Intelligent operation and maintenance management method for digital twin water quality automatic monitoring station

Abstract

The invention relates to the technical field of operation and maintenance management, in particular to an intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station. The method comprises the following steps: acquiring an equipment sensor signal set through a sensor cluster pre-deployed in an automatic monitoring station, and executing digital signal conversion; executing sensor signal collaboration according to a digital signal conversion result to obtain equipment sensing data; determining an abnormal sampling point according to the numerical value change amplitude of the equipment sensing data; performing drift correction on the abnormal sampling point, and replacing a drift correction result to a corresponding sampling value in the equipment sensing data to obtain sensing correction data; combining the sensing correction data with a preset equipment mechanism model to construct an equipment operation twin model; and executing equipment operation simulation by using the equipment operation twin model, and determining an operation and maintenance management strategy according to an operation simulation result. According to the invention, the overall operation and maintenance efficiency, reliability and fine management capability of the automatic water quality monitoring station can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of operation and maintenance, and particularly relates to an intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station. BACKGROUND

[0002] At present, the application of digital twin technology in water quality monitoring stations mainly involves deploying multiple types of sensors to collect real-time equipment operation parameters such as temperature, pressure, and vibration in multiple dimensions, and transmitting the parameters to a digital twin model constructed based on multi-physical field coupling modeling and finite element analysis technology to achieve high-precision simulation of equipment operation status.

[0003] However, the existing digital twin water quality monitoring system still faces multiple technical bottlenecks in actual application. First, in terms of data quality, affected by factors such as electromagnetic interference, water corrosion, and environmental fluctuations, the monitoring data collected by sensors is prone to drift, missing, and abnormal fluctuations. The measurement standards between equipment from different manufacturers are not unified, resulting in insufficient data comparability and consistency. At the same time, the traditional transmission method has a delay, making it difficult to guarantee the real-time and reliability of the monitoring data. Second, in terms of model accuracy, there is a significant deviation between the water quality simulation model and the actual environmental characteristics. The existing modeling method is difficult to accurately depict the multi-factor coupling effect in complex water areas. The model parameter calibration process relies on manual adjustment and lacks a self-adaptive optimization mechanism. The correlation accuracy between multi-scale models is low, and the fitting effect from point data to basin scale is limited. Third, in terms of system integration, due to differences in construction period and communication protocols between different monitoring stations, data exchange and resource sharing between heterogeneous systems are limited, causing difficulties in integrating monitoring data and insufficient system compatibility and expandability. Finally, in terms of operation and maintenance, the device status evaluation and fault warning still rely on static thresholds, with a high false alarm rate. Maintenance decisions rely on human experience and lack data-driven intelligent analysis and prediction support. The efficiency of spare parts management and deployment is low, making it difficult to achieve fine and preventive maintenance of monitoring equipment. SUMMARY

[0004] Therefore, it is necessary to provide an intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station to solve at least one of the above technical problems.

[0005] To achieve the above-mentioned purpose, an intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station includes the following steps: Step S1: Collecting equipment sensor signal sets through a pre-deployed sensor cluster at the automatic monitoring station, and transmitting the equipment sensor signal sets to a pre-deployed edge node to perform digital signal conversion; Step S2: Performing sensor signal coordination based on the digital signal conversion result to obtain equipment sensor data; Step S3: determining an abnormal sampling point according to the numerical variation amplitude of the device sensor data; performing drift correction on the abnormal sampling point, and replacing the drift correction result to the corresponding sampling value in the device sensor data to obtain sensor correction data; Step S4: constructing a device operation twin model in combination with the sensor correction data and a preset device mechanism model; performing device operation simulation using the device operation twin model, and determining an operation and maintenance strategy according to the operation simulation result.

[0006] The present application realizes high-frequency and low-latency collection of water quality automatic monitoring station device operation parameters through multi-sensor cluster collection + edge node digital signal conversion, can effectively reduce data drift and abnormal fluctuations caused by electromagnetic interference, water corrosion or environmental fluctuations, and at the same time, can unify and standardize data from devices of different manufacturers, improve data consistency and comparability, and thus guarantee the real-time and reliability of monitoring data. Secondly, through collaborative sensor signal processing and drift correction mechanism, abnormal sampling points can be identified and offset correction values can be calculated based on adjacent sensor signals and historical data, data noise cross-cancellation and fluctuation component integration can be realized, and device sensor data can be closer to the actual operation state, providing high-precision basic data for subsequent simulation and operation and maintenance analysis. Thirdly, in combination with the device mechanism model and the operation twin model, multi-dimensional operation simulation of the device is performed on the digital twin platform, and the simulation result is compared and verified with the real-time collected data, the model parameters can be self-adaptively calibrated through iterative optimization of the simulation result, the operation state of the water quality monitoring device can be accurately simulated, and the model precision and multi-scale correlation ability can be significantly improved, and the simulation from point data to basin scale is more reliable. In addition, through the dynamic matching mechanism of the operation and maintenance knowledge base and the spare parts library, high-similarity operation and maintenance scenarios can be identified in combination with the operation simulation result, and historical operation and maintenance strategies can be dynamically adjusted based on the operation state deviation, and intelligent generation of operation and maintenance strategies can be realized. This method can automatically judge the operation and maintenance deviation, and can classify and distribute remote control tasks or high-priority maintenance tasks according to the deviation amplitude, can significantly improve the fault response efficiency and the scientific nature of maintenance decision, can reduce the false alarm rate, and can improve the intelligent level of preventive maintenance and spare parts management. The present application forms an end-to-end closed-loop operation and maintenance management system, covers the whole process from data collection, data correction, digital twin simulation, operation and maintenance strategy generation to task distribution and verification optimization, realizes real-time monitoring, intelligent analysis and active maintenance of the device operation state, and improves the overall operation and maintenance efficiency, reliability and fine management ability of the water quality automatic monitoring station. BRIEF DESCRIPTION OF DRAWINGS

