Electric power facility intelligent lock management and control method based on multi-dimensional authority data processing

Abstract

The invention relates to an electric power facility intelligent lock management and control method based on multi-dimensional right limit data processing, and the method comprises the steps: obtaining a curve of magnetic induction signal intensity and waveform changing with time during the operation of an intelligent lock, and carrying out the preliminary filtering processing of data through the analysis of waveform distortion and intensity attenuation characteristics, and obtaining a change trend sequence; the waveform distortion degree increment and the signal strength reduction rate caused by wear of the intelligent lock body are extracted through the change trend sequence, and the grading index of the wear degree of the intelligent lock body is determined; the lock cylinder gap change is combined with historical door body matching state data, the dynamic mapping relation between the magnetic induction signal intensity and the gap change is calculated, and the magnetic induction threshold value range is dynamically adjusted; and whether the intensity of the magnetic induction signal meets the target locking judgment standard or not is judged in real time, the lock control parameters are dynamically updated when the lock body runs, and optimized lock control state data are obtained.

Description

Technical Field

[0001] This application relates to the field of electrical digital data technology, and particularly to an intelligent lock control method for power facilities based on multi-dimensional permission data processing. Background Art

[0002] The intelligent lock control of power facilities is a key area for ensuring the safe and efficient operation of energy infrastructure, and its importance is reflected in the direct impact on equipment safety, operation stability, and maintenance efficiency. With the improvement of the intelligence level of power facilities, as a core component, the intelligent lock not only needs to ensure high security but also adapt to the complex environment during long-term operation. Currently, intelligent lock control technologies based on multi-dimensional permission data have been widely applied, but existing methods have significant limitations in dealing with the performance degradation problem during the long-term use of the lock body. Most solutions rely on static thresholds or fixed decision-making logics, which are difficult to adapt to the dynamic changes caused by the wear of mechanical components, and are prone to false judgments or safety hazards. Especially when the clearance between the lock core and the door body increases, traditional methods cannot effectively capture the subtle features of early performance degradation. The core challenges in this field focus on how to accurately perceive and predict the performance changes caused by the wear of the lock body. The increase in the clearance between the lock core and the door body leads to an increase in the waveform distortion degree of the magnetic induction signal, making the determination of the closed state ambiguous, and the change pattern of the key extraction resistance also changes accordingly. These technical factors directly affect the reliability and safety of the lock. If relevant parameters cannot be dynamically adjusted in a timely manner, it may lead to locking failure or an increase in maintenance costs.

[0003] When dealing with these dynamic changes, existing technologies often lack an adaptive prediction mechanism and are difficult to identify wear characteristics and take effective interventions at an early stage. Therefore, how to construct a predictive maintenance algorithm that can dynamically adjust the magnetic induction threshold parameter and the locking determination tolerance based on the change in the magnetic induction signal intensity has become a key issue for extending the service life of equipment while ensuring safety. Summary of the Invention

[0004] In order to solve the problems existing in the above-mentioned prior art, the purpose of this application is to provide an intelligent lock control method for power facilities based on multi-dimensional permission data processing.

[0005] The intelligent lock control method for power facilities based on multi-dimensional permission data processing described in this application includes: After obtaining the magnetic induction signal sequence during the operation of the intelligent lock, perform denoising to generate a denoised signal sequence; Calculate and extract the attenuation parameters of the hyperbolic attenuation model of the signal amplitude in the original magnetic induction signal sequence for the denoised signal sequence, and fit the change trend sequence; Taking the change trend sequence as the input, calculate the difference between the current waveform distortion degree and the historical reference value as the distortion increment and decompose it, construct a support vector regression model, output the wear quantification score of the lock body wear degree, and divide the wear index according to the score; After dividing the wear index, generate a gap change curve according to the judgment result of whether the predicted value of the lock core gap exceeds the preset threshold; Taking the predicted value of the lock core gap as a parameter, establish a database of the door body matching state; and construct a gap-magnetic induction intensity mapping model, output the signal intensity compensation coefficient and perform segmented optimization, and dynamically adjust the interval boundary of the segmented optimization; When the segmented optimization belongs to the case where the high-region threshold is continuously exceeded, output the tolerance increment of the gap between the lock core and the door body in the next 7 days; calculate the magnetic induction fluctuation compensation amount according to the tolerance increment; generate the target determination standard curve; Based on the target determination standard of the target determination standard curve, divide the magnetic induction signal intensity level, and dynamically adjust the torque output of the intelligent lock motor to drive the lock core or the lock tongue; According to the lock control state data of the torque output, construct a deep neural network misjudgment probability model, output the misjudgment score, and dynamically adjust the weight factor of the wear parameter according to the risk level until the torque fluctuation of the intelligent lock motor driving the lock core or the lock tongue meets the preset parameters.

