Whole-process management and control system for electric power engineering construction project

Abstract

The invention relates to the technical field of electric power engineering management and control, and discloses an electric power engineering construction project whole process management and control system. The system comprises an engineering parameter monitoring module, a risk analysis module, a progress control module, a quality evaluation module, an early warning output module and a background supervision terminal. The engineering parameter monitoring module collects multi-dimensional real-time parameters through a sensor array and sends the multi-dimensional real-time parameters to the risk analysis module and the background supervision terminal. The risk analysis module generates a progress control instruction according to the deviation between the real-time parameter and a preset standard parameter and sends the progress control instruction to the progress control module. The progress control module executes progress adjustment operation, and the quality evaluation module monitors the adjustment process and generates a quality signal. The early warning output module receives the quality early warning signal, generates early warning information and sends the early warning information to the background supervision terminal, and the background supervision terminal gives an alarm. According to the system, collaborative operation of engineering parameter real-time monitoring, risk analysis, progress adjustment, quality evaluation and early warning is realized, and a whole-process management and control closed loop is formed.

Description

Technical Field

[0001] The present invention relates to the technical field of power engineering management and control, and in particular to a full-process management and control system for power engineering construction projects. Background Art

[0002] In the field of power engineering construction, full-process project management involves multiple dimensions, including construction progress, project quality, and safety risks. Each link is closely interconnected and influenced by complex factors. Traditional management and control models rely heavily on manual inspections and empirical judgment, making it difficult to capture and dynamically analyze project parameters in real time. For example, during the installation of transmission lines, if changes in key parameters such as tower foundation settlement and conductor tension are not monitored in a timely manner, construction deviations may occur. Fluctuations in indicators such as equipment installation accuracy and concrete curing environment during substation construction can also potentially impact project quality.

[0003] In existing technologies, some control measures use single-parameter monitoring equipment, which can only track a certain indicator of a specific link and lack the ability to integrate and analyze multi-dimensional data. When deviations occur in engineering parameters, they cannot be quickly linked to schedule adjustment needs, resulting in a disconnect between the construction rhythm and actual working conditions. At the same time, risk assessments often rely on fixed threshold judgments, which are difficult to adapt to engineering variables under different geological conditions and climatic environments, and are prone to misjudgments or omissions. In addition, there is a lack of closed-loop linkage between quality assessment and schedule control, and the execution effect after schedule adjustment cannot be fed back in a timely manner, which may cause secondary rework and affect overall construction efficiency.

[0004] With the expansion of the scale of power projects and the increase in technical complexity, the shortcomings of traditional management and control models in data processing timeliness, risk response speed and multi-link coordination have become increasingly prominent. An integrated management and control solution is needed that can achieve real-time parameter monitoring, intelligent risk analysis, dynamic progress adjustment and closed-loop quality evaluation. Summary of the Invention

[0005] The purpose of the present invention is to provide a whole-process control system for electric power engineering construction projects to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above objectives, the present invention provides a full-process management and control system for power engineering construction projects, the system comprising: Engineering parameter monitoring module, risk analysis module, progress control module, quality assessment module, early warning output module and background supervision terminal; The engineering parameter monitoring module collects multi-dimensional parameters in the power engineering construction process in real time through the sensor array, and sends the real-time parameter values ​​to the risk analysis module and the background monitoring terminal, which visualizes the real-time parameter values; The risk analysis module obtains parameter deviations based on real-time parameter values ​​and preset standard parameter values, calculates progress control instructions based on the parameter deviations using a random forest algorithm, and sends the progress control instructions to the progress control module; The progress control module performs construction progress adjustment operations according to the progress control instructions to optimize the construction progress; the quality assessment module monitors the progress adjustment operation process, judges the execution performance of the progress adjustment operation, generates a quality pass signal or a quality warning signal based on the performance, and sends the quality pass signal or quality warning signal to the warning output module; When the warning output module receives the quality warning signal, it generates warning information and sends it to the background monitoring end, and the background monitoring end issues an alarm according to the warning information.

[0007] Preferably, the specific evaluation process of the quality evaluation module includes: marking the corresponding progress adjustment operation as an abnormal operation or a standard operation through a dynamic threshold evaluation mechanism, setting an evaluation period, and calculating the ratio of the number of abnormal operations to the total number of progress adjustment operations within the evaluation period to obtain the abnormal operation rate; when the abnormal operation rate exceeds the preset abnormal threshold, generating a quality warning signal; when the abnormal operation rate does not exceed the preset abnormal threshold, performing a difference calculation between the execution efficiency value of the corresponding progress adjustment operation and the intermediate value of the preset efficiency benchmark value to obtain the absolute value to obtain the efficiency evaluation value, and marking the ratio of the stability value of the corresponding operation process to the preset stability benchmark as the stability evaluation value; performing a weighted average calculation on the efficiency evaluation values ​​of all progress adjustment operations within the evaluation period to obtain a comprehensive efficiency value, and performing a weighted average calculation on the stability evaluation values ​​of all progress adjustment operations within the evaluation period to obtain a comprehensive stability value; obtaining a quality output value by fusing the abnormal operation rate, the comprehensive efficiency value and the comprehensive stability value, generating a quality warning signal when the quality output value exceeds the preset quality threshold, and generating a quality qualified signal when the quality output value does not exceed the preset quality threshold.

[0008] Preferably, the specific evaluation process of the dynamic threshold evaluation mechanism includes: obtaining the moment when the progress control module receives the progress control instruction and recording it as the starting moment, and obtaining the moment when the progress control module completes the corresponding progress adjustment operation and recording it as the ending moment, and marking the time interval between the starting moment and the ending moment as the monitoring period; marking the ratio of the progress control instruction to the monitoring period as the execution efficiency value, and obtaining the stability value through operation smoothness analysis; when the execution efficiency value is not in the preset efficiency value range or the stability value exceeds the preset stability limit, marking the corresponding progress adjustment operation as an abnormal operation; when the execution efficiency value is in the preset efficiency value range and the stability value does not exceed the preset stability limit, marking the corresponding progress adjustment operation as a standard operation.

[0009] Preferably, the specific analysis process of the operation stability analysis includes: constructing a two-dimensional coordinate system with the time axis as the horizontal axis and the real-time parameter value as the vertical axis, extracting the change trajectory of the engineering construction parameters during the corresponding progress adjustment operation, placing the change trajectory in the coordinate system, and the starting point of the change trajectory is located on the vertical axis; setting multiple sampling points on the change trajectory, marking the longitudinal distance between adjacent sampling points as the change amplitude value, performing cluster analysis on all the change amplitude values ​​to obtain a fluctuation evaluation value, and marking the frequency of occurrence of the change amplitude value that is not within the preset change amplitude range as an abnormal frequency value; and obtaining the stability value of the corresponding progress adjustment operation by combining the fluctuation evaluation value and the abnormal frequency value.

[0010] Preferably, the early warning output module is communicatively connected to the safety auxiliary analysis module, and the early warning output module sends a quality qualified signal to the safety auxiliary analysis module; when the safety auxiliary analysis module receives the quality qualified signal, it performs a safety auxiliary evaluation on the progress control module, generates a safety alarm signal or an auxiliary qualified signal through the evaluation, and sends the safety alarm signal or the auxiliary qualified signal to the background supervision end; when the background supervision end receives the safety alarm signal, it issues a safety alarm.

[0011] Preferably, the specific evaluation process of the safety auxiliary evaluation includes: collecting the vibration intensity and noise level values ​​of the progress control module during its operation; when the vibration intensity or noise level value exceeds the corresponding preset safety standard, judging that the progress control module is in an abnormal state; obtaining the duration of the progress control module in the abnormal state during the evaluation period and calculating the ratio thereof with the total operation time during the evaluation period to obtain the abnormal duration rate, marking the number of occurrences of the single duration of the progress control module in the abnormal state during the evaluation period exceeding the corresponding preset single duration threshold as an over-limit abnormal value, and marking the longest single duration of the progress control module in the abnormal state during the evaluation period as an abnormal duration value; obtaining a safety evaluation value by weighted calculation of the abnormal duration rate, the over-limit abnormal value and the abnormal duration value; generating a safety alarm signal when the safety evaluation value exceeds the preset safety threshold; generating an auxiliary qualified signal when the safety evaluation value does not exceed the preset safety threshold.

[0012] Preferably, the background supervision end is communicatively connected to the equipment diagnosis feedback module. When a quality warning signal or a safety alarm signal is generated, the equipment diagnosis feedback module performs an elimination diagnosis analysis on the progress control module, determines through analysis whether an equipment elimination signal is generated, and sends the equipment elimination signal to the background supervision end; when the background supervision end receives the equipment elimination signal, it issues an equipment elimination alarm.

[0013] Preferably, the specific analysis process of the equipment diagnosis feedback module includes: obtaining the production date of the progress control module, calculating the time difference between the current date and the production date to obtain the equipment life value, and marking the total time that the progress control module is in operation in the historical stage as the cumulative operation value; obtaining the environmental risk value of the progress control module through analysis, and marking the number of times that the maintenance interval of the progress control module in the historical stage exceeds the preset maintenance interval standard as the maintenance abnormality coefficient; obtaining the diagnostic evaluation value by linearly combining the equipment life value, the cumulative operation value, the environmental risk value and the maintenance abnormality coefficient; generating an equipment elimination signal when the diagnostic evaluation value exceeds the preset diagnostic threshold.

[0014] Preferably, the method for analyzing and obtaining the environmental risk value includes: collecting the temperature parameters and humidity parameters of the environment in which the progress control module is located, marking the deviation value of the temperature parameter compared to the set standard temperature value as the temperature evaluation value, and marking the deviation value of the humidity parameter compared to the set standard humidity value as the humidity evaluation value; collecting the pollutant concentration of the environment in which the progress control module is located and marking it as the pollution evaluation value; obtaining the environmental risk value by normalizing the temperature evaluation value, the humidity evaluation value and the pollution evaluation value; when the environmental risk value exceeds the preset environmental risk threshold, judging that the progress control module is in a high-risk state; obtaining the total time that the progress control module is in a high-risk state in the historical stage and marking it as the environmental risk value.

[0015] Preferably, the risk analysis module is communicatively connected to the environment adaptation module; the environment adaptation module collects meteorological data, geological change data and equipment operating condition data of the construction site in real time; the environment adaptation module analyzes the correlation between environmental parameters and historical accident data through a convolutional neural network, and outputs an environmental disturbance coefficient; the risk analysis module couples the environmental disturbance coefficient with the parameter deviation and dynamically adjusts the risk judgment threshold of the random forest algorithm.

[0016] Compared with the prior art, the present invention has the following beneficial effects: Through the collaborative operation of multiple modules, a comprehensive management and control system covering the entire construction process has been established. The engineering parameter monitoring module utilizes a sensor array to simultaneously collect multi-dimensional real-time parameters, breaking the limitations of traditional single-parameter monitoring. This allows back-end supervisors to intuitively grasp the real-time status of each project link and achieve a comprehensive understanding of the construction process.

[0017] The risk analysis module uses a random forest algorithm to generate progress control instructions based on the deviation between real-time parameters and preset standard parameters, making the generated progress adjustment instructions more closely aligned with the actual project conditions. This data-based analysis method avoids the subjectivity of relying on empirical judgment and can dynamically adapt to adjustment requirements based on parameter changes, making progress control more targeted.

[0018] The progress control module executes construction schedule adjustments based on instructions, flexibly responding to fluctuations in project parameters to ensure that the construction rhythm matches parameter changes, mitigating schedule delays or advances caused by parameter deviations. The quality assessment module monitors the progress of schedule adjustments in real time, generating corresponding signals based on performance evaluation. This enables instant linkage between schedule adjustments and quality assessments, preventing quality issues from slipping out of control after schedule adjustments.

[0019] Upon receiving a quality warning signal, the early warning output module promptly generates and transmits warning information to the backend supervisory system, enabling supervisors to quickly identify any quality anomalies and facilitate timely intervention. The backend supervisory system's visual display of real-time parameters and response to warning information eliminates reliance on manual inspections, enabling centralized monitoring of project dynamics and improving both the timeliness and accuracy of management and control. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a working principle diagram of the whole-process control system for electric power engineering construction projects according to the present invention; Figure 2 is a flow chart of the dynamic threshold evaluation mechanism; Figure 3 Flowchart for operational stability analysis; Figure 4 Flowchart for equipment diagnostic feedback module analysis; Figure 5 Flowchart for device diagnostic feedback analysis. DETAILED DESCRIPTION

[0021] 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.

[0022] See also Figure 1 The present invention provides a full-process management and control system for power engineering construction projects, which includes: an engineering parameter monitoring module, a risk analysis module, a progress control module, a quality assessment module, an early warning output module, and a background supervision terminal. The specific implementation method is as follows: The engineering parameter monitoring module uses a sensor array deployed at the construction site to collect real-time multi-dimensional parameters during power project construction, including voltage, current, temperature, humidity, mechanical vibration, and the operating status of construction equipment. This sensor array utilizes a distributed network and transmits real-time parameter values ​​to the risk analysis module and backend monitoring via Industrial Internet of Things protocols. The backend monitoring platform integrates a data visualization platform, displaying real-time parameter trends using dynamic charts and 3D modeling. The risk analysis module includes a built-in parameter comparison unit that calculates the deviation between real-time parameter values ​​and thresholds in a pre-set standard parameter library to generate a parameter deviation matrix. A random forest algorithm model, loaded with a training set of historical engineering data, outputs progress control instructions based on the parameter deviation matrix, including acceleration, deceleration, or suspension of construction. The progress control module interacts with construction equipment via an industrial control bus to execute progress adjustments. During the progress adjustment process, the quality assessment module collects execution data, including response delay, operational consistency, and parameter fluctuation characteristics, to generate quality assessment results. The warning output module triggers a multi-level warning mechanism based on the quality assessment results, and the backend monitoring terminal transmits warning information via audible and visual alarms and mobile devices.

[0023] Example 1: See Figure 2 The quality assessment module dynamically monitors and categorizes the execution of the progress control module, employing a periodic assessment mechanism to quantitatively analyze the compliance and stability of construction adjustment operations. The assessment cycle is configured based on the actual needs of the project, with a default setting of 24-hour continuous monitoring. The system automatically records the receipt and execution time of each progress control instruction, accurate to the millisecond, and calculates the actual duration of each adjustment operation through time difference calculation. Data during the monitoring period is stored in a circular buffer, managed using the first-in-first-out principle to ensure the integrity and continuity of real-time data.

[0024] The calculation process of the execution efficiency value comprehensively considers the two dimensions of instruction complexity and actual execution time. The system's built-in instruction complexity classification system divides progress control instructions into three levels: basic, standard, and complex, and each level corresponds to a different time reference value. The degree of deviation between the actual execution time and the reference value is normalized to generate an efficiency score, which is dynamically mapped to a preset efficiency value range. The efficiency value range is automatically matched according to the project type. For example, a wide range is used for substation construction projects, while a narrow range is applicable to transmission line projects. When the execution efficiency value is detected to be outside the reasonable range for the current project type, the system automatically triggers the abnormal operation flag.

[0025] Stability analysis uses a multi-dimensional parameter trajectory monitoring method. The system captures the operating parameters of construction equipment at a high sampling frequency, including key indicators such as robotic arm displacement, hydraulic pressure, and motor speed. These real-time data streams are converted into time series waveforms, forming a continuous change trajectory in a two-dimensional coordinate system. The trajectory analysis process uses sliding window technology, and the window size is adaptively adjusted according to the parameter type. Smaller windows are used for rapidly changing dynamic parameters, while larger windows are suitable for slowly changing steady-state parameters. After the data points in each window are smoothed, the change gradient between adjacent sampling points is calculated, and abnormal fluctuation patterns are identified through clustering algorithms.

[0026] Abnormal operations are determined using compound conditional logic. The system monitors both execution efficiency values ​​and stability indicators. When any indicator exceeds the preset safety boundary, it is determined to be an abnormal operation. The stability boundary value is not a fixed threshold, but a floating range that is dynamically adjusted based on the historical data of equipment operation. The system maintains a dynamic baseline database that records the parameter fluctuation characteristics of various types of equipment under normal operating conditions. The degree of deviation between the real-time monitoring data and the baseline is weighted and calculated to generate a stability score. The statistics of abnormal frequency use a time decay algorithm. Abnormal events that have occurred recently have a higher statistical weight, so that the evaluation results can better reflect the current actual situation.

[0027] A hierarchical weighting strategy is employed during the comprehensive evaluation phase. The system organically integrates the evaluation results of three dimensions: abnormal operation rate, comprehensive efficiency value, and comprehensive stability value. The abnormal operation rate directly reflects the frequency of problems during the schedule adjustment process and occupies a fundamental weight in the overall evaluation. The comprehensive efficiency value is calculated using a time-weighted average algorithm, with efficiency data from recent operations having a higher contribution. The comprehensive stability value uses a sliding window average method to ensure the stability of the evaluation results. After the three-dimensional evaluation data is standardized, different integration weights are assigned according to the specific characteristics of the project. The resulting quality output value comprehensively reflects the execution quality of schedule control operations.

[0028] The early warning trigger mechanism employs a multi-level response strategy. The system issues warnings based on the severity of quality output values, with three levels: observation, warning, and emergency. Observation-level warnings generate only system log records for subsequent analysis; warning-level warnings trigger visual notifications on the backend supervisory system; and emergency-level warnings activate both audible and visual alarms and mobile device notifications. Warning level thresholds are dynamically configured based on project risk assessment results, with increased warning sensitivity in high-risk areas. The system also features an automatic warning upgrade function, automatically raising the warning level when continuous monitoring detects a deteriorating trend in quality indicators.

[0029] The data visualization interface provides a multi-dimensional view of quality monitoring. The backend supervisory platform integrates a professional data analysis dashboard, displaying the spatial distribution of abnormal operations using heat maps, temporal trends of quality indicators using line graphs, and correlations between efficiency and stability values ​​using three-dimensional scatter plots. Operators can use the timeline zoom function to view detailed data for any period. The system supports multi-view linkage analysis, and clicking on a specific abnormal event retrieves a complete operation log and parameter record. All visualization elements support custom configuration, allowing users to adjust display content and layout based on their focus.

[0030] The historical data backtracking function supports quality analysis. The system fully preserves the original data and calculation results of all assessment cycles, establishing a traceable quality archive. Analysts can compare quality performance across different periods and project sections, identifying potential issues through trend analysis. The system provides intelligent comparison tools that automatically highlight indicators with significant changes, helping to identify the causes of quality fluctuations. The data export function allows evaluation results to be generated into a standard report format, including statistical charts of key indicators and textual analysis conclusions, meeting the documentation requirements of project management.

[0031] An adaptive learning mechanism continuously optimizes the evaluation model. The system's built-in machine learning algorithm continuously absorbs new monitoring data and automatically adjusts evaluation parameters and weightings. When significant changes in the engineering environment or equipment configuration are detected, the system initiates a specialized learning mode to quickly adapt to current conditions. The model optimization process utilizes online learning, gradually updating the evaluation rules without interrupting monitoring functionality. The system regularly generates model health reports and recommends parameter adjustments that require manual intervention, ensuring the accuracy and reliability of evaluation results.

[0032] The abnormality diagnosis assistance function provides problem analysis support. When a quality warning is detected, the system automatically initiates a root cause analysis process, locating the possible source of the fault through correlation analysis technology. The diagnostic process comprehensively examines equipment status data, environmental monitoring data, and operation records to generate a diagnostic report containing possible causes and resolution suggestions. The system maintains a library of typical fault cases and uses a similarity matching algorithm to provide reference solutions for the current problem. Diagnostic results are presented in a structured format, including a failure probability assessment, impact analysis, and recommended resolution priorities, assisting managers in making quick decisions.

[0033] Quality assessment results are deeply integrated with the project management system. Generated assessment data is synchronized to the project management system in real time, serving as a crucial basis for project progress and quality control. The system provides standardized data interfaces, enabling the push of quality assessment results to third-party analysis platforms. Assessment data is automatically linked to corresponding project phases and construction units, establishing a complete quality traceability chain. Managers can use a comprehensive query interface to compare quality performance across projects, providing data support for resource allocation and process improvement. The system also supports the customization and sharing of quality assessment templates, facilitating the rapid deployment of consistent assessment standards across similar projects.

[0034] Example 2: See Figure 3 The operational stability analysis unit uses multi-parameter fusion monitoring technology to comprehensively evaluate the operating status of construction equipment. This unit collects the dynamic response characteristics of the mechanical system through a high-precision sensor network. The sampling frequency is differentiated according to the parameter type. Vibration parameters are collected at a high frequency of kHz, while temperature parameters are sampled at the second level. The collected raw signal is digitally filtered and converted into a standardized engineering unit value to eliminate the impact of differences in sensor range and accuracy. The signal processing link includes a noise suppression algorithm to effectively separate the actual operating signal of the equipment from environmental interference components, thereby improving the credibility of the monitoring data.

[0035] Change trajectory analysis utilizes a spatiotemporal correlation modeling approach. The system constructs a multidimensional feature space, mapping time series data into quantifiable operating state trajectories. Trajectory feature extraction utilizes a sliding time window technique, with the window width adaptively adjusted based on the device's response characteristics. A narrow window is used for rapid dynamic processes, while a wide window is used for slow steady-state processes. Feature extraction of each data point within the window generates a state vector, which includes time-domain features such as mean, variance, and crest factor, as well as frequency-domain features derived through a fast Fourier transform. These feature vectors constitute the state space point set for the device's operation, and cluster analysis is used to identify typical operating modes.

[0036] The fluctuation assessment algorithm utilizes a multi-scale analysis approach. The system simultaneously examines the impact of short-term fluctuations and long-term trends on device stability. Short-term fluctuation analysis focuses on parameter changes on a timescale of seconds, deriving the instantaneous rate of change by calculating the differential sequence of adjacent sampling points. Long-term trend analysis focuses on parameter drift on a timescale of hours, extracting the trend term using a polynomial fitting method. The quantitative assessment of the degree of fluctuation comprehensively considers the distribution characteristics of the instantaneous rate of change and the slope of the trend term to generate a comprehensive fluctuation index. This index is normalized and compared with historical normal operation data to determine whether the current fluctuation level is within a reasonable range.

[0037] Abnormal frequency detection utilizes adaptive threshold technology. The system dynamically maintains a baseline for the normal fluctuation range of various parameters. This baseline is not a fixed value, but rather a dynamic range that changes with the device's operating status. The detection algorithm calculates the deviation of parameter values ​​from the dynamic baseline in real time. When the deviation persists for more than a set duration, it is recorded as a valid abnormal event. Abnormal event determination considers both amplitude and duration, avoiding false alarms caused by transient interference while capturing persistent abnormal conditions. The system records the number and duration of abnormal events within each monitoring cycle and calculates an abnormal activity index to reflect the overall operating status of the device.

[0038] The vibration monitoring function of the Safety Assist Analysis module utilizes multi-sensor data fusion technology. A triaxial vibration sensor array is deployed at key equipment locations, and spatial correlation analysis is used to eliminate local interference. After time-frequency analysis of the vibration signal, the energy distribution of characteristic frequency bands is extracted for evaluation. The system establishes a database of equipment vibration characteristics, containing typical vibration patterns under different load conditions. Real-time monitoring data is compared with this database to identify abnormal vibration components. Vibration intensity assessment not only considers amplitude but also analyzes the spectral distribution of vibration energy, providing a more comprehensive picture of the equipment's mechanical condition.

[0039] The noise monitoring system utilizes acoustic fingerprinting technology. An array of acoustic sensors is deployed around the equipment to collect the time-domain waveform and spectral characteristics of operating noise. Using an acoustic feature extraction algorithm, the system builds an acoustic fingerprint database of the equipment's normal operation. Real-time acoustic signals are then pattern-matched against the fingerprint database to calculate a similarity score. Noise level assessment not only focuses on the overall sound pressure level but also analyzes changes in acoustic characteristics within specific frequency bands, enabling early detection of potential faults such as bearing wear and gear mesh anomalies. Acoustic monitoring data is cross-validated with vibration monitoring results to improve the accuracy of anomaly diagnosis.

[0040] Abnormal state determination utilizes a multi-indicator collaborative decision-making mechanism. The system integrates multi-dimensional monitoring data, including vibration intensity, noise level, and temperature changes, and classifies the state using a decision tree algorithm. Multiple buffer levels are implemented during the determination process to prevent frequent state switching caused by measurement fluctuations. Confirmation of an abnormal state requires both duration and indicator severity to prevent misjudgment due to transient interference. The system records the start time, duration, and peak intensity of each abnormal state, creating a complete abnormal event log that provides detailed data for subsequent analysis.

[0041] The safety assessment model is continuously optimized using incremental learning. The system continuously absorbs new monitoring data and automatically adjusts assessment parameters and decision boundaries. The model training process utilizes an online learning algorithm, gradually updating assessment rules without interrupting monitoring functionality. The system regularly self-checks model performance and verifies assessment accuracy through backtesting historical data. When significant changes are detected in the engineering environment or equipment configuration, a dedicated learning mode is automatically triggered to quickly adapt to the new operating conditions. Model optimization records are fully preserved, supporting version rollback and performance comparison.

[0042] Early warning information generation utilizes a hierarchical classification strategy. Based on the severity of safety assessment results, the system categorizes warnings into three levels: caution, warning, and danger. Different warning levels correspond to different response processes and handling authorities. Warnings include detailed diagnostic data and recommended actions, assisting managers in making rapid decisions. Warnings are issued through a multi-channel parallel mechanism, including local audio and visual alarms, mobile device push notifications, and remote monitoring center notifications, ensuring timely dissemination of information. Warning status is displayed in real time on a panoramic project view, visually demonstrating the problem location and impact area.

[0043] The device status visualization interface provides a multi-dimensional monitoring view. The system integrates a professional data analysis dashboard, using trend charts to display historical changes in key parameters, spectrum charts to analyze vibration characteristics, and 3D models to display overall device status. Visual elements support interactive operation, allowing users to click on device models to view detailed monitoring data. The system offers a multi-view comparison function, allowing users to compare the operating status of the same device over time or compare the status of similar devices horizontally. All views support custom configuration, allowing users to adjust the display content and layout based on their focus.

[0044] The historical data management function fully records the status of the equipment throughout its lifecycle. The system automatically archives all monitoring data and assessment results to establish a traceable equipment health profile. Data storage utilizes a hierarchical structure, with raw sampling data, feature extraction results, and assessment conclusions stored separately, balancing the requirements for data integrity and access efficiency. Data retrieval supports combined queries based on multiple criteria, such as time range, equipment type, and anomaly level, allowing for rapid identification of events of interest. The data export function generates standardized equipment status reports, including statistical charts of key parameters and textual analysis conclusions.

[0045] Example 3: See Figure 4The vibration intensity monitoring system uses multi-dimensional signal processing technology to perform refined perception of the operating status of construction machinery. Three groups of orthogonally installed accelerometers are arranged at key parts of the equipment, and the sampling frequency is set to 5120Hz to meet the needs of mechanical vibration feature extraction. The original vibration signal first undergoes analog anti-aliasing filtering and then is digitized by a 24-bit analog-to-digital converter. The digital signal processing link adopts a five-stage cascade filtering architecture to remove irrelevant components such as power supply interference, environmental noise, and high-frequency glitches in turn. The extraction of effective vibration signals adopts an adaptive threshold algorithm, which dynamically adjusts the cutoff frequency according to the equipment speed to ensure that the characteristic vibration components can be accurately captured under different working conditions. The vibration intensity quantification uses an improved envelope demodulation technology to convert the high-frequency resonance signal into a low-frequency envelope waveform that can be intuitively evaluated. After the envelope signal undergoes Hilbert transform, the integral of the square of its amplitude is used to characterize the vibration energy:

[0046] in Indicates the vibration energy index, is the Hilbert transform operator, is the preprocessed vibration signal, to This indicator effectively integrates the vibration amplitude and duration information, and can better reflect the actual damage potential than simple peak detection.

[0047] The noise monitoring subsystem uses acoustic array technology to achieve directional sound pickup. An equilateral tetrahedron array consisting of four microphones is installed within 1 meter of the equipment, and the beamforming algorithm is used to suppress environmental noise interference. The sound pressure level measurement adopts the equivalent continuous A-weighting mode, and a Leq value is calculated every 125 milliseconds. The spectrum analysis adopts 1 / 12 octave resolution, focusing on monitoring the characteristic frequency bands related to the mechanical structure of the equipment. The abnormal noise is identified by the spectrum comparison method, and the difference between the real-time spectrum and the reference template is quantified by the dynamic time warping algorithm. The system maintains a noise feature database, which contains the acoustic characteristics of typical faults such as gear meshing, bearing wear, loosening and collision, and realizes early fault warning through pattern matching.

[0048] The abnormal state determination engine employs a multi-level confidence strategy. Feature extraction is performed on the raw monitoring data to generate a twelve-dimensional feature vector, including vibration energy, noise eigenvalues, and temperature change rate. The determination model first independently evaluates each feature and calculates its normalized distance from the normal range. This information is then integrated into a comprehensive anomaly score using a feature fusion network. This network utilizes a radial basis function neural network architecture, which is adaptable to nonlinear relationships. Determination results are categorized into three levels: normal, caution, and abnormal, each corresponding to a different response strategy. Hysteresis intervals are set for state transitions to avoid frequent state changes caused by measurement fluctuations. A detailed event record is generated for each state change, including metadata such as the triggering characteristics, change magnitude, and duration.

[0049] The safety assessment model utilizes a time series deep learning approach. A long short-term memory network architecture is used to model the temporal evolution of equipment status. The input sequence consists of monitoring features from the past 60 minutes, and the output is a prediction of the status trend over the next 15 minutes. Model training utilizes a curriculum learning strategy, first learning steady-state operating patterns and then gradually introducing transition states and abnormal cases. During online application, the difference between the predicted results and the actual monitoring data is used to assess model confidence. A continued increase in the difference automatically triggers model retraining. The assessment results are output as risk probabilities, with values ​​ranging from 0 to 1 corresponding to different levels of risk, ranging from safe to dangerous.

[0050] The equipment diagnostic feedback system builds a full lifecycle health profile. Static data, such as each device's identity, technical parameters, and maintenance records, is stored alongside real-time monitoring data. Archival data is managed using a time-partitioning strategy, with recent data maintained as high-resolution raw records and historical data gradually aggregated into statistical features. The lifespan prediction model analyzes the impact of multiple factors, including operating intensity, environmental stress, and maintenance quality, to estimate remaining useful life. The diagnostic report generation module automatically extracts trends in key indicators and compares them against benchmark data for similar equipment to identify abnormal aging patterns.

[0051] Environmental risk monitoring utilizes a multi-sensor fusion solution. A temperature and humidity composite sensor is installed in key heat dissipation locations on the equipment, sampling at a 1Hz frequency and achieving measurement accuracy of ±0.5°C and ±3%RH. The pollutant monitoring unit integrates a laser scattering particulate matter sensor and an electrochemical gas sensor to detect PM2.5, SO2, NOx, and other harmful gases. Environmental data preprocessing includes temperature compensation, humidity correction, and cross-interference elimination. Risk value calculation utilizes fuzzy logic, mapping the degree of deviation of each environmental parameter into a membership function. A comprehensive risk rating is generated using an inference rule base.

[0052] The maintenance decision-making system employs a condition-based maintenance strategy. Maintenance demand forecasts comprehensively consider the current equipment status, historical maintenance results, and project schedule constraints. Maintenance tasks automatically generate work instructions containing specific inspection items, operating procedures, and acceptance criteria. The maintenance execution process is recorded in real time via mobile devices, including key operations such as part replacement, parameter adjustment, and anomaly detection. Maintenance effectiveness evaluation compares changes in equipment status indicators before and after maintenance to quantify the effectiveness of maintenance actions. A maintenance knowledge base continuously accumulates typical cases and best practices, supporting similarity-based maintenance solution recommendations.

[0053] The data visualization platform utilizes a hierarchical display design. The top-level panoramic view displays the overall status distribution of all devices, using color coding to intuitively identify the locations of abnormal devices. The second-level device view displays detailed monitoring curves and evaluation results for individual devices, supporting multi-parameter overlay analysis. The bottom-level data view provides raw signal waveforms and spectral characteristics, enabling professionals to conduct in-depth diagnostics. Drill-down operations are supported between views, allowing users to quickly navigate from macro-level status to micro-level details. Display content supports custom layouts and themes to adapt to different user viewing habits.

[0054] The system's integrated architecture utilizes a microservices design pattern. Each functional module operates as an independent service, communicating loosely through a message bus. The data interface adheres to the RESTful specification, providing standard services such as device status query, alarm subscription, and maintenance record synchronization. Interface security utilizes two-way authentication and data encryption to ensure the confidentiality and integrity of monitoring data transmission. The system supports a horizontally scalable architecture, enabling the addition of service instances to accommodate larger-scale monitoring needs. Service health is monitored in real time, and abnormalities automatically trigger failover and recovery processes.

[0055] The user permission system enables refined access control. Role definitions distinguish between different functions, such as system administrators, equipment experts, and field operators. Permissions are assigned down to the level of specific function buttons and data fields. Operational audit logs record key events such as user logins, function usage, and data access. Authentication mechanisms support multiple methods, including passwords, digital certificates, and biometrics. Session management detects abnormal login behavior, such as frequent failed attempts or access at unusual times, and automatically triggers security measures.

[0056] System maintenance mechanisms ensure long-term reliable operation. Automatic update services regularly check for software patches and security updates, deploying them in phases and batches after testing. Performance monitoring dashboards display CPU, memory, network, and other resource usage, providing early warning of potential performance bottlenecks. Capacity planning tools analyze data growth trends and predict storage resource requirements. Disaster recovery plans define emergency procedures for different failure scenarios, and recovery drills are conducted regularly. Service-level agreements clearly define system availability and performance metrics, guiding operational and maintenance priority decisions.

[0057] Example 4: See Figure 5 The equipment diagnostic feedback module enables full lifecycle health management of construction machinery through multi-dimensional data analysis. For example, the system collects complete status records of a certain hydraulic pile driver since its commissioning, creating a structured health profile. This equipment has been in service for over five years on a coastal cross-sea bridge project, accumulating 12,345 operating hours and undergoing three major overhauls and 17 routine maintenance sessions.

[0058] Environmental monitoring data shows that the equipment was exposed to harsh operating conditions of high salt spray and high humidity for extended periods. Temperature and humidity sensors record environmental parameters every ten minutes, with abnormal environmental periods marked according to the following rules: When the temperature exceeds 40°C or the humidity remains above 90% for three consecutive hours, the system automatically registers it as a high-risk period. The pollutant monitoring unit detected multiple incidents of seawater droplet adhesion, with chloride ion concentrations peaking at 285 ppm. The environmental risk value is calculated based on the duration, intensity, and combined effects of these factors.

[0059] The equipment maintenance record database contains a complete service history, with key information from each maintenance operation stored in a structured format. The system automatically detects abnormal maintenance intervals, such as a hydraulic system filter replacement that was delayed by 127 operating hours, exceeding the manufacturer's recommended maintenance interval by 23%. The maintenance anomaly factor is calculated based on the frequency, severity, and impact of such deviations. The equipment diagnostic engine correlates maintenance records with concurrent operating data to assess the potential impact of delayed maintenance on equipment status.