[0007] Other features, objects and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments made with reference to the accompanying drawings: Fig. 1 The present application is a schematic diagram of the step flow of the digital twin water quality automatic monitoring station intelligent operation and maintenance management method. Fig. 2 Figure 1 is a profile schematic diagram of a water quality automatic monitoring station in an embodiment of the present application; The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0008] The technical method of the present application will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0009] In addition, the accompanying drawings are only schematic illustrations of the present application, and are not necessarily drawn to scale. The same reference numerals in the drawings represent the same or similar parts, and thus repeated descriptions thereof will be omitted. Some block diagrams shown in the drawings are functional entities, which do not necessarily have to correspond to physically or logically independent entities. The functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0010] It should be understood that although the terms "first", "second" and the like can be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, without departing from the scope of the exemplary embodiments, a first element can be referred to as a second element, and similarly a second element can be referred to as a first element. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0011] To achieve the above-mentioned object, please refer to Figs. 1-2 The present application provides a digital twin water quality automatic monitoring station intelligent operation and maintenance management method, which comprises the following steps: Step S1: Collecting equipment sensor signal set through the sensor cluster pre-deployed in the automatic monitoring station, and transmitting the equipment sensor signal set to the pre-deployed edge node to perform digital signal conversion; In this embodiment, if the automatic monitoring station is in a continuous monitoring period, the sensor cluster in the monitoring area starts synchronous collection. The sensor cluster includes temperature, vibration, current and sound pressure sensing units, which are respectively arranged in the equipment bearing part, the outer wall of the shell, the power supply port and the shell cavity, forming a monitoring system covering the three-dimensional state of heat, force and electricity of the equipment. Each sensor collects analog signals at a sampling frequency of 2Hz, and sends them to the edge node after preliminary denoising by the collection controller. The edge node is configured with a 16-bit A / D conversion chip to perform digital signal conversion, and the sampling time stamp of each signal is unified to the same time base. In order to eliminate high-frequency interference, a band-pass filter is added during digital conversion, and the filter passband range is 1-200Hz. The final digital signal data packet is aggregated into a device sensor signal set according to the sensor number, providing the original input for subsequent collaborative analysis.

[0012] Step S2: performing sensor signal collaboration according to the digital signal conversion result to obtain device sensing data; In one embodiment, the sensor signal after digital signal conversion in the edge node is subjected to sensor signal collaboration processing. The converted signal set is subjected to effective sensor digital signal screening, taking the signal-to-noise ratio (SNR) and stability coefficient of the signal in each sampling period as the screening condition. If the SNR of the sensor output signal is less than 20dB or the stability coefficient exceeds 0.15, it is determined that the signal has abnormal drift or random pulse interference, and is rejected. For the screened effective sensor signal, interpolation reconstruction is performed according to its sampling time stamp to strictly align each signal sequence on the same time axis. After signal screening, noise cross-cancellation processing is performed. This process takes the amplitude deviation of adjacent sensors in the same sampling period as the coupling parameter, calculates the cross-correlation function to identify the synchronous noise component. If a certain noise component appears in adjacent sensors and the phase difference is less than 5°, it is determined as a noise signal, which is cancelled by inverse phase superposition. The cancelled signal is subjected to band-limited smoothing processing (cut-off frequency set to 180Hz) to form a multi-dimensional signal matrix with consistent timing and noise suppression. Then, according to the physical quantity type of different sensors (such as temperature, current, vibration, sound pressure), the corresponding data sub-matrix is constructed, and is spliced according to the sampling time sequence to form the device sensing data.

[0013] Step S3: determining the abnormal sampling point according to the numerical change amplitude of the device sensing data; performing drift correction on the abnormal sampling point, and replacing the drift correction result to the corresponding sampling value in the device sensing data to obtain the sensing correction data; In this embodiment, the abnormal sampling points are detected according to the value change range of the device sensor data. Specifically, taking every five sampling points as an analysis unit, if the value change range of the current sampling point exceeds 1.5 times of the average change range of the adjacent interval, the current sampling point is marked as an abnormal sampling point. For the abnormal sampling point, the amplitude difference value and the direction angle change value of the adjacent sampling points before and after the abnormal sampling point are extracted as the offset identification parameters. The average change range of the same type of sensor in the same continuous sampling period is searched in parallel and the direction change distribution density is taken as the drift correction reference value. Then, according to the preset weight ratio (amplitude correction weight = 0.6, direction correction weight = 0.4), the comprehensive offset correction value is calculated. If is greater than the drift threshold value, the value replacement correction is performed, and the corrected amplitude and direction angle are backfilled to the corresponding position of the device sensor data. The generated sensor correction data after correction retains the original time sequence structure, ensures the data continuity and consistency, and thus provides reliable input for subsequent twin simulation.

[0014] Step S4: combining the sensor correction data with the preset device mechanism model to construct a device operation twin model; using the device operation twin model to perform device operation simulation, and determining the operation and maintenance strategy according to the operation simulation result.

[0015] In this embodiment, a device operation twin model is constructed by combining the sensor correction data with the preset device mechanism model. The mechanism model is established based on the design parameters of the device manufacturer, and includes a speed-load-temperature rise coupling relationship matrix and a power-current mapping function, which are used to describe the physical constraint conditions of device operation. The twin model takes the sensor correction data as the input layer, takes the mechanism constraint relationship as the intermediate layer, and forms a dynamic mapping of the device operation state through parameter fitting. In the simulation stage, an iterative operation is performed every 30 seconds, and the output operation simulation result includes the real-time power change curve, the temperature rise curve and the vibration energy spectrum distribution. The system identifies the operation abnormal trend of the device according to the deviation of the simulation output and the historical normal operation sample, and then determines the corresponding operation and maintenance strategy. The strategy is based on the correction data to realize operation load distribution optimization, temperature control trigger threshold adjustment and maintenance cycle revision, so as to form a complete operation and maintenance decision closed loop.

[0016] Optionally, after determining the operation and maintenance strategy in step S4, the method further includes: sending the operation and maintenance strategy to the automatic monitoring station management platform, and distributing maintenance task sheets to the execution units according to the operation and maintenance deviation amplitude of the operation and maintenance strategy and the existing operation and maintenance strategy in the automatic monitoring station management platform; In this embodiment, the strategy is transmitted to the automatic monitoring station management platform through edge node encryption. After receiving, the platform first matches the existing operation and maintenance management strategy library according to the strategy number, calculates the operation and maintenance management deviation amplitude of the three key parameters of device running intensity, temperature control threshold and maintenance cycle setting. The deviation amplitude is calculated in the form of normalized difference, and the normalized threshold range is set to [0, 1]. When the deviation amplitude is in the range of 0.15-0.35, it is defined as moderate deviation. When the deviation amplitude is in the moderate range, the classification and distribution process is automatically triggered. According to the deviation amplitude level, the task list is classified, if the deviation amplitude is in the remote adjustable range, a remote control task list is generated with control instructions and encrypted control links; if the deviation amplitude exceeds the upper limit of the remote adjustable range, a high-priority maintenance task list is automatically generated, which contains maintenance scheme number, device repair instructions and required spare parts number information. Each task list is pushed to the corresponding execution unit through the task distribution interface of the platform, and the execution unit starts the subsequent maintenance operation after receiving and confirming. Through this deviation grading distribution method, remote and on-site maintenance resources can be reasonably scheduled to avoid maintenance response delay.

[0017] Synchronously collect real-time device sensor signal set by using sensor cluster, and perform digital signal conversion and sensor signal coordination to obtain real-time device sensing data; Compare real-time device sensing data with operation simulation result to verify the operation simulation result; In this embodiment, the real-time collection program of the sensor cluster is triggered synchronously. Each sensor outputs a digital signal with a sampling period of 200 ms, and performs digital signal conversion and signal coordination processing through an edge node. Digital signal conversion includes A / D conversion and timestamp synchronization, which ensures that all signals have a unified sampling reference; signal coordination processing performs effective signal filtering and noise cross-cancellation operations according to the spatial distribution relationship of each sensor in the same sampling period to generate a real-time device sensing data matrix. The dimension of the matrix is defined as T x N, where T is the number of time sampling points, N is the number of sensors, and each element is the amplitude data of the sensor after space-time synchronization processing.

[0018] In a further embodiment, the simulation result of the previous round is called for comparison and verification. First, linear interpolation alignment is performed on the real-time device sensing data and the simulation result on the same time axis to eliminate the sampling period difference. Then, the average absolute error of both on the key indicators (including flow response time, pump temperature rise rate, power supply current fluctuation coefficient) is calculated as the verification numerical error. If the verification numerical error of any indicator exceeds 0.08, it is determined that the simulation result has deviation. At this time, the error distribution diagram and the corresponding deviation source positioning report are output, which includes the deviation direction of each parameter and the corresponding device number. The verification result is used to feedback the correction direction of the simulation model, ensuring that the calculation result of the twin model is consistent with the on-site running state. Through this comparison and verification, dynamic self-calibration of the running twin model under real-time monitoring conditions can be realized.

[0019] It is worth noting that the verification result can also be uploaded to the operation and maintenance knowledge base to update the knowledge base.

[0020] According to the verification numerical error in the verification result, the operation and maintenance strategy is iteratively optimized until the verification numerical error is less than the preset error threshold.

[0021] In this embodiment, if the verification result shows that the simulation result has deviation, the iterative optimization process of the operation and maintenance strategy is started. The correction weight is determined according to the deviation proportion of each indicator in the error distribution diagram. If the temperature rise deviation accounts for more than 40% of the total deviation, the weight of the temperature control related parameters is increased to 0.6; if the load deviation accounts for less than 20%, the weight of the running intensity classification parameters is reduced to 0.2. The device running intensity classification, cooling trigger threshold and maintenance cycle setting value are recalculated according to the weight adjustment result to generate a new operation and maintenance strategy. The updated strategy is verified again on the platform side to calculate the verification numerical error. If the error is less than the preset error threshold 0.05, it is determined that the optimization is completed; if the threshold requirement is not met, the optimization process is repeated until the convergence condition is met. The optimized strategy is updated to the operation and maintenance knowledge base, and the global strategy is synchronized through version number control to realize dynamic convergence of the digital twin system with the real device state in multiple iterations.

[0022] Fig. 2 A profile schematic diagram of the water quality automatic monitoring station in the embodiment of the present application is shown in Fig. 2 101 is a tower top meteorological sensor of the water quality automatic monitoring station, which is arranged at the top of the meteorological monitoring tower of the water quality automatic monitoring station and can monitor meteorological parameters such as air temperature, humidity and air pressure; ​102 is a wireless transmission device of the water quality automatic monitoring station, equipped with an industrial-grade wireless communication module (supporting 4G / 5G, LoRa, etc. Multi-protocol), through a high-gain antenna to realize real-time convergence and transmission of sensor cluster signals, for transmitting the sensor signals collected by the sensor cluster to the edge node; 103 is a sensor cluster deployed on the pipeline and device at different depths of the water quality automatic monitoring station in the water. The area is an array of turbidity sensors (range 0~4000NTU, accuracy ±2%FS), dissolved oxygen sensors (range 0~20mg / L, accuracy ±0.3mg / L), pH sensors (range 0~14, accuracy ±0.1pH), conductivity sensors (range 0~50000μS / cm, accuracy ±1%FS) and other multi-type water quality sensors on the sampling pipeline, multi-parameter sensing probe device at different depths of the monitoring well.

[0023] Optionally, the classified distribution maintenance task sheet includes: If the operation and maintenance deviation amplitude is in the preset remote adjustment interval, the operation and maintenance strategy corresponding remote execution unit is determined through the automatic monitoring station management platform, and the control instructions in the operation and maintenance strategy and the remote control link corresponding to the remote execution unit are integrated into a remote control task sheet, which is distributed to the corresponding remote execution unit; In this embodiment, if the automatic monitoring station management platform detects that the operation and maintenance deviation amplitude is in the preset remote adjustment interval (for example, 5% to 15%), the corresponding execution unit identifier is queried in the task execution index table according to the operation and maintenance strategy number, and the remote control link address is extracted according to the network topology configuration file of the execution unit. Subsequently, the platform extracts the parameter adjustment instructions that can be realized by remote control from the operation and maintenance strategy, including pump start-stop timing adjustment, valve opening degree fine tuning, and data sampling frequency correction control instructions. The control instructions are encapsulated in the form of structured instruction blocks, each instruction block including instruction type, target device number, execution delay, and security check code. The control instructions and the remote control link are integrated to generate a remote control task sheet, which has task number, task priority, and feedback channel information. After generating the task sheet, the platform transmits it to the control terminal of the corresponding remote execution unit through the edge gateway encryption, and records the delivery time and execution status in the task management table. After receiving the task sheet, the remote execution unit confirms the legality of the instruction by checking the security code, and executes the operation within the specified time window, while feeding back the execution status to the platform task monitoring interface in real time, realizing closed-loop control of remote control tasks.

[0024] If the operation and maintenance management deviation amplitude exceeds the remote adjustment interval, the maintenance operation guide, equipment maintenance scheme and spare parts configuration requirement in the operation and maintenance management strategy are integrated into a high-priority maintenance task list, and the high-priority maintenance task list is pushed to the on-site maintenance execution unit in the corresponding area of the automatic monitoring station through the automatic monitoring station management platform.

[0025] In another embodiment, if it is monitored that the operation and maintenance management deviation amplitude exceeds the upper limit of the remote adjustment interval, the platform automatically switches to the on-site maintenance task generation mode. At this time, the maintenance priority level of the affected equipment is determined according to the equipment operation intensity difference, temperature rise abnormality degree and maintenance time delay amplitude in the deviation data. The platform extracts the maintenance operation guide corresponding to the level from the operation and maintenance management strategy, including the pump body filter cleaning process, power supply line contact detection step and sensor calibration process. At the same time, the repair template and equipment structure parameter file in the operation and maintenance management knowledge base are called to automatically match the adapted equipment repair scheme, and the required spare parts configuration list is determined according to the current equipment model and running time. After integration of all contents, a high-priority maintenance task list is generated, and the task list structure includes: ① maintenance object number; ② maintenance operation guide; ③ repair step sequence and estimated time consumption; ④ spare parts configuration requirement list; ⑤ task priority identification and responsibility unit code. After the task list is generated, it is transmitted to the on-site maintenance execution unit terminal in the corresponding area through the on-site task pushing interface of the automatic monitoring station management platform, and a task receipt mechanism is attached when pushing, requiring the execution unit to confirm the task receiving status within 10 minutes after receiving. The on-site maintenance personnel implement the repair according to the operation guide in the task list after receiving the task list, and upload the execution record and equipment operation recovery data after completing the task. The platform compares the equipment sensor data before and after maintenance to verify the task effect and update the operation and maintenance management strategy version, realizing the reverse supplement of on-site maintenance data to the operation and maintenance strategy library.

[0026] Optionally, the sensor signal coordination in step S2 includes: According to the transmission timestamp sequence of each sensor signal in the digital signal conversion result to the edge node, the execution time of each sensor digital signal is synchronized and paired; In this embodiment, each sensor digital signal will be accompanied by a transmission timestamp during transmission to the edge node. The sequence of these timestamps will be sorted and aligned to establish a time synchronization table with a sampling period as the unit. In this embodiment, the sampling period can be set to 500 milliseconds, and the timestamp deviation is allowed to be within ±5 milliseconds, ensuring that minor transmission delays will not affect the synchronization result. By aligning the timestamps of each sensor digital signal, a sensor time synchronization matrix can be generated, where the rows represent sensor numbers, the columns represent sampling time points, and the matrix elements are digital signal values at corresponding time points. This matrix provides a basic data structure for subsequent signal coordination analysis.

[0027] Acquire the spatial distribution information of each sensor in the sensor cluster, and detect the signal change trend of adjacent sensors in the same sampling period based on the spatial distribution information of each sensor in the time synchronization pairing results; In this embodiment, the spatial distribution information of the sensor cluster can be obtained through the automatic monitoring station management platform, including the specific deployment coordinates of the sensors in the monitoring station (such as the three-dimensional positions of x, y, and z) and a list of neighboring sensors. Based on this spatial information and the sensor time synchronization matrix, the signal change trend of adjacent sensors within each sampling period is detected. The change trend can be calculated by the numerical difference between adjacent sampling points and represented by the change in direction vector and amplitude. In this embodiment, the threshold for the direction vector angle is set to 30°, and the threshold for the amplitude change rate is ±10%, to determine whether the signals of adjacent sensors show a consistent trend. The detection result will form a signal trend consistency matrix, where the matrix element value is 0 or 1, respectively indicating whether the signal trends of adjacent sensors are consistent within the sampling period.

[0028] Based on the detection results of signal change trends, effective sensor digital signals are selected, and noise cross-cancellation is performed on the effective sensor digital signals to obtain the results of cooperative sensor digital signals.

[0029] In a further embodiment, valid sensor digital signals are screened based on a signal trend consistency matrix. For each sensor, if its signal change trend is consistent with that of more than one-third of its neighboring sensors in the same sampling period and the amplitude change is within a set threshold range, then the sensor's digital signal is determined to be a valid sensor digital signal. In noise cross-cancellation, the fluctuation components of the digital signals of the valid sensors and their three nearest neighboring sensors are extracted, the residual mean sequence is calculated, and its moving average is used as the environmental background noise benchmark. By using the fluctuation components of adjacent sensors as weights to perform inverse superposition on the fluctuations of the valid sensor digital signals, and integrating the results of each sampling period, a collaborative sensor digital signal result matrix is ​​obtained, which is used as the input for device operation simulation.

[0030] Optionally, filtering valid sensor digital signals includes: Calculate the consistency rate of the direction of change and the similarity of the magnitude of change of the digital signals of the sensor and its adjacent sensors within the same sampling period; In this embodiment, the signal change direction and amplitude changes of each sensor within a continuous sampling period are extracted. The change direction can be calculated from the difference in values ​​between adjacent sampling points and represented by an angle; the change amplitude is the absolute value of the difference in values ​​between adjacent sampling points. Subsequently, for each sensor, the consistency rate of its change direction and the similarity of its change amplitude with those of adjacent sensors within the same sampling period are calculated. The consistency rate of change direction is obtained by statistically analyzing the proportion of sampling points with a change direction difference of less than 30° within the sampling period; the similarity of change amplitude is obtained by statistically analyzing the proportion of adjacent sampling points with an amplitude change difference of less than 10%.

[0031] If the direction of change of any sensor is consistent with that of more than three adjacent sensors, and the similarity of the change amplitude is less than the similarity of the change amplitude, then the sensor is determined to be a valid sensor, and the sensor digital signal corresponding to the valid sensor is designated as the valid sensor digital signal.

[0032] In a further embodiment, if the direction of change of any sensor is consistent with that of more than three neighboring sensors by more than 80%, and the similarity of the magnitude of change is less than 10%, then the sensor is determined to be a valid sensor. The digital signal of the valid sensor will be extracted to form a valid sensor digital signal.

[0033] Optionally, performing noise cross-cancellation includes: Extract the signal strength sequence of the effective sensor and the three nearest adjacent sensors within the same sampling period from the digital signal conversion results; Calculate the residual mean sequence between the effective sensor digital signal and the adjacent sensor digital signals based on the signal strength sequence; In this embodiment, the digital signal intensity sequence of each effective sensor and its three closest adjacent sensors is extracted from the digital signal conversion result. The spatial distance can be calculated based on the preset three-dimensional coordinates of the sensor in the automatic monitoring station, ensuring that the selected adjacent sensors are the three closest devices with a distance of 0.5 to 1.5 meters. The sampling period is set to 500 milliseconds to form the adjacent sensor digital signal, which contains the signal intensity sequence of the sensor within the sampling period. Subsequently, the signal intensity residuals of each effective sensor and its adjacent sensors are calculated within the same sampling period to obtain the residual sequence. The mean of the residual sequence is further calculated to form a residual mean sequence matrix, where each element represents the average residual between a certain effective sensor and the signals of its neighboring sensors within a certain sampling period.

[0034] The moving average value in the residual mean sequence is used as the environmental background noise benchmark, and the fluctuation component of each sensor's digital signal under the environmental background noise benchmark is extracted. In a further embodiment, the residual mean sequence is averaged using a sliding window, with the window length set to 5 sampling periods, to obtain an environmental background noise reference sequence, which is used to extract the fluctuation components of each sensor's digital signal. The fluctuation components are calculated by subtracting the background noise reference obtained from the moving average from each sensor's digital signal, resulting in a high-frequency fluctuation component matrix, which represents the fluctuation components of each sensor's digital signal under the environmental background noise reference.

[0035] The fluctuation components of adjacent sensor digital signals are used as weighting terms to perform weighted inverse superposition on the fluctuation components of the effective sensor digital signals, and the weighted inverse superposition results of the effective sensor digital signals are integrated to obtain the device sensing data.

[0036] In a further embodiment, the fluctuation components of the digital signals from adjacent sensors are used as weighting terms to perform weighted inverse superposition on the fluctuation components of the effective sensor digital signals. The weighting coefficient can be set to be proportional to the reciprocal of the spatial distance between adjacent sensors, with higher weights for closer sensors, and the coefficient range is 0.25~0.4. After weighted inverse superposition, the superposition result is integrated with the original fluctuation components of the effective sensors to form the device sensing data.

[0037] It is worth noting that the integration of the superposition result with the original fluctuation components of the effective sensor can be achieved through element-by-element fusion. The weighted inverse superposition result is weighted and averaged with the fluctuation components of the effective sensor using a preset coefficient (e.g., 0.6) and a supplementary coefficient (e.g., 0.4) to obtain the corrected signal fluctuation matrix. Subsequently, the corrected fluctuation matrix is ​​added point-by-point to the baseline value of the original digital signal from the effective sensor (i.e., the reference value sequence after digital signal conversion) to obtain the device sensing data.

[0038] Optionally, determining abnormal sampling points in step S3 includes: Extract the numerical sequence of each sensor within a continuous sampling period from the device's sensing data, and calculate the magnitude and direction of change of each sensor between adjacent sampling points based on the numerical sequence; In this embodiment, the numerical sequence of each sensor within a continuous sampling period is extracted from the generated device sensing data. The sampling period is set to T = 1 second, and the total number of sampling points is N = 120, forming a time series vector of length 120. For each time series, the numerical difference between adjacent sampling points is calculated to obtain the amplitude change value of each sampling point and the direction of change between adjacent points. The direction of change can be represented by the sign of the vector difference, indicating an upward or downward trend.

[0039] Based on the variation amplitude and direction of each sensor between adjacent sampling points, calculate the average variation amplitude and direction distribution density of similar sensors within the same continuous sampling period; In a further embodiment, the amplitude change values ​​and change directions of all similar sensors are statistically analyzed within the same continuous sampling period. The average amplitude change value and the change direction distribution density are calculated respectively. The change direction distribution density can be divided into several angle segments in the range of 0° to 180° and the occurrence frequency of each segment is counted to obtain the probability distribution of the change direction of each sampling point in the group of sensors, which is convenient for identifying abnormal direction deviations.

[0040] If the change amplitude between any sampling point and its adjacent sampling points exceeds 1 to 3 times the average change amplitude, and the angle difference between the change direction and the average change direction angle of similar sensors in the same continuous sampling period in the change direction distribution density is higher than the direction difference angle difference threshold, then the sampling point is determined to be an abnormal sampling point.

[0041] In a further embodiment, for each sampling point, it is determined whether its amplitude change value exceeds the range of 1.5 to 2.5 times the average amplitude. If it does, the amplitude is considered abnormal. Simultaneously, the angle difference between the direction of change of this sampling point and the average direction of change of similar sensors is calculated. If the angle difference is greater than 30° (as a directional difference angle threshold), the sampling point is determined to be an abnormal sampling point. This method ensures that only when both amplitude and directional abnormalities are met can a point be marked as abnormal, avoiding false positives.

[0042] Optionally, performing drift correction in step S3 includes: Calculate the difference in amplitude and the change in direction angle between abnormal sampling points and adjacent sampling points within the continuous sampling period based on the numerical sequence of each sensor. In this embodiment, the numerical sequence of each abnormal sampling point and its adjacent sampling points before and after it is extracted from the device sensing data before sensing correction. Assuming the continuous sampling period is T = 1 second, the value of each sampling point is denoted as... The values ​​of the sampling points before and after are respectively denoted as and Calculate the amplitude difference between the abnormal sampling point and its adjacent sampling points. And calculate the change in direction angle. The direction angle can be represented by the angle between the slope vector of adjacent points and the positive direction, which makes it easy to judge the deviation of the upward or downward trend of the value.

[0043] Based on the amplitude difference and directional angle change between abnormal sampling points and their adjacent sampling points, and taking the average amplitude and directional distribution density of similar sensors within the same continuous sampling period as reference values, offset correction values ​​are calculated according to preset weight ratios; the offset correction values ​​include amplitude correction values ​​and directional angle correction values. In a further embodiment, the amplitude difference of each abnormal sampling point is... and change in direction and angle Average amplitude change compared to similar sensors within the same continuous sampling period Distribution density in the direction of change A comparative analysis was conducted, using preset weighting ratios. , For amplitude correction value and direction angle correction value Perform weighted calculations separately: amplitude correction value Direction angle correction value ;in For density based on angle distribution The determined average change direction of similar sensors, As a correction reference, ensure that the offset correction takes into account both the changes of the abnormal sampling points themselves and the trends of the group of sensors, and avoid errors introduced by extreme values ​​at a single point.

[0044] If the offset correction value of the abnormal sampling point is within the preset correction tolerance range, then the offset correction value is taken as the drift correction result.

[0045] Of particular importance, the drift correction process also includes: If the offset correction value of an abnormal sampling point is not within the correction tolerance range, the weight ratio is iteratively corrected until the difference between the offset correction value and the amplitude of the adjacent sampling point is within the correction tolerance range.

[0046] In a further embodiment, the calculated amplitude correction value is... and direction angle correction value Tolerance judgment is performed, assuming the amplitude tolerance range is ±0.05 units. If the error falls within this tolerance range, the correction is considered reasonable; otherwise, the weight ratio can be adjusted iteratively. and This continues until the correction value falls within the tolerance range. Finally, the magnitude correction value is... With direction angle correction value It is applied to abnormal sampling points, replaces the original values, generates drift-corrected sampling point data, and updates the device sensing data for subsequent sensing correction data generation and simulation.

[0047] Optionally, determining the operation and maintenance management strategy in step S4 includes: The operation and maintenance management scenario with high similarity to the simulation results is matched by a pre-set operation and maintenance management knowledge base and spare parts library. In this embodiment, feature extraction is performed on the simulation results, including the operating load, temperature rise curve, and maintenance timeliness data sequence of each device. It is assumed that the continuous sampling period is T=1 second, the unit of operating load is kW, the unit of temperature rise is °C, and the unit of maintenance timeliness is days. These feature vectors are then compared with the feature vectors of each historical operation and maintenance management scenario in a pre-built operation and maintenance management knowledge base. Cosine similarity or Euclidean distance is used as the matching standard. When the similarity is higher than 0.85, it is determined to be a high-similarity operation and maintenance management scenario. Simultaneously, available spare parts information for each scenario is extracted from the spare parts library and associated with historical policies in the operation and maintenance management knowledge base to form a complete scenario-policy-spare parts mapping table for subsequent policy correction.

[0048] Based on the differences between highly similar operation and maintenance management scenarios and operation simulation results, operational status deviations are determined; these operational status deviations include operational load deviations, temperature rise deviations, and maintenance timeliness deviations. In a further embodiment, a difference analysis is performed between the matched high-similarity operation and maintenance management scenarios and the simulation results. The operational state deviation is calculated, including: operational load deviation. Temperature rise deviation Maintenance timeliness deviation Among them, the simulated operating load Simulated temperature rise Simulation maintenance timeliness Historical operating load is used as a basis for simulation results. Historical temperature rise Historical maintenance period This is the historical average. If the operating load deviation exceeds ±5kW, the temperature rise deviation exceeds ±2°C, or the maintenance timeliness deviation exceeds ±3 days, then a strategy correction is required.

[0049] Historical operation and maintenance management strategies corresponding to highly similar operation and maintenance management scenarios are extracted from the operation and maintenance management knowledge base, and the historical operation and maintenance management strategies are corrected according to the deviation of the operation status to obtain the operation and maintenance management strategy.

[0050] In this embodiment, historical operation and maintenance management strategies corresponding to highly similar scenarios are extracted from the operation and maintenance management knowledge base, including equipment operation intensity grading schemes, cooling trigger threshold settings, and maintenance cycle settings. These historical strategies are then dynamically adjusted based on operational status deviations: if... If the threshold is exceeded, the equipment operation intensity classification will be linearly adjusted by increasing or decreasing the threshold; if If the threshold is exceeded, adjust the equipment cooling trigger threshold to ensure temperature control remains within a safe range; if If the deviation is too large, adjust the maintenance cycle setting to ensure that the equipment maintenance timeliness matches the current operating status. Finally, combine the adjustment results with weighted coefficients. , , (For example, 0.5, 0.3, and 0.2 respectively) are weighted and merged to generate the final operation and maintenance management strategy, including the revised operation intensity classification table, cooling trigger threshold, and maintenance cycle.

[0051] Optionally, correcting historical operation and maintenance management strategies based on operational status deviations includes: Based on the deviation in operating status, the corresponding historical operation and maintenance management strategy is dynamically adjusted. If the deviation in operating load exceeds the normal load threshold, the equipment operation intensity classification in the historical operation and maintenance management strategy is corrected. If the temperature rise deviation exceeds the temperature control threshold, the equipment cooling trigger threshold in the historical operation and maintenance management strategy is adjusted. If the maintenance timeliness deviation is lower than the preset maintenance timeliness threshold, the maintenance cycle setting value in the historical operation and maintenance management strategy is corrected. In this embodiment, the historical operation and maintenance management strategy is dynamically adjusted based on the operational status deviation. Taking the equipment operation intensity classification as an example, it is assumed that the historical strategy divides the load into five levels: 0–20kW, 20–40kW, 40–60kW, 60–80kW, and 80–100kW. The simulation results calculate the operational load deviation... When the load exceeds ±5kW, the load ranges are adjusted linearly. For example, if... If the power is increased by 8kW, the original 40–60kW range will be moved up by 8kW, making the range 48–68kW, thereby correcting the operating intensity classification table to adapt to actual load changes.

[0052] In another embodiment, the device cooling trigger threshold in the historical strategy is adjusted based on the temperature rise deviation. Assuming the historical cooling trigger threshold is 60°C, the simulation results show a temperature rise deviation... When the temperature exceeds ±2°C, the threshold is adjusted linearly according to the deviation magnitude, for example... If the temperature is +3°C, the trigger threshold will be adjusted to 63°C to ensure that the device can start the cooling mechanism in time under high temperature conditions, while avoiding frequent starts that would increase energy consumption.

[0053] In another embodiment, the maintenance cycle setting in the historical strategy is corrected to address the maintenance timeliness deviation. Assuming the historical maintenance cycle is set to 30 days, simulation results calculate the maintenance timeliness deviation. If the maintenance cycle is -5 days, it will be shortened to 25 days to ensure that the equipment is maintained in a timely manner even under low time sensitivity.

[0054] Based on highly similar operation and maintenance management scenarios, determine the equipment performance impact of operating load deviation, temperature rise deviation, and maintenance timeliness deviation respectively; In this embodiment, based on the correlation between load variation and equipment failure rate or performance degradation in highly similar operation and maintenance management scenarios, the performance impact of the equipment under the current deviation is calculated. For example, the average lifespan reduction percentage of similar equipment in historical data when the load exceeds the rated value by ±5% is used as a reference, and linear interpolation or curve fitting is used to obtain the result. Corresponding performance impact The unit is the equipment performance impact ratio (0–1). This is for temperature rise deviation. Based on the equipment temperature control response, energy consumption, and losses when the temperature rise exceeds the cooling trigger threshold in highly similar operation and maintenance management scenarios, a mapping table is established between temperature rise deviation and equipment stability or lifespan. According to the current... By comparing with the mapping table, the impact of temperature rise deviation on equipment performance can be obtained. This reflects the relative impact of temperature control deviation on equipment operating efficiency and safety. It also addresses maintenance timeliness deviations. This study statistically analyzes the impact of earlier or later maintenance cycles on equipment failure rates in highly similar operation and maintenance management scenarios, and converts the failure probabilities corresponding to each cycle deviation into performance impact quantities. This is used to quantitatively represent the impact of delayed or advanced maintenance on the overall performance and availability of the equipment.

[0055] Using the impact of equipment performance as a weighting coefficient, a weighted fusion calculation is performed on the correction results of equipment operation intensity classification, equipment cooling trigger threshold adjustment results, and maintenance cycle setting value correction results to obtain the operation and maintenance management strategy.

[0056] In this embodiment, the adjustment results are weighted and fused according to their respective weighting coefficients to obtain the final operation and maintenance management strategy. Specifically, the adjustment result for equipment operating intensity classification is multiplied by the weighting coefficient. +Cooling trigger threshold adjustment result× +Maintenance cycle setting correction result× This process generates a comprehensive strategy table applicable to the current operating status. The strategy table includes revised load grading, cooling thresholds, and maintenance cycles, and can be directly used for remote control task assignment or on-site maintenance execution, ensuring safe and efficient operation of the equipment under different operating conditions.

[0057] Most importantly, the methods for obtaining the operations and maintenance management knowledge base include: Acquire historical equipment operation data of automatic monitoring stations, and extract historical equipment operation numerical data, historical maintenance records and manufacturer design parameters from the historical equipment operation data of automatic monitoring stations; In this embodiment, historical operating data of the equipment is obtained from the historical database of the automatic monitoring station. This data includes various numerical signal sequences collected by sensors, equipment status logs, and maintenance records. Historical equipment operating data is stored in the form of a time-series matrix, where rows represent different sampling time points, columns represent various equipment indicators (such as water flow rate, pump load, motor speed, temperature, etc.), and matrix elements are floating-point values. Historical maintenance records are stored in an event list structure, with each record containing maintenance start time, maintenance end time, maintenance type, and operator information. Manufacturer design parameters are stored in the form of a parameter dictionary, including equipment rated load, temperature control threshold, recommended operating cycles, and safe operating limits.

[0058] Align the timestamps between historical equipment operation data and historical maintenance records, and determine the correlation between equipment maintenance strategies and numerical changes based on the magnitude of changes in historical equipment operation data and the maintenance timestamps in historical maintenance records. In a further embodiment, historical equipment operation data and historical maintenance records are aligned according to timestamps. During the alignment process, using second-level timestamps as a benchmark, the operation data interval corresponding to each maintenance event is extracted to form an operation-maintenance comparison dataset. In this comparison dataset, the change in each equipment indicator before and after maintenance (e.g., load change) is calculated. Temperature rise change ) and maintenance operation interval The correlation between the parameters was analyzed to determine the response characteristics of each maintenance strategy to equipment operating parameters. The correlation was calculated using the Pearson correlation coefficient or the weighted mean square error method, with a reference threshold set between 0.6 and 0.8 to ensure a significant correlation between the selected maintenance strategies and the numerical changes.

[0059] Effective maintenance strategies are screened based on the correlation between equipment maintenance strategies and numerical changes, and a correspondence is established between effective maintenance strategies and abnormal equipment operation modes in historical equipment operation data. In a further embodiment, effective maintenance strategies are screened based on the aforementioned correlation analysis. An effective maintenance strategy refers to a historical maintenance plan that highly matches the abnormal operating mode of the equipment. The criteria for determining an effective maintenance strategy include: the key parameters of the equipment returning to normal range after the strategy is implemented; the frequency of strategy re-execution being lower than a preset threshold (e.g., no more than 4 times per year); and the strategy conforming to the manufacturer's design parameter constraints. These effective maintenance strategies are then correlated with the abnormal operating modes in the historical equipment operating data. Each abnormal operating mode is represented by a state mode vector, such as [high load, abnormal temperature rise, excessive vibration]. The corresponding maintenance strategy is represented by a list of operation steps, such as [adjust pump flow, start cooling device, replace filter element].

[0060] It is worth noting that the methods for identifying abnormal operating modes of equipment include: Key indicator columns, such as pump load, motor speed, temperature rise, and vibration value, are extracted from the historical equipment operation data matrix. These indicators are then standardized to allow for unified analysis of data with different dimensions. The standardized indicator columns are divided into continuous sampling windows, for example, each window containing sampling points within a 60-second period, forming a window feature vector. Within each window, the mean, variance, maximum, and minimum values ​​of each indicator are calculated as features to describe the equipment's operating status during that time period. Example operating parameters: window length 60 seconds; feature calculations include mean, variance, and extreme values. Then, based on the statistical characteristics of historical operating data and the manufacturer's design parameters, abnormal thresholds are set for each indicator. For example, pump load exceeding ±15% of rated load, temperature rise exceeding the design threshold by ±2°C, and motor vibration exceeding the safety value by ±5%. If any indicator in the window's feature vector exceeds the threshold, the window is considered to be in an abnormal operating state. Subsequently, continuous abnormal windows are clustered chronologically to form abnormal operating patterns. Clustering methods can employ hierarchical clustering based on Euclidean distance or density clustering (DBSCAN) to group similar feature vectors into the same abnormal operating pattern. Example of operating parameters: In DBSCAN, epsilon is set to 0.5, and the minimum number of sample points is set to 3. Finally, each abnormal operating mode is represented as a state mode vector, including the type and magnitude of each abnormal indicator, such as [high load, abnormal temperature rise, vibration exceeding limits], and the time distribution and duration of the mode are recorded.

[0061] The knowledge base relationship framework is constructed by using the correspondence between effective maintenance strategies and abnormal operating modes of equipment in historical equipment operation data as the basis, and the manufacturer's design parameters as constraints.

[0062] In a further embodiment, the correspondence between effective maintenance strategies and abnormal equipment operation modes is used as the core framework of the knowledge base, with manufacturer design parameters added as constraints to form a complete operation and maintenance management knowledge base. The knowledge base structure adopts a key-value mapping format: the key is an abnormal operation mode vector, and the value is the corresponding maintenance strategy and parameter constraint set; it also includes metadata, including strategy execution effects, historical failure rates, and applicable equipment models. Example of operating parameters: time alignment accuracy of 1 second, correlation threshold of 0.7, abnormal operation mode feature dimension of 5~10, and maintenance strategy constraints including load threshold ±10%, temperature control threshold ±2°C, and maintenance cycle ±7 days.

[0063] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0064] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for intelligent operation and maintenance management of a digital twin automatic water quality monitoring station, characterized in that, Includes the following steps: Step S1: Collect the sensor signal set of the device by pre-deploying the sensor cluster at the automatic monitoring station, and transmit the sensor signal set of the device to the pre-deployed edge node to perform digital signal conversion; Step S2: Perform sensor signal coordination based on the digital signal conversion result to obtain device sensing data; Step S3: Determine abnormal sampling points based on the magnitude of changes in the device's sensor data; Perform drift correction on abnormal sampling points and replace the drift correction results with the corresponding sampling values ​​in the device's sensing data to obtain the sensing correction data; Step S4: Combine the sensor correction data with the preset device mechanism model to construct a device operation twin model; The equipment operation twin model is used to perform equipment operation simulation, and the operation and maintenance management strategy is determined based on the simulation results.

2. The intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station according to claim 1, characterized in that, After determining the operation and maintenance management strategy in step S4, the following steps are also included: The operation and maintenance management strategy is sent to the automatic monitoring station management platform, and maintenance task orders are distributed to the execution units according to the deviation between the operation and maintenance management strategy and the existing operation and maintenance management strategy in the automatic monitoring station management platform. By using a sensor cluster to synchronously acquire real-time device sensor signal sets and performing digital signal conversion in coordination with sensor signals, real-time device sensing data can be obtained. Compare real-time device sensor data with simulation results to verify the simulation results; The operation and maintenance management strategy is iteratively optimized based on the verification numerical error in the verification results until the verification numerical error is less than the preset error threshold.

3. The intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station according to claim 2, characterized in that, The categorized distribution of maintenance task orders includes: If the deviation of operation and maintenance management is within the preset remote adjustment range, the remote execution unit corresponding to the operation and maintenance management strategy will be determined through the automatic monitoring station management platform, and the control instructions in the operation and maintenance management strategy and the remote control links corresponding to the remote execution units will be integrated into a remote control task sheet and distributed to the corresponding remote execution units. If the deviation in operation and maintenance management exceeds the remote adjustment range, the maintenance operation guidelines, equipment repair plans, and spare parts configuration requirements in the operation and maintenance management strategy will be integrated into a high-priority maintenance task order, and the high-priority maintenance task order will be pushed to the on-site maintenance execution unit in the corresponding area of ​​the automatic monitoring station through the automatic monitoring station management platform.

4. The intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station according to claim 1, characterized in that, The sensor signal coordination in step S2 includes: Based on the transmission timestamp sequence of each sensor signal transmitted to the edge node in the digital signal conversion result, the execution time of each sensor digital signal is synchronized and paired. Acquire the spatial distribution information of each sensor in the sensor cluster, and detect the signal change trend of adjacent sensors in the same sampling period based on the spatial distribution information of each sensor in the time synchronization pairing results; Based on the detection results of signal change trends, effective sensor digital signals are selected, and noise cross-cancellation is performed on the effective sensor digital signals to obtain the results of cooperative sensor digital signals.

5. The intelligent operation and maintenance management method for a digital twin automatic water quality monitoring station according to claim 4, characterized in that, Filtering valid sensor digital signals includes: Calculate the consistency rate of the direction of change and the similarity of the magnitude of change of the digital signals of the sensor and its adjacent sensors within the same sampling period; If the direction of change of any sensor is consistent with that of more than three adjacent sensors, and the similarity of the change amplitude is less than the similarity of the change amplitude, then the sensor is determined to be a valid sensor, and the sensor digital signal corresponding to the valid sensor is designated as the valid sensor digital signal.

6. The intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station according to claim 4, characterized in that, Performing noise cross-cancellation includes: Extract the signal strength sequence of the effective sensor and the three nearest adjacent sensors within the same sampling period from the digital signal conversion results; Calculate the residual mean sequence between the effective sensor digital signal and the adjacent sensor digital signals based on the signal strength sequence; The moving average value in the residual mean sequence is used as the environmental background noise benchmark, and the fluctuation component of each sensor's digital signal under the environmental background noise benchmark is extracted. The fluctuation components of adjacent sensor digital signals are used as weighting terms to perform weighted inverse superposition on the fluctuation components of the effective sensor digital signals, and the weighted inverse superposition results of the effective sensor digital signals are integrated to obtain the device sensing data.

7. The intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station according to claim 1, characterized in that, Step S3 involves identifying anomalous sampling points, including: Extract the numerical sequence of each sensor within a continuous sampling period from the device's sensing data, and calculate the magnitude and direction of change of each sensor between adjacent sampling points based on the numerical sequence; Based on the variation amplitude and direction of each sensor between adjacent sampling points, calculate the average variation amplitude and direction distribution density of similar sensors within the same continuous sampling period; If the change amplitude between any sampling point and its adjacent sampling points exceeds 1 to 3 times the average change amplitude, and the angle difference between the change direction and the average change direction angle of similar sensors in the same continuous sampling period in the change direction distribution density is higher than the direction difference angle difference threshold, then the sampling point is determined to be an abnormal sampling point.

8. The intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station according to claim 1, characterized in that, Step S3 includes performing drift correction, which includes: Calculate the difference in amplitude and the change in direction angle between abnormal sampling points and adjacent sampling points within the continuous sampling period based on the numerical sequence of each sensor. Based on the amplitude difference and directional angle change between abnormal sampling points and their adjacent sampling points, and taking the average amplitude and directional distribution density of similar sensors within the same continuous sampling period as reference values, offset correction values ​​are calculated according to preset weight ratios; the offset correction values ​​include amplitude correction values ​​and directional angle correction values. If the offset correction value of the abnormal sampling point is within the preset correction tolerance range, then the offset correction value is taken as the drift correction result.

9. The intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station according to claim 1, characterized in that, Step S4, which determines the operation and maintenance management strategy, includes: The operation and maintenance management scenario with high similarity to the simulation results is matched by a pre-set operation and maintenance management knowledge base and spare parts library. Based on the differences between highly similar operation and maintenance management scenarios and operation simulation results, operational status deviations are determined; these operational status deviations include operational load deviations, temperature rise deviations, and maintenance timeliness deviations. Historical operation and maintenance management strategies corresponding to highly similar operation and maintenance management scenarios are extracted from the operation and maintenance management knowledge base, and the historical operation and maintenance management strategies are corrected according to the deviation of the operation status to obtain the operation and maintenance management strategy.

10. The intelligent operation and maintenance management method for a digital twin water quality automatic monitoring station according to claim 9, characterized in that, Correcting historical operation and maintenance management strategies based on operational status deviations includes: Based on the deviation in operating status, the corresponding historical operation and maintenance management strategy is dynamically adjusted. If the deviation in operating load exceeds the normal load threshold, the equipment operation intensity classification in the historical operation and maintenance management strategy is corrected. If the temperature rise deviation exceeds the temperature control threshold, the equipment cooling trigger threshold in the historical operation and maintenance management strategy is adjusted. If the maintenance timeliness deviation is lower than the preset maintenance timeliness threshold, the maintenance cycle setting value in the historical operation and maintenance management strategy is corrected. Based on highly similar operation and maintenance management scenarios, determine the equipment performance impact of operating load deviation, temperature rise deviation, and maintenance timeliness deviation respectively; Using the impact of equipment performance as a weighting coefficient, a weighted fusion calculation is performed on the correction results of equipment operation intensity classification, equipment cooling trigger threshold adjustment results, and maintenance cycle setting value correction results to obtain the operation and maintenance management strategy.

Images