[0006] The advantages of a power facility intelligent lock control method based on multi-dimensional permission data processing described in this application are as follows: by obtaining the magnetic induction signal intensity and the waveform change curve, analyzing the waveform distortion degree and the intensity attenuation characteristics, extracting the lock body wear index, establishing a wear prediction model, combining with the historical door body matching state data, calculating the dynamic mapping relationship between the magnetic induction signal and the gap change, adjusting the magnetic induction threshold range, according to the lock body wear degree and the predicted value of the lock core gap change, identifying the lock core tolerance range, obtaining the target locking determination standard, real-time judging whether the magnetic induction signal meets the determination standard, dynamically updating the lock control parameters, evaluating the misjudgment probability, judging the safety hazard risk, and adjusting the threshold weight factor when necessary, and re-optimizing the control strategy. The present invention can effectively predict the wear state of the intelligent lock body, dynamically optimize the lock control parameters, improve the locking accuracy and safety, and extend the service life of the intelligent lock. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 is the flow chart of a power facility intelligent lock control method based on multi-dimensional permission data processing described in this application Figure 1 ; Figure 2 is the flow chart of a power facility intelligent lock control method based on multi-dimensional permission data processing described in this application Figure 2 . DETAILED DESCRIPTION OF THE EMBODIMENTS

[0008] Such asFigure 1 - Figure 2 As shown in the figure, a method for controlling an intelligent lock of a power facility based on multi-dimensional permission data processing according to the present application includes the following steps: S101. Obtain the curve of the magnetic induction signal intensity and waveform changing with time during the operation of the intelligent lock, preprocess the signal data to obtain a change trend sequence; S102. Extract the change in the magnetic induction signal characteristics caused by the wear of the intelligent lock body based on the change trend sequence, quantify the waveform distortion increment and the signal intensity attenuation rate, and determine the grading index of the lock body wear degree; S103. Use the historical intelligent lock body wear data to construct a prediction model, combine the wear degree grading index and the change trend sequence to predict the wear state of the intelligent lock body within a preset time, and generate a change curve of the lock core gap through fitting analysis; S104. Combine the change in the lock core gap and the historical door body matching state data to construct a dynamic mapping relationship between the magnetic induction signal intensity and the gap parameters, and dynamically optimize the magnetic induction threshold range to improve the determination accuracy of the intelligent lock; S105. According to the lock body wear degree grading index and the predicted value of the lock core gap, when the adjusted magnetic induction threshold exceeds the preset upper limit, identify the lock core tolerance range and generate a target locking determination standard to ensure the reliable operation of the intelligent lock; S106. Real-time monitor whether the magnetic induction signal intensity meets the target locking determination standard, and dynamically adjust the lock control parameters to optimize the operation state of the intelligent lock; S107. Combine the optimized lock control state data and the change trend sequence, evaluate the misjudgment probability of the intelligent lock operation state and determine whether the safety risk is lower than the preset threshold. If the risk exceeds the standard, adjust the weight factors of the magnetic induction threshold and the determination tolerance, and re-optimize the lock control parameters to ensure the safe operation of the intelligent lock.

[0009] As Figure 1 - Figure 2 shown, in step S101, obtain the curve of the magnetic induction signal intensity and waveform changing with time during the operation of the intelligent lock, preprocess the signal data to obtain a change trend sequence.