[0060] The lifespan prediction model analyzes wear data from core equipment components. For example, for a hydraulic main pump, the system tracks historical efficiency trends and, based on accumulated operating time and load intensity, estimates the remaining service life. Vibration monitoring data reveals a gradual increase in the pump's axial vibration over the past three months, and spectrum analysis reveals a significant increase in the harmonic components of the impeller pass frequency. These characteristics are fed into the prediction model, which outputs a prediction that the component is likely to experience performance degradation within the next 600-800 operating hours.

[0061] The diagnostic assessment process utilizes a multi-level decision-making mechanism. The system first checks whether the equipment's basic health indicators exceed absolute safety limits, then analyzes for signs of deterioration, and finally assesses the impact of environmental stress and maintenance history. The assessment results are categorized into four levels: Normal, Concern, Warning, and Immediate Action, each with corresponding recommendations. For equipment in a warning state, the system generates a diagnostic report containing detailed analysis data for the maintenance team's reference.

[0062] The elimination diagnostic analysis comprehensively considers the equipment's technical status and economic factors. It systematically assesses the potential repair costs, downtime losses, and engineering risks of continuing to use the equipment, comparing them with the equipment's residual value and replacement costs. The diagnostic process references average service life data for similar equipment and makes adjustments based on the specific operating conditions of the project. When the system detects that a piece of equipment has entered a period of accelerated wear and that maintenance costs are projected to exceed 40% of the purchase price of new equipment, it generates a recommendation for equipment elimination.

[0063] Decision support features provide multi-dimensional analytical views. The device status overview displays historical trends of key parameters in a timeline format, with color-coded anomalies and maintenance records. A comparative analysis tool overlays the current device status with the typical aging curve for similar equipment, visually demonstrating the degree of deviation. A predictive view simulates the remaining life under different maintenance strategies, helping managers assess the long-term impact of various options.

[0064] The maintenance knowledge base integrates equipment manufacturer specifications, historical maintenance cases, and expert experience. When the system detects a specific failure mode, it automatically retrieves relevant solutions and precautions. Knowledge entries contain structured information such as a description of the failure phenomenon, possible cause analysis, detection methods, and repair steps. Maintenance personnel can rate and supplement this knowledge to continuously optimize the knowledge base.

[0065] Early warning information is released using a tiered response mechanism. General alerts are sent to the equipment administrator via system messages; critical alerts are sent simultaneously to the maintenance supervisor and project leader; and emergency alerts trigger on-site audible and visual alarms and notify the relevant emergency response team. Alerts include the equipment number, anomaly description, severity, and recommended actions, enabling rapid problem location and initiation of response processes.

[0066] Data visualization tools provide interactive analysis capabilities. The device health dashboard supports dynamic filtering by time range, parameter type, and anomaly level. Trend charts can overlay multiple related parameters to help analyze fault correlations. Geographic views display the status distribution of distributed devices, facilitating resource allocation. All charts support drill-down operations, allowing quick navigation from summary data to detailed records.

[0067] The system's integrated architecture enables seamless integration with the equipment management platform. Basic equipment information, maintenance work orders, and spare parts inventory data are synchronized in real time. The work order management module automatically converts diagnostic results into maintenance tasks, assigning appropriate priorities and resources. The spare parts demand forecasting function analyzes wear patterns and generates pre-emptive procurement recommendations, reducing waiting times.

[0068] The configuration management interface supports flexible rule adjustments. Users can customize evaluation parameter weights, warning thresholds, and elimination criteria for device types. Rule changes are implemented in phases after impact analysis to avoid sudden changes to existing evaluation results. Version control records all configuration changes and supports rapid rollback to previous versions.

[0069] Mobile apps extend system functionality to the jobsite. Technicians use tablets to view real-time equipment status, review maintenance history, and record inspection data. Augmented reality overlays key parameters onto the live view of the equipment to assist in fault location. Offline mode ensures basic operations are still possible in areas with poor network coverage, with data automatically synchronized upon reconnection.

[0070] Report generation tools output analytical documents that comply with industry standards. Regular health reports summarize device status assessment results, trend analysis, and maintenance recommendations. Special diagnostic reports detail the fault analysis process and resolution plan. Customizable reports allow users to select focus indicators and presentation formats to meet management needs at different levels.

[0071] Example 5: The environmental adaptation module builds a multi-source environmental data acquisition network to achieve all-round environmental perception of the power engineering construction site. In the transmission line construction project, the system deployed distributed meteorological monitoring stations, geological sensor arrays and construction equipment working condition acquisition terminals to form a three-dimensional monitoring system. The meteorological monitoring node collects wind speed, precipitation, temperature and lightning activity data every five minutes and transmits it to the central gateway via the LoRa wireless network. The geological monitoring unit uses an observation network composed of microseismic sensors and inclinometers. The monitoring frequency is set to 20Hz, which can capture millimeter-level displacement changes on the surface of the construction area. Equipment working condition monitoring reads twelve operating parameters such as oil pressure, speed, hydraulic temperature, etc. of large machinery such as excavators and cranes in real time through the CAN bus interface.

[0072] Environmental disturbance analysis utilizes a spatiotemporal correlation modeling approach. A convolutional neural network model receives 72 hours of time-series data on environmental parameters as input and, after multiple layers of feature extraction, outputs environmental disturbance coefficients. The input data is first normalized to eliminate dimensional differences between sensors. A sliding window is then used to generate a 256×256 pixel two-dimensional feature map, where the time dimension represents the monitoring period and the spatial dimension reflects the sensor locations. The network architecture consists of alternating convolutional and pooling layers, progressively extracting multi-scale environmental features, ranging from local anomalies to global trends. The model training phase utilizes a dataset of 200 historical engineering cases, each annotated with the type and severity of an actual environmental accident.

[0073] The dynamic adjustment mechanism for risk assessment thresholds utilizes a fuzzy inference strategy. Parameter deviation data and environmental disturbance coefficients are fed into a fuzzy interface and converted into fuzzy variables with varying degrees of membership. The rule base contains 36 empirically derived adjustment rules, such as "When wind speed parameter deviation is large and the environmental disturbance coefficient indicates continued strong winds, moderately lower the risk assessment threshold." The defuzzification process converts the fuzzy output into a precise threshold adjustment, limited to ±15% of the base threshold. The system records the decision-making basis and actual results of each threshold adjustment and continuously optimizes the rule weights through online learning.

[0074] A hierarchical early warning system has been established by analyzing the correlation between meteorological data and construction safety. Wind speed monitoring data is filtered through a moving average and then compared in real time with lifting operation safety standards. When the instantaneous wind speed exceeds 80% of the equipment's rated operating limit, the system automatically sends an alert to the relevant operators. Rainfall monitoring data, combined with soil moisture sensor readings, assesses the stability risk of foundation pit slopes. The lightning warning system calculates the distance between the construction area and the center of the thunderstorm and predicts dangerous periods based on approach speed and intensity. All meteorological warning information comes with specific work restriction recommendations, such as suspending high-altitude work or requiring equipment to be reinforced for windproofing.

[0075] Analysis of geological monitoring data focuses on identifying potential risks of surface deformation. An array of microseismic sensors uses time-difference-of-arrival (TDOA) technology to locate vibration sources, distinguishing vibrations caused by construction machinery from natural geological activity. A network of inclinometers monitors tilt changes in the area surrounding tower foundations, initiating specialized assessments when a continuous trend of displacement is detected. The system compares geological monitoring data with engineering design parameters, focusing particularly on monitoring soft soil areas and fill sections. Geological risk warnings include assessments of the potential impact range and projected rate of development, guiding the implementation of targeted prevention and control measures on-site.

[0076] Environmental adaptability analysis of equipment operating conditions focuses on performance changes under extreme conditions. The hydraulic system monitoring unit automatically increases the oil temperature sampling frequency in low-temperature environments to track fluidity changes during the startup phase. In hot weather, the system focuses on monitoring engine cooling efficiency and temperature rise in electrical equipment. In humid environments, insulation resistance testing is performed more frequently to prevent short-circuit risks. Correlation analysis between equipment status data and environmental parameters can identify performance inflection points under specific climate conditions, providing a basis for adjusting operating parameters. If the system detects that the equipment is approaching its operating limits under current environmental conditions, it recommends reducing the load or shortening the continuous operation time.

[0077] The environmental risk visualization interface utilizes multi-layer map overlay technology. The base layer displays a satellite image of the construction site, the middle layer overlays a heat map of real-time environmental monitoring data, and the upper layer labels risk warning areas and equipment locations. Users can use the time slider to view the historical evolution of environmental conditions or select specific parameter types for specialized analysis. The 3D visualization function supports rendering of terrain elevation data, visually displaying the spatial distribution of geological monitoring points. Detailed information for each environmental element can be viewed with a single click, including sensor readings, change curves, and associated equipment status.

[0078] The early warning information distribution system implements hierarchical and categorized push notifications. General environmental alerts are sent to relevant team leaders; critical alerts that could impact construction safety are simultaneously notified to the project safety engineer; and emergency alerts requiring evacuation trigger site-wide broadcasts and mobile notifications. Alerts include a description of the environmental anomaly, an estimated impact area, recommended response measures, and a predicted duration. Information is pushed through on-site LED displays, radio intercoms, and mobile app notifications, ensuring timely access to all personnel. Alert response status is tracked in real time, and unconfirmed alerts are automatically escalated.

[0079] The environmental data archiving system establishes a complete, traceable record. Raw sensor data undergoes quality inspection and is stored in a time-series database, retaining millisecond timestamps and device identification information. Derived data products, such as environmental disturbance coefficients and risk threshold adjustment records, are stored in a document database to maintain data relevance. Long-term archived data is compressed and indexed, enabling rapid retrieval of similar historical environmental scenarios. The data access interface provides combined query capabilities by time range, geographic region, and anomaly type, facilitating post-analysis and case studies.

[0080] Systematic maintenance mechanisms ensure the ongoing reliability of the environmental monitoring network. Sensor nodes are equipped with self-diagnostic capabilities, regularly reporting battery charge, signal strength, and measurement accuracy. Base station equipment utilizes redundant configurations, with automatic switchover between active and standby nodes ensuring data continuity. Network communication quality is monitored in real time, with transmission frequency and power automatically adjusted in weak signal areas. Regular calibration processes ensure measurement data accuracy, with calibration records including complete information such as operator, reference source data, and correction parameters. Troubleshooting work orders are automatically distributed to the maintenance team, tracking the entire problem resolution process.

[0081] The mobile app provides real-time access to on-site environmental information. Construction workers can view the environmental risk level and specific hazard alerts for their current location via their mobile phones. The navigation function guides workers to avoid high-risk areas or quickly reach the nearest emergency shelter. Offline mode caches critical environmental data, providing basic safety guidance even during network outages. On-site photo upload allows for supplemental reporting of environmental anomalies, with images automatically tagged with location and time information. The voice announcement function provides instant voice alerts when the environment changes suddenly, compensating for the limitations of visual cues.

[0082] The integration of environmental data with the engineering management system enables collaborative decision-making. The scheduling module automatically adjusts work sequencing based on environmental forecast data, advancing or delaying sensitive processes during high-risk weather periods. The resource scheduling system optimizes equipment allocation based on environmental conditions, preventing performance-limited machinery from operating in harsh environments. The safety management system automatically converts environmental warnings into inspection tasks, requiring targeted inspections of affected areas. The quality management system records the impact of environmental parameters on construction processes, providing context for acceptance evaluations.

[0083] Configuration management tools support flexible adjustment of environmental rules. Users can customize environmental parameter sensitivity weights for different project types and set project warning thresholds. Equipment environmental adaptability profiles maintain operating limit data for various types of machinery and are updated in sync with manufacturer technology updates. An emergency plan library manages the handling procedures for various environmental emergencies and regularly organizes simulation drills. All configuration changes record the reason for the change and the effective date, supporting configuration rollbacks and version comparisons.

[0084] This environmental adaptive system is distinguished by transforming discrete environmental factor monitoring into a comprehensive risk assessment. Using machine learning methods, it explores the inherent correlations between environmental parameters and project safety. The system not only monitors current environmental conditions but also analyzes changing trends and cumulative effects, enabling early identification of potential risks. Multi-dimensional correlation analysis of environmental data, equipment status, and construction progress supports more informed environmental adaptability decisions. A dynamic adjustment mechanism ensures consistent risk assessments while adapting to unique environmental conditions, balancing safety requirements with project efficiency.

[0085] Data analysis methods emphasize the spatiotemporal characteristics of environmental impacts. The rate and periodicity of parameter changes are analyzed in the temporal dimension to distinguish between transient disturbances and sustained trends. Spatial correlation models are established across monitoring points to distinguish between local anomalies and regional environmental changes. Combined temporal and spatial analysis can track the propagation path and impact range of risk sources, for example, determining the direction of strong winds and the duration of their impact on subsequent construction areas.

[0086] The system architecture was designed with the specificities of field construction environments in mind. Sensor nodes feature an IP65 rating and operate in a wide operating temperature range, adapting to ambient temperatures ranging from -30°C to 70°C. Data transmission utilizes a multi-mode hybrid network, automatically selecting LoRa, 4G, or satellite communications based on site conditions. Edge computing nodes deploy pre-processing algorithms to reduce network data transmission. The power system integrates solar panels and long-lasting lithium batteries to ensure normal operation even in continuous rainy weather. Lightning protection features include multi-level surge protection and a grounding system to reduce the risk of damage from lightning strikes.

[0087] 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.

[0088] 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. A whole-process management and control system for electric power engineering construction projects, characterized in that: It includes engineering parameter monitoring module, risk analysis module, progress control module, quality assessment module, early warning output module and background supervision terminal; The engineering parameter monitoring module collects multi-dimensional parameters in the power engineering construction process in real time through the sensor array, and sends the real-time parameter values ​​to the risk analysis module and the background monitoring terminal, which visualizes the real-time parameter values; The risk analysis module obtains parameter deviations based on real-time parameter values ​​and preset standard parameter values, calculates progress control instructions based on the parameter deviations using a random forest algorithm, and sends the progress control instructions to the progress control module; The progress control module performs construction progress adjustment operations according to the progress control instructions to optimize the construction progress; the quality assessment module monitors the progress adjustment operation process, judges the execution performance of the progress adjustment operation, generates a quality pass signal or a quality warning signal based on the performance, and sends the quality pass signal or quality warning signal to the warning output module; When the warning output module receives the quality warning signal, it generates warning information and sends it to the background monitoring end, and the background monitoring end issues an alarm according to the warning information.

2. The whole process control system of electric power engineering construction project according to claim 1 is characterized in that: The specific evaluation process of the quality assessment module includes: marking the corresponding progress adjustment operation as an abnormal operation or a standard operation through a dynamic threshold evaluation mechanism, setting an evaluation period, and calculating the ratio of the number of abnormal operations to the total number of progress adjustment operations within the evaluation period to obtain the abnormal operation rate; when the abnormal operation rate exceeds the preset abnormal threshold, a quality warning signal is generated; when the abnormal operation rate does not exceed the preset abnormal threshold, the difference between the execution efficiency value of the corresponding progress adjustment operation and the middle value of the preset efficiency benchmark value is calculated to obtain the absolute value to obtain the efficiency evaluation value, and the ratio of the stability value of the corresponding operation process to the preset stability benchmark is marked as the stability evaluation value; the efficiency evaluation values ​​of all progress adjustment operations within the evaluation period are weighted averaged to obtain the comprehensive efficiency value, and the stability evaluation values ​​of all progress adjustment operations within the evaluation period are weighted averaged to obtain the comprehensive stability value; the quality output value is obtained by fusing the abnormal operation rate, the comprehensive efficiency value and the comprehensive stability value, and a quality warning signal is generated when the quality output value exceeds the preset quality threshold, and a quality qualified signal is generated when the quality output value does not exceed the preset quality threshold.

3. The whole process control system of electric power engineering construction project according to claim 2 is characterized in that: The specific evaluation process of the dynamic threshold evaluation mechanism includes: recording the moment when the progress control module receives the progress control instruction as the starting moment, and recording the moment when the progress control module completes the corresponding progress adjustment operation as the ending moment, and marking the time interval between the starting moment and the ending moment as the monitoring period; marking the ratio of the progress control instruction to the monitoring period as the execution efficiency value, and obtaining the stability value through operation smoothness analysis; when the execution efficiency value is not in the preset efficiency value range or the stability value exceeds the preset stability limit, marking the corresponding progress adjustment operation as an abnormal operation; when the execution efficiency value is in the preset efficiency value range and the stability value does not exceed the preset stability limit, marking the corresponding progress adjustment operation as a standard operation.

4. The whole process control system of electric power engineering construction project according to claim 3 is characterized in that: The specific analysis process of the operation stability analysis includes: constructing a two-dimensional coordinate system with the time axis as the horizontal axis and the real-time parameter value as the vertical axis, extracting the change trajectory of the engineering construction parameters during the corresponding progress adjustment operation, placing the change trajectory in the coordinate system, and the starting point of the change trajectory is located on the vertical axis; setting multiple sampling points on the change trajectory, marking the longitudinal distance between adjacent sampling points as the change amplitude value, performing cluster analysis on all change amplitude values ​​to obtain a fluctuation assessment value, and marking the frequency of occurrence of the change amplitude value that is not within the preset change amplitude range as an abnormal frequency value; and obtaining the stability value of the corresponding progress adjustment operation by combining the fluctuation assessment value and the abnormal frequency value.

5. The whole process control system of electric power engineering construction project according to claim 1 is characterized in that: The early warning output module is communicatively connected to the safety auxiliary analysis module, and the early warning output module sends a quality qualified signal to the safety auxiliary analysis module; When the safety auxiliary analysis module receives the quality qualified signal, it performs a safety auxiliary evaluation on the progress control module, generates a safety alarm signal or an auxiliary qualified signal through the evaluation, and sends the safety alarm signal or the auxiliary qualified signal to the background supervision end; when the background supervision end receives the safety alarm signal, it issues a safety alarm.

6. The whole process control system of electric power engineering construction project according to claim 5 is characterized in that: The specific evaluation process of the safety auxiliary evaluation includes: collecting the vibration intensity and noise level values ​​of the progress control module during its operation; when the vibration intensity or noise level value exceeds the corresponding preset safety standard, judging that the progress control module is in an abnormal state; obtaining the duration of the progress control module in the abnormal state during the evaluation period and calculating the ratio thereof with the total operation time during the evaluation period to obtain the abnormal duration rate, marking the number of occurrences of the single duration of the progress control module in the abnormal state during the evaluation period exceeding the corresponding preset single duration threshold as an over-limit abnormal value, and marking the longest single duration of the progress control module in the abnormal state during the evaluation period as an abnormal duration value; obtaining a safety evaluation value by weighted calculation of the abnormal duration rate, the over-limit abnormal value and the abnormal duration value; generating a safety alarm signal when the safety evaluation value exceeds the preset safety threshold; generating an auxiliary qualified signal when the safety evaluation value does not exceed the preset safety threshold.

7. The whole process control system of electric power engineering construction project according to claim 1 is characterized in that: The background supervision end is communicated with the equipment diagnosis feedback module. When a quality warning signal or a safety alarm signal is generated, the equipment diagnosis feedback module performs an elimination diagnosis analysis on the progress control module, determines through analysis whether an equipment elimination signal is generated, and sends the equipment elimination signal to the background supervision end; when the background supervision end receives the equipment elimination signal, it issues an equipment elimination alarm.

8. The whole process control system of electric power engineering construction project according to claim 7 is characterized in that: The specific analysis process of the equipment diagnosis feedback module includes: obtaining the production date of the progress control module, calculating the time difference between the current date and the production date to obtain the equipment life value, and marking the total length of time the progress control module has been in operation in the historical stage as the cumulative operation value; obtaining the environmental risk value of the progress control module through analysis, and marking the number of times the maintenance interval of the progress control module in the historical stage exceeds the preset maintenance interval standard as the maintenance anomaly coefficient; obtaining the diagnostic evaluation value by linearly combining the equipment life value, the cumulative operation value, the environmental risk value and the maintenance anomaly coefficient; and generating an equipment elimination signal when the diagnostic evaluation value exceeds the preset diagnostic threshold.

9. The whole process control system of electric power engineering construction project according to claim 8 is characterized in that: The analysis and acquisition method of the environmental risk value includes: collecting the temperature parameters and humidity parameters of the environment in which the progress control module is located, marking the deviation value of the temperature parameter compared to the set standard temperature value as the temperature assessment value, and marking the deviation value of the humidity parameter compared to the set standard humidity value as the humidity assessment value; collecting the pollutant concentration of the environment in which the progress control module is located and marking it as the pollution assessment value; obtaining the environmental risk value by normalizing the temperature assessment value, the humidity assessment value and the pollution assessment value; when the environmental risk value exceeds the preset environmental risk threshold, judging that the progress control module is in a high-risk state; obtaining the total time that the progress control module is in a high-risk state in the historical stage and marking it as the environmental risk value.

10. The whole process management and control system of electric power engineering construction project according to claim 2, characterized in that: The risk analysis module is communicatively connected to the environmental adaptation module; the environmental adaptation module collects meteorological data, geological change data, and equipment operating condition data from the construction site in real time; the environmental adaptation module analyzes the correlation between environmental parameters and historical accident data through a convolutional neural network and outputs an environmental disturbance coefficient; the risk analysis module couples the environmental disturbance coefficient with the parameter deviation and dynamically adjusts the risk judgment threshold of the random forest algorithm.

Images