-cycle mapping module; Multi-source sensing module: Using wired or wireless communication, it uses a variety of sensors to collect multi-source data from electromechanical equipment, including mechanical vibration and electrical parameters, to fully reflect the equipment's operating status and provide raw data for the edge computing module. Edge data collection and preprocessing module: After receiving the raw data, it completes initial screening, format unification, time synchronization, and abnormal signal filtering at the local edge, reducing the pressure on the cloud and outputting standardized data streams for further processing by the data cleaning module; Data cleaning and multi-dimensional feature extraction module: Deeply cleans pre-processed data, uses multiple methods to remove interference, extracts key feature parameters such as time domain and frequency domain, and forms a multi-dimensional fault feature set, laying the foundation for the state modeling module; Equipment status dynamic modeling module: Based on characteristic parameters, combined with finite element simulation and other technologies, a dynamic model is established to track equipment status trends, calculate health indicators, and provide a basic basis for fault diagnosis for the intelligent fault prediction module; Intelligent fault prediction and trend analysis module: This module uses the CNN-LSTM model to integrate health indicators and historical data, predicts fault trends, outputs health scores and fault probabilities, and drives the adaptive warning threshold module to adjust thresholds. Adaptive warning threshold generation and dynamic adjustment module: Based on health scores, forecast trends and other data, it uses models such as Bayesian dynamic regression to generate and update thresholds, passes dynamic thresholds to the intelligent decision-making module, and adjusts warning strategies; Intelligent decision-making and remote collaboration module: Determines equipment status based on dynamic thresholds, pushes warning information, supports remote interaction and manual intervention, and transmits adjusted operation strategies to the low-carbon operation and maintenance module to achieve intelligent management and control; Intelligent low-carbon operation and maintenance management module: Optimizes equipment scheduling based on operation strategies, reduces energy consumption, and extends lifespan. It also provides real-time feedback of operation and maintenance results and equipment status to the digital twin module to assist in full lifecycle management. Digital twin system integration and full-cycle mapping module: Aggregates information from multiple modules to build a digital twin, realizes visualization and analysis functions, and feeds virtual-reality mapping data back to the fault prediction and threshold adjustment module to form an optimization closed loop.

[0006] Preferably, the multi-source sensing module includes: (1) Data collection methods and types: Using wired and wireless communication technologies, deploy multiple types of sensors such as vibration, voltage, and thermal imaging to perceive the operating status of electromechanical equipment in real time. Starting from mechanical vibration, electrical parameters, and thermodynamic state, multi-source data is collected in all directions. For example, vibration sensors capture the vibration characteristics of equipment to determine the condition of mechanical components, and thermal imagers monitor temperature distribution to prevent overheating failures, providing rich data for equipment status analysis; (2) Multi-source data acquisition model: Construct a multi-source data acquisition model and use a weighted fusion strategy to process multi-source heterogeneous sensor data. Through this model, different types of data can be integrated to give full play to the advantages of each sensor and improve data accuracy and reliability. Multi-source data acquisition model formula: ; Where: For the The total amount of sensor data of each device (multi-source fusion result); For the The data collected by the jth sensor of a device at time t; For the The weight coefficient of the jth sensor of a device is set according to the sensor type; The total number of sensors for each device; (3) Data transmission and function: The original, multimodal sensor data after collection and fusion has the characteristics of fully reflecting the operating status of the equipment and will be transmitted downward to the edge computing module. This data serves as the basic input for subsequent data processing, providing support for modules such as edge data collection and preprocessing, feature extraction, and state modeling. It plays a key starting role in the data processing process of the entire electromechanical equipment adaptive intelligent early warning system.

[0007] Preferably, the edge data collection and preprocessing module includes: (1) Data reception and preliminary screening: The edge data acquisition and preprocessing module receives multi-source heterogeneous data uploaded by the multi-source perception module. These data contain information on various aspects of equipment operation, but may contain errors or invalid records. The module starts the preliminary screening process and quickly identifies and eliminates obviously erroneous data based on preset rules, laying the foundation for subsequent accurate processing and ensuring the reliability of the starting point of data processing; (2) Time synchronization preprocessing and precise calibration: To solve the problem of time asynchrony of multi-source sensor data, the module uses a time synchronization preprocessing formula for precise calibration to achieve precise alignment of data in the time dimension, ensuring the accuracy and effectiveness of data analysis; Time synchronization preprocessing formula: ; Where: Time-synchronized sensor data; The original collected data; is the time delay correction value of the jth sensor; (3) Format unification, anomaly filtering, and data transfer: After completing time synchronization, the module unifies the data formats of different sensor types to ensure that the data conforms to standard specifications. At the same time, an anomaly detection algorithm based on statistics and signal characteristics is used to scan the data in real time to eliminate sudden interference and noise. After processing, the standardized data stream is output to the data cleaning module, which not only ensures data quality but also significantly reduces the data processing load on the cloud.

[0008] Preferably, the data cleaning and multi-dimensional feature extraction module includes: (1) Deep data cleaning operation: This module receives edge preprocessed data and starts the deep cleaning process. It uses the denoising algorithm to initially remove random noise in the data, and then uses wavelet denoising filtering. Based on the specific wavelet denoising filtering principle, it analyzes the signal in detail at different scales and accurately filters out interference signals. Combined with the time domain averaging method, it performs comprehensive processing on multiple collected data, effectively enhances the effective signal, eliminates various non-working condition interference and abnormal points, and greatly improves the data quality; Wavelet denoising filter formula: ; Where: is the filtered signal; is the wavelet coefficient of the approximate signal; is the wavelet coefficient of the detail signal; is the basis function of scale A; is the basis function of scale j; j is the scale parameter of wavelet decomposition, representing the different scale levels of signal analysis, and its value range is [A, J], where A is the starting scale and J is the maximum decomposition scale; (2) Multi-dimensional feature parameter extraction: After completing the deep cleaning of the data, this module starts to extract key feature parameters. From the time domain dimension, basic statistical features such as mean and variance are extracted to characterize the basic characteristics of the signal; in the frequency domain, the spectrum distribution is deeply analyzed to mine the frequency information of equipment vibration and electrical signals; through time-frequency domain analysis, the time and frequency dimension information are integrated to present the dynamic change of signal frequency over time; for thermal imaging data, features such as temperature extremes and temperature gradients are extracted to comprehensively evaluate the thermal status of the equipment; (3) Fault feature set output and transfer: After data cleaning and multi-dimensional feature extraction, this module integrates and generates multi-dimensional fault feature sets. These feature sets contain key information about the equipment's operating status and can accurately reflect the characteristics of potential equipment failures. This multi-dimensional fault feature set is transferred downward to the state modeling module, providing indispensable key data support for the subsequent construction of a dynamic equipment state model, helping to achieve accurate modeling and analysis of equipment status.

[0009] Preferably, the device state dynamic modeling module includes: (1) Multi-technology integration to build dynamic models: This module is based on the key feature parameters output by the data cleaning and multi-dimensional feature extraction module, and comprehensively uses finite element simulation, electromechanical system physical modeling and digital twin real-time synchronization technology. Finite element simulation starts from the details of the equipment structure and discretizes the changes in mechanical and thermal performance under different working conditions; electromechanical system physical modeling builds a mathematical model based on the principle characteristics of the equipment to accurately describe the operation process; digital twin real-time synchronization ensures that the virtual model is consistent with the actual equipment status, and together builds a dynamic equipment operation status model; (2) Model-based in-depth tracking of equipment status: The established dynamic model has a strong ability to track equipment status. By continuously analyzing the changes in various parameters in the model, it can capture the status trends of the equipment during operation in real time. Combined with the equipment health index model, the equipment status is comprehensively evaluated from multiple dimensions, and the equipment's operating status is converted into intuitive and measurable indicators. Potential performance degradation or abnormal trends can be discovered in a timely manner, providing a forward-looking basis for equipment maintenance; Equipment health index model formula: ; In the formula The device health index (HealthIndex); For the The weight of each state parameter; For the state parameters (such as vibration RMS, current, etc.); is the total number of state parameters; (3) Health indicator output and subsequent connection: Based on real-time tracking of equipment status change trends, this module calculates real-time equipment health indicators based on the equipment health index model. These indicators directly reflect the current health level of the equipment and are passed down to the intelligent fault prediction module. As the basic criteria for fault diagnosis, they provide an important reference for the intelligent fault prediction module, helping it to more accurately predict potential equipment failures and promote the efficient operation of the entire early warning system.

[0010] Preferably, the intelligent fault prediction and trend analysis module includes: (1) Deep learning model fusion technology: This module innovatively uses the CNN-LSTM multi-level deep learning model to deeply integrate the real-time equipment health indicators and historical fault data output by the equipment status model. With its powerful feature extraction capabilities, CNN can automatically mine local and spatial features in the data and accurately identify the fault modes contained in the equipment operation data; LSTM has unique advantages in processing time series data and can effectively learn the long-term change trend of equipment status in the time dimension. The two complement each other and lay a solid technical foundation for fault prediction; (2) Equipment failure trend prediction process: Based on the fusion of data and models, this module conducts in-depth predictions on the potential failure trends of equipment in the future. By extracting features from real-time health indicators and historical data using CNN, combined with time series analysis of equipment operation trends using LSTM, subtle signals such as equipment performance degradation and abnormal parameter changes are captured from massive data. The future development trend of equipment failures is deduced, and a health score and failure probability reflecting the health status of the equipment are dynamically output, providing forward-looking predictions for equipment operation and maintenance. (3) Prediction results drive threshold optimization: The equipment failure trend data and failure probability output by this module will be passed down to the adaptive warning threshold module. These prediction results serve as the key basis for dynamically adjusting the warning threshold, driving the threshold to be optimized in real time based on the actual operating conditions and performance change trends of the equipment. Through this linkage mechanism, the warning system can closely follow the actual operation of the equipment, improve the timeliness and accuracy of the warning, and achieve dynamic and precise control of the operating risks of electromechanical equipment.

[0011] Preferably, the adaptive warning threshold generation and dynamic adjustment module includes: (1) Multi-source data integration and analysis: This module collects the equipment health score and forecast trend data output by the intelligent fault prediction and trend analysis module, and integrates the equipment's real-time operating parameters and environmental change information. These multi-dimensional data comprehensively reflect the equipment's operating status, providing a solid data foundation for threshold generation and adjustment, and are a key prerequisite for achieving accurate early warning. Through in-depth analysis of the data, subtle changes in the equipment status are captured, indicating the direction for subsequent threshold adjustments; (2) Dynamic model-driven threshold generation: Using Bayesian dynamic regression and exponential smoothing models, early warning and alarm thresholds are generated based on integrated multi-source data. The Bayesian dynamic regression model combines prior knowledge with real-time data to achieve dynamic optimization of thresholds through probabilistic reasoning; the exponential smoothing model uses weighted averaging of historical data to quickly respond to changes in data trends. Both models are based on the principle of dynamic threshold adjustment formulas, which enable thresholds to closely follow fluctuations in equipment status, always maintain adaptability to the actual operating conditions of the equipment, and improve the accuracy and timeliness of early warnings; Dynamic threshold adjustment formula: ; Where: is the updated threshold; The threshold value at the previous moment is the dynamic adjustment coefficient; Predict the current probability of failure; is the preset baseline probability; (3) Threshold transfer and strategy optimization: The generated dynamic threshold is promptly transferred to the intelligent decision-making module, becoming an important basis for determining the equipment status. Based on the dynamic threshold, the intelligent decision-making module adjusts the early warning strategy in real time. When the equipment operating parameters reach the threshold, the corresponding early warning measures are quickly triggered. This linkage mechanism ensures that the early warning system can accurately reflect the operating status of the equipment under different working conditions and environments, effectively avoiding false alarms and missed alarms, and achieving precise control of the operating risks of electromechanical equipment.

[0012] Preferably, the intelligent decision-making and remote collaboration module includes: (1) Dynamic threshold-driven intelligent decision-making: This module uses the dynamic thresholds transmitted by the adaptive warning threshold generation and dynamic adjustment module as a benchmark to accurately determine the real-time operating status of electromechanical equipment. Once the equipment operating parameters reach the warning or alarm threshold, the system immediately triggers the intelligent push mechanism and sends the corresponding level of warning or alarm information to relevant personnel to ensure that equipment abnormalities can be detected in time and buy time for subsequent disposal; (2) Remote interaction and strategy adjustment: This module supports remote WEB and APP interaction, giving users the ability to remotely control devices. Users can not only view the device's operating status in real time and receive warning information, but also manually intervene in the device's operating strategy as needed, such as flexibly adjusting operating parameters and switching working modes. At the same time, the system can also automatically optimize the operating strategy based on preset rules, and the adjusted strategy is passed down to the low-carbon operation and maintenance module, comprehensively improving the safety and reliability of equipment operation.

[0013] Preferably, the intelligent low-carbon operation and maintenance control module includes: (1) Strategy-based intelligent operation and maintenance scheduling: This module starts the intelligent operation and maintenance scheduling mechanism based on the adjustment strategy output by the intelligent decision-making and remote collaboration modules. It gives priority to high-efficiency equipment, distributes the load scientifically and rationally according to the equipment performance and work requirements, and automatically shuts down equipment with high energy consumption and low efficiency. By optimizing and adjusting the equipment operation combination and working parameters, it reduces unnecessary energy consumption, effectively reduces the overall energy consumption of the system while ensuring the normal operation of the equipment, and thus extends the service life of the equipment; (2) Carbon emission optimization and data feedback: During the intelligent operation and maintenance scheduling process, the system carbon emission optimization model is used to calculate and accurately control carbon emissions during the operation and maintenance phase in real time. This model comprehensively considers multiple factors such as equipment operating energy consumption and working hours, dynamically monitors carbon emissions, and optimizes the operation and maintenance strategy accordingly. At the same time, the low-carbon operation and maintenance results and equipment operating status are synchronously fed back to the digital twin module in real time, providing detailed data for the full life cycle management of the equipment, helping to achieve the low-carbon and intelligent operation and maintenance goals of electromechanical equipment.

[0014] System carbon emission optimization model formula: ; Where: is the total carbon emissions of the current system; is the carbon emission factor of the i-th equipment; is the power load of the i-th device; is the time step; n is the total number of devices involved in carbon emission calculation in the system; Preferably, the digital twin system integration and full-cycle mapping module includes: (1) Multi-source data fusion to build digital twins: This module aggregates in real time the equipment operation data collected by the multi-source sensing module, the fault prediction results of the intelligent fault prediction and trend analysis module, and the operation and maintenance status information of the intelligent low-carbon operation and maintenance control module. Based on this, a highly realistic equipment digital twin is constructed to fully map the status of the equipment throughout its life cycle. The equipment operation details are presented through three-dimensional visualization, the fault evolution is analyzed with the help of historical data, and intelligent optimization analysis is carried out to propose operation and maintenance plans to achieve comprehensive control of the equipment status; (2) Error correction drives closed-loop optimization: During the operation of the digital twin, the deviation between the actual state of the equipment and the virtual model is monitored and dynamically corrected using the virtual-real synchronization error correction formula to ensure that the virtual model accurately reflects the actual state of the equipment. The corrected equipment virtual-real mapping data is fed back to the fault prediction and threshold adjustment module to provide more accurate data support, forming a full life cycle adaptive optimization closed loop, and continuously improving the system warning accuracy and the scientific nature of equipment operation and maintenance; Virtual and real synchronization error correction formula: ; Where: is the virtual-real synchronization error; is the actual status parameter of the device; Predict state parameters for digital twins.

[0015] The beneficial effects of the present invention are as follows: 1. The present invention deploys multiple sensors through a multi-source sensing perception module to collect multi-dimensional data of the equipment, which is cleaned and calibrated by the edge data acquisition and preprocessing module, and deeply processed by the data cleaning and multi-dimensional feature extraction module to ensure data integrity and accuracy; the digital twin system integration and full-cycle mapping module aggregates data in real time to build a virtual model to achieve comprehensive perception of the equipment status; data is quickly transmitted between modules, and edge computing collaboration from data acquisition to preliminary processing greatly shortens the response time, enabling the monitoring system to present the equipment operation status in real time and accurately, providing a reliable basis for subsequent decision-making, and effectively improving the real-time performance and data integrity of the monitoring system.

[0016] 2. The present invention constructs an equipment operation model through a dynamic modeling module of equipment status combined with finite element simulation and digital twin technology. The intelligent fault prediction and trend analysis module adopts a CNN-LSTM model to integrate equipment health indicators and historical fault data. CNN identifies fault modes and LSTM learns status trends. The combination of the two accurately mines data features and predicts potential faults in advance. The physical model assists in verifying the prediction results, avoiding misjudgments caused by single data or algorithms, significantly reducing the false alarm rate and missed alarm rate, achieving accurate early warning of equipment failures, and ensuring the safe and stable operation of electromechanical equipment.

[0017] 3. The present invention uses an adaptive warning threshold generation and dynamic adjustment module based on the results of the intelligent fault prediction and trend analysis module, combined with equipment operating conditions and environmental data, and using Bayesian dynamic regression or exponential smoothing models to dynamically generate warning thresholds. This mechanism breaks the limitations of fixed thresholds and adjusts in real time according to the actual operating status of the equipment. When the equipment operating conditions or the environment change, the threshold is optimized accordingly to ensure accurate adaptation of the warning strategy. The intelligent decision-making and remote collaboration module uses dynamic thresholds to timely determine the equipment status, achieve accurate warnings, comprehensively enhance the system's intelligence and adaptability, and improve the efficiency and reliability of electromechanical equipment operation and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of the adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data of the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] like Figure 1 As shown, an embodiment of the present invention provides an adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data. The system consists of a multi-source sensor perception module, an edge data acquisition and preprocessing module, a data cleaning and multi-dimensional feature extraction module, an equipment status dynamic modeling module, an intelligent fault prediction and trend analysis module, an adaptive early warning threshold generation and dynamic adjustment module, an intelligent decision-making and remote collaboration module, an intelligent low-carbon operation and maintenance management module, and a digital twin system integration and full-cycle mapping module. Multi-source sensing module: Using wired or wireless communication, it uses a variety of sensors to collect multi-source data from electromechanical equipment, including mechanical vibration and electrical parameters, to fully reflect the equipment's operating status and provide raw data for the edge computing module. Edge data collection and preprocessing module: After receiving the raw data, it completes initial screening, format unification, time synchronization, and abnormal signal filtering at the local edge, reducing the pressure on the cloud and outputting standardized data streams for further processing by the data cleaning module; Data cleaning and multi-dimensional feature extraction module: Deeply cleans pre-processed data, uses multiple methods to remove interference, extracts key feature parameters such as time domain and frequency domain, and forms a multi-dimensional fault feature set, laying the foundation for the state modeling module; Equipment status dynamic modeling module: Based on characteristic parameters, combined with finite element simulation and other technologies, a dynamic model is established to track equipment status trends, calculate health indicators, and provide a basic basis for fault diagnosis for the intelligent fault prediction module; Intelligent fault prediction and trend analysis module: This module uses the CNN-LSTM model to integrate health indicators and historical data, predicts fault trends, outputs health scores and fault probabilities, and drives the adaptive warning threshold module to adjust thresholds. Adaptive warning threshold generation and dynamic adjustment module: Based on health scores, forecast trends and other data, it uses models such as Bayesian dynamic regression to generate and update thresholds, passes dynamic thresholds to the intelligent decision-making module, and adjusts warning strategies; Intelligent decision-making and remote collaboration module: Determines equipment status based on dynamic thresholds, pushes warning information, supports remote interaction and manual intervention, and transmits adjusted operation strategies to the low-carbon operation and maintenance module to achieve intelligent management and control; Intelligent low-carbon operation and maintenance management module: Optimizes equipment scheduling based on operation strategies, reduces energy consumption, and extends lifespan. It also provides real-time feedback of operation and maintenance results and equipment status to the digital twin module to assist in full lifecycle management. Digital twin system integration and full-cycle mapping module: Aggregates information from multiple modules to build a digital twin, realizes visualization and analysis functions, and feeds virtual-reality mapping data back to the fault prediction and threshold adjustment module to form an optimization closed loop.

[0021] Based on multi-source sensor data, an adaptive intelligent early warning system for electromechanical equipment is constructed to achieve efficient operation and maintenance and accurate early warning. At the data perception level, the multi-source sensing module deploys multiple sensors to collect multi-dimensional data from equipment. This data is processed layer by layer through the edge data acquisition and preprocessing module and the data cleaning and multi-dimensional feature extraction module to ensure data integrity and accuracy. The digital twin system integration and full-cycle mapping module constructs a virtual model to achieve comprehensive perception of equipment status, and rapid data transmission improves real-time monitoring. In terms of fault prediction, the equipment status dynamic modeling module integrates finite element simulation and digital twin technology. The intelligent fault prediction and trend analysis module uses a CNN-LSTM model, combining equipment health indicators and historical data to accurately identify fault characteristics and reduce false alarm and missed alarm rates. In terms of early warning mechanisms, the adaptive early warning threshold generation and dynamic adjustment module combines multi-source data to dynamically generate thresholds, breaking the limitations of fixed thresholds. The intelligent decision-making and remote collaboration module uses this to accurately determine equipment status, enhancing the system's intelligence and adaptability and ensuring reliable equipment operation.

[0022] The multi-source sensing module utilizes a combination of wired and wireless communication technologies, deploying various types of sensors, including vibration, voltage, and thermal imaging, to provide real-time and comprehensive sensing of the operating status of electromechanical equipment from multiple dimensions, including mechanical vibration, electrical parameters, and thermodynamic state. For example, vibration sensors can capture equipment vibration characteristics to determine the condition of mechanical components, while thermal imagers can monitor temperature distribution to prevent overheating. A weighted fusion strategy is used to process multi-source heterogeneous data, fully leveraging the strengths of each sensor to effectively improve data accuracy and reliability. The collected and fused raw, multimodal sensor data fully represents the device's operating status and is transmitted to the edge computing module, providing essential input for subsequent data preprocessing, feature extraction, and state modeling, supporting the efficient operation of the entire early warning system.

[0023] Among them, the edge data acquisition and preprocessing module is mainly responsible for the preprocessing of multi-source heterogeneous data. This module first takes over the data uploaded by the multi-source perception module. Although this data covers many aspects of equipment operation information, there are errors or invalid records. The module uses a preliminary screening program to quickly eliminate invalid data according to preset rules to ensure the reliability of data processing. To address the time asynchrony issue of multi-source sensor data, we use a time synchronization preprocessing formula to perform precise calibration, achieving precise alignment of data in the time dimension, providing effective support for subsequent analysis. The module unifies the data format to make it conform to standard specifications, and uses anomaly detection algorithms based on statistics and signal characteristics to filter sudden interference and noise in real time, outputting standardized data streams to the data cleaning module. While ensuring data quality, it significantly reduces the cloud data processing load and lays a solid foundation for data processing of the entire early warning system.

[0024] The data cleaning and multi-dimensional feature extraction module is mainly responsible for deep processing of the edge pre-processed data. This module first uses a denoising algorithm to preliminarily remove random noise, and then based on the wavelet denoising filtering principle, it conducts a detailed analysis of the signal at different scales to accurately filter out interference signals. At the same time, combined with the time domain averaging method, it enhances effective signals and eliminates abnormal data, significantly improving data quality. After completing deep cleaning, the module extracts key feature parameters from multiple dimensions such as time domain, frequency domain, time-frequency domain and thermal imaging data to comprehensively characterize the operating status of the equipment. The multi-dimensional fault feature set obtained through cleaning and extraction is integrated and output, and then passed to the state modeling module to provide core data for building a dynamic equipment state model, laying a solid foundation for subsequent accurate judgment of equipment status and prediction of potential faults.

[0025] The equipment status dynamic modeling module is responsible for accurately modeling the equipment status and outputting key indicators. This module is based on the key characteristic parameters output by the data cleaning and multi-dimensional feature extraction module, integrating finite element simulation, electromechanical system physical modeling, and digital twin real-time synchronization technology. Finite element simulation analyzes the performance changes of equipment under working conditions from structural details, physical modeling accurately describes the operation process based on principles, and digital twins ensure real-time synchronization between the virtual model and the equipment status, jointly constructing a dynamic equipment operation status model. Based on this model, the module continuously and in-depth tracks device status. Using the device health index model, it comprehensively assesses device health from multiple dimensions, capturing performance degradation or abnormal trends in real time. The resulting real-time device health indicators are then passed down to the intelligent fault prediction module, providing the core basis for fault diagnosis and enabling the early warning system to accurately predict potential equipment failures.

[0026] The intelligent fault prediction and trend analysis module optimizes forward-looking judgments and early warnings for equipment failures. This module innovatively utilizes the CNN-LSTM multi-level deep learning model to deeply integrate real-time health indicators and historical fault data output by the equipment status model. This module leverages the CNN's ability to automatically extract local and spatial features and accurately identify fault modes, while combining the LSTM's advantages in processing time series data and learning long-term trends in equipment status. Deeply mine subtle abnormal signals in equipment operating data, deduce potential future failure trends, and dynamically output health scores and failure probabilities. The prediction results are passed to the adaptive early warning threshold module, which optimizes the threshold in real time based on equipment operating conditions and performance trends. This enables precise control of electromechanical equipment operating risks and improves the timeliness and accuracy of the early warning system.

[0027] The adaptive warning threshold generation and dynamic adjustment module first collects the equipment health score and predicted trend data output by the intelligent fault prediction and trend analysis module, and integrates the equipment's real-time operating parameters and environmental change information. By deeply analyzing multi-dimensional data, it accurately captures subtle changes in equipment status. Using Bayesian dynamic regression and exponential smoothing models, combined with dynamic threshold adjustment formulas, we generate warning and alarm thresholds based on integrated data, ensuring that the thresholds closely align with the dynamic changes in equipment operating status. The resulting dynamic thresholds are passed to the intelligent decision-making module, serving as the basis for determining equipment status and driving real-time adjustments to the warning strategy, ensuring that the warning system effectively avoids false alarms and missed alerts under varying operating conditions and environments.

[0028] Among them, the intelligent decision-making and remote collaboration module uses the dynamic threshold transmitted by the adaptive warning threshold generation and dynamic adjustment module as a basis to accurately determine the real-time operating status of the equipment. When the equipment parameters reach the threshold, the intelligent push mechanism is quickly triggered to issue warning or alarm information in time.

[0029] Supporting remote web and app interaction, users can monitor device status in real time, receive alerts, and manually intervene in operational strategies. The system can also automatically optimize strategies based on preset rules. Adjusted strategies are then passed to the low-carbon operation and maintenance module, comprehensively improving the safety and reliability of device operations.

[0030] Among them, the intelligent low-carbon operation and maintenance management and control module starts intelligent operation and maintenance scheduling based on the adjustment strategy output by the intelligent decision-making and remote collaboration module, gives priority to high-efficiency equipment, reasonably distributes the load, shuts down inefficient equipment, and optimizes equipment operation combinations and parameters, effectively reducing energy consumption and extending equipment life while ensuring the normal operation of the equipment.

[0031] Leveraging a systemic carbon emissions optimization model, the system integrates factors such as equipment energy consumption and operating hours to accurately calculate and control carbon emissions during the operation and maintenance phase in real time, dynamically optimizing the operation and maintenance strategy accordingly. Simultaneously, low-carbon operation and maintenance results and equipment operating status are fed back to the digital twin module, providing data support for the full lifecycle management of the equipment.

[0032] Among them, the digital twin system integration and full-cycle mapping module aggregates in real time the equipment operation data of the multi-source sensing perception module, the prediction results of the intelligent fault prediction and trend analysis module, and the operation and maintenance status information of the intelligent low-carbon operation and maintenance control module to build a highly realistic digital twin and realize three-dimensional visualization, historical backtracking and intelligent optimization of the equipment's full life cycle status.

[0033] During operation, the virtual-reality synchronization error correction formula is used to monitor and correct the deviation between the actual status of the equipment and the virtual model in real time to ensure accurate mapping. The corrected virtual-reality mapping data is fed back to the fault prediction and threshold adjustment module to form a closed-loop optimization, continuously improving the system's early warning accuracy and scientific operation and maintenance.

[0034] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0035] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data, characterized by: The system consists of a multi-source sensing module, an edge data acquisition and preprocessing module, a data cleaning and multi-dimensional feature extraction module, an equipment status dynamic modeling module, an intelligent fault prediction and trend analysis module, an adaptive warning threshold generation and dynamic adjustment module, an intelligent decision-making and remote collaboration module, an intelligent low-carbon operation and maintenance management module, and a digital twin system integration and full-cycle mapping module. Multi-source sensing module: uses sensors to collect multi-source data of electromechanical equipment, including mechanical vibration and electrical parameters, to obtain raw data; Edge data acquisition and preprocessing module: performs preliminary screening, format unification, time synchronization, and abnormal signal filtering on the received multi-source data, and outputs preprocessed data to the data cleaning and multi-dimensional feature extraction module; Data cleaning and multi-dimensional feature extraction module: Deeply cleans pre-processed data, removes interference, extracts key feature parameters in the time domain and frequency domain, and forms a multi-dimensional fault feature set; Equipment status dynamic modeling module: builds an equipment dynamic operation model based on the multi-dimensional fault feature set, calculates equipment health indicators in real time and transmits them to the intelligent fault prediction and trend analysis module; Intelligent Fault Prediction and Trend Analysis Module: This module integrates equipment health indicators and historical data, uses deep learning models to predict equipment failure trends, generates equipment health scores and failure probabilities, and transmits these scores to the Adaptive Warning Threshold Generation and Dynamic Adjustment Module in real time. Adaptive warning threshold generation and dynamic adjustment module: Based on health scores and predicted trend data, dynamic thresholds are generated using the Bayesian dynamic regression model; Intelligent decision-making and remote collaboration module: Determines the equipment operating status based on dynamic thresholds, generates operation and maintenance adjustment strategies, and pushes them to the intelligent low-carbon operation and maintenance management module in real time. It also pushes early warning information to the remote management terminal to support remote interaction and manual intervention. Intelligent low-carbon operation and maintenance control module: Optimizes equipment scheduling based on operation strategies, and provides real-time feedback of operation and maintenance results and equipment status to the digital twin system integration and full-cycle mapping module; Digital twin system integration and full-cycle mapping module: Builds a digital twin of the equipment, monitors the virtual-reality synchronization error of the equipment in real time, and transmits the virtual-reality mapping data back to the intelligent fault prediction and trend analysis module after correction, forming a closed-loop optimization control chain.

2. The adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data according to claim 1, characterized in that: The multi-source sensing module includes: (1) Data collection methods and types: Using wired and wireless communication technologies, deploying vibration, voltage, and thermal imaging sensors to perceive the operating status of electromechanical equipment in real time, starting from mechanical vibration, electrical parameters, and thermodynamic state, collecting multi-source data in all directions; (2) Multi-source data acquisition model: Construct a multi-source data acquisition model, use a weighted fusion strategy to process multi-source heterogeneous sensor data, and integrate different types of data through this model; Multi-source data acquisition model formula: ； Where: is the total amount of sensor data of the i-th device; The data collected by the jth sensor of the i-th device at time t; is the weight coefficient of the jth sensor of the i-th device; The total number of sensors for each device; (3) Data transmission and function: The original, multimodal sensor data after collection and fusion is transmitted to the preprocessing module.

3. The adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data according to claim 1, characterized in that: The edge data acquisition and preprocessing module includes: (1) Data reception and preliminary screening: If there are errors or invalid records in the received multi-source heterogeneous data, the preliminary screening program is initiated to quickly identify and eliminate obviously erroneous data based on preset rules; (2) Time synchronization preprocessing and precise calibration: To solve the time asynchrony problem of multi-source sensor data, a time synchronization preprocessing formula is used for precise calibration; Time synchronization preprocessing formula: ； Where: Time-synchronized sensor data; The original collected data; is the time delay correction value of the jth sensor; (3) Format unification and anomaly filtering: After completing time synchronization, the module unifies the data formats of different types of sensors, uses anomaly detection algorithms based on statistics and signal characteristics, scans data in real time, and eliminates sudden interference and noise.

4. The adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data according to claim 1, characterized in that: The data cleaning and multi-dimensional feature extraction module includes: (1) Deep data cleaning operation: After receiving the data processed by the preprocessing module, the denoising algorithm is used to initially remove the random noise in the data. Then, with the help of wavelet denoising filtering, the signal is carefully analyzed at different scales to accurately filter out interference signals; Wavelet denoising filter formula: ； Where: is the filtered signal; is the wavelet coefficient of the approximate signal; is the wavelet coefficient of the detail signal; is the basis function of scale A; is the basis function of scale j; j is the scale parameter of wavelet decomposition, and its value range is [A, J], where A is the starting scale and J is the maximum decomposition scale; (2) Multi-dimensional feature parameter extraction: After completing deep data cleaning, the frequency domain is used to analyze the spectrum distribution and mine the frequency information of equipment vibration and electrical signals; time and frequency domain analysis is used to integrate time and frequency dimension information; for thermal imaging data, temperature extremes and temperature gradient features are extracted; (3) Fault feature set output and transmission: After data cleaning and multi-dimensional feature extraction, a multi-dimensional fault feature set is generated and integrated, and the multi-dimensional fault feature set is transmitted to the equipment status dynamic modeling module.

5. The adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data according to claim 1, characterized in that: The device status dynamic modeling module includes: (1) Multi-technology integration to build dynamic models: Based on the key feature parameters output by the data cleaning and multi-dimensional feature extraction modules, finite element simulation, electromechanical system physical modeling and digital twin real-time synchronization technology are comprehensively used. Finite element simulation discretizes and analyzes the changes in mechanical and thermal performance under different working conditions; electromechanical system physical modeling builds a mathematical model based on the principle characteristics of the equipment; and digital twin real-time synchronization ensures that the virtual model is consistent with the actual equipment status; (2) Model-based in-depth tracking of equipment status: By analyzing the changes in various parameters in the model and combining it with the equipment health index model, the equipment's operating status is converted into intuitive and measurable indicators, allowing for timely detection of potential performance degradation or abnormal trends; Equipment health index model formula: ； Where: is the equipment health index; For the The weight of each state parameter; For the Status parameters; is the total number of state parameters; (3) Health indicator output and subsequent connection: Real-time equipment health indicators are calculated based on the equipment health index model. These indicators are passed to the intelligent fault prediction module as the basic criteria for fault diagnosis.

6. The adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data according to claim 1, characterized in that: The intelligent fault prediction and trend analysis module includes: (1) Deep learning model fusion technology: Using the CNN-LSTM deep learning model, deep integration of real-time equipment health indicators and historical fault data; (2) Equipment failure trend prediction process: Deeply predict the potential failure trend of the equipment in the future. By extracting the features of real-time health indicators and historical data through CNN and combining it with LSTM to analyze the time series of equipment operation trends, the future failure development trend of the equipment is deduced. (3) Prediction results drive threshold optimization: Equipment failure trend data and failure probability are transmitted to the adaptive warning threshold module, and the driving threshold is optimized in real time based on the actual operating conditions and performance change trends of the equipment.

7. The adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data according to claim 1, characterized in that: The adaptive warning threshold generation and dynamic adjustment module includes: (1) Multi-source data integration and analysis: Collect the equipment health score and prediction trend data output by the intelligent fault prediction and trend analysis module, and integrate the equipment's real-time operating parameters and environmental change information. Through in-depth analysis of the data, it can capture subtle changes in the equipment status. (2) Dynamic model-driven threshold generation: The Bayesian dynamic regression model achieves dynamic optimization of the threshold through probabilistic reasoning; the exponential smoothing model uses weighted average of historical data to quickly respond to changes in data trends. Both models are based on the principle of dynamic threshold adjustment formula, which enables the threshold to closely follow the fluctuation of device status; Dynamic threshold adjustment formula: ； Where: is the updated threshold; is the threshold value at the previous moment; is the dynamic adjustment coefficient; Predict the current probability of failure; is the preset baseline probability; (3) Threshold transfer and strategy optimization: The generated dynamic threshold is transferred to the intelligent decision-making module. Based on the dynamic threshold, the intelligent decision-making module adjusts the early warning strategy in real time. When the equipment operating parameters reach the threshold, the corresponding early warning measures are quickly triggered.

8. The adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data according to claim 1, characterized in that: The intelligent decision-making and remote collaboration module includes: (1) Dynamic threshold-driven intelligent decision-making: Based on the received dynamic threshold, the real-time operating status of the electromechanical equipment is determined. When the equipment operating parameters reach the warning or alarm threshold, the system immediately triggers the intelligent push mechanism; (2) Remote interaction and strategy adjustment: Users can view the equipment operation status in real time, receive early warning information, and manually intervene in the equipment operation strategy. The system automatically optimizes the operation strategy based on preset rules, and the adjusted strategy is transmitted to the low-carbon operation and maintenance module.

9. The adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data according to claim 1, characterized in that: The intelligent low-carbon operation and maintenance control module includes: (1) Policy-based intelligent operation and maintenance scheduling: Based on the received adjustment strategy, high-efficiency equipment is prioritized, load is distributed according to equipment performance and work requirements, and equipment with high energy consumption and low efficiency is automatically shut down; (2) Carbon emission optimization and data feedback: With the help of the system carbon emission optimization model, the carbon emissions in the operation and maintenance phase are calculated and accurately controlled in real time, and the operation and maintenance strategy is optimized accordingly. The low-carbon operation and maintenance results and equipment operating status are synchronously fed back to the digital twin module in real time; System carbon emission optimization model formula: ； Where: is the total carbon emissions of the current system; is the carbon emission factor of the i-th equipment; is the power load of the i-th device; is the time step; n is the total number of devices involved in carbon emission calculation in the system.

10. The adaptive intelligent early warning system for electromechanical equipment based on multi-source sensor data according to claim 1, characterized in that: The digital twin system integration and full-cycle mapping module includes: (1) Multi-source data fusion to build digital twins: real-time aggregation of equipment operation data, fault prediction results, and operation and maintenance status information to build equipment digital twins and fully map the status of the entire life cycle of the equipment; (2) Error correction-driven closed-loop optimization: During the operation of the digital twin, the deviation between the actual state of the equipment and the virtual model is monitored and dynamically corrected using the virtual-real synchronization error correction formula. The corrected equipment virtual-real mapping data is fed back to the fault prediction and threshold adjustment module. Virtual and real synchronization error correction formula: ； Where: is the virtual-real synchronization error; is the actual status parameter of the device; Predict state parameters for digital twins.

Images