[0010] Further, in step S101, in the embodiment of the present application, when the control function of the intelligent lock is triggered, the system first collects the magnetic induction signal data during operation through a magnetic induction sensor, and preprocesses the data to obtain a change trend sequence; wherein the above system can be an intelligent lock control system, which can be controlled by a computer or an MCU chip, and is signal-connected to the intelligent lock, and the signal connection method can be a direct connection or an indirect connection; Collect the original magnetic induction signal data sequence from the magnetic induction sensor of the intelligent lock. The original magnetic induction signal data sequence is the magnetic induction signal data sequence collected from the magnetic induction sensor and not processed, and record the signal amplitude, sampling time and frequency parameters; In the embodiments of the present application, the magnetic induction sensor continuously collects the original magnetic induction signal data sequence during the operation of the intelligent lock at a sampling interval of 10 milliseconds, and records the signal amplitude, sampling time, and frequency parameters of each data point. For example, during the closing process of the lock, the signal amplitude may gradually decrease from 2.8 millitesla to 0.6 millitesla, lasting about 300 milliseconds. The embodiments of the present application do not overly limit the specific parameters of sampling and can be adjusted according to the actual scenario; The original magnetic induction signal data sequence is filtered by a digital filter to generate a denoised signal sequence; To remove noise interference, the embodiments of the present application use a Butterworth digital filter to perform third-order high-pass filtering and second-order low-pass filtering on the original signal, filtering out high-frequency noise above 500 Hz and low-frequency interference below 5 Hz, and generating a filtered signal sequence; High-pass filtering can remove the baseline offset caused by geomagnetic field fluctuations or temperature drift, and low-pass filtering can filter out sensor thermal noise or external electromagnetic interference; Perform spectral analysis on the denoised signal sequence to extract the waveform distortion degree and signal attenuation characteristics; In the embodiments of the present application, spectral analysis is performed on the denoised signal sequence through a 512-point discrete Fourier transform, the amplitude ratio of the fundamental wave component to the harmonic component is calculated to obtain the waveform distortion degree index, and at the same time, the attenuation curve of the signal amplitude changing with time is recorded. Under normal circumstances, the fundamental frequency of the intelligent lock is about 50 Hz, the second harmonic component does not exceed 15% of the fundamental wave, and the third harmonic does not exceed 8%; Establish a signal intensity attenuation model based on the waveform distortion degree and attenuation curve, and fit the change trend curve; According to the waveform distortion degree and attenuation curve, the embodiments of the present application set the signal amplitude determination threshold interval to ±20% of the nominal value, for example, 2.0 to 3.0 millitesla. A hyperbolic fitting method is used to establish a signal intensity attenuation model, including an initial intensity coefficient of 2.5 millitesla and an attenuation rate coefficient of 0.008 / ms. Subsequently, the signal sequence is segmented by a 100-ms time window, the signal-to-noise ratio of each segment is calculated, the signal segments with a signal-to-noise ratio higher than 15 dB are selected, the peak-valley value data points (peak value about 2.8 millitesla, valley value about 2.2 millitesla) are extracted, and a change trend curve including the time stamp, magnetic induction intensity, and waveform distortion degree value is fitted by a cubic spline interpolation method; In the embodiments of the present application, the trend curve generated through the above steps can accurately characterize the magnetic field characteristics of the intelligent lock in different working states, provide a reliable data basis for wear prediction and parameter optimization, and can dynamically adjust the lock control parameters according to the trend curve to ensure the locking accuracy and security, and significantly improve the reliability and service life of the intelligent lock; In the embodiments of the present application, the design of the lock determines the state of the lock (unlocked or locked) based on the movement of the bolt in and out. The magnetic induction signal is collected by a magnetic induction sensor (Hall sensor) built into the smart lock. Its principle is based on the magnetic field change caused by the mechanical movement of the lock core. When the lock core rotates or the bolt moves, the relative position between the internal magnet and the sensor changes, resulting in the change of the magnetic induction intensity (unit: millitesla) and waveform output by the sensor.

[0011] The bolt is the part of the smart lock that directly cooperates with the lock catch on the door frame and is responsible for the physical locking or unlocking action. When the lock is in the closed state, the bolt will insert into the lock catch on the door frame; when the lock is opened, the bolt will retract into the lock body.

[0012] The permanent magnet is installed on the bolt or a component linked to the bolt. As the bolt moves in and out, the position of the permanent magnet relative to the magnetic induction sensor will change. The magnetic induction sensor (Hall sensor) is fixedly installed inside the lock body at a position where it can detect the magnetic field change caused by the permanent magnet, and the magnetic induction sensor is installed inside the lock body at a position where it can detect the magnetic field change when the bolt is fully extended or retracted. For example, when the bolt is fully extended, the magnetic induction sensor can detect the strongest magnetic field intensity; when the bolt is fully retracted, the weakest magnetic field intensity is detected. The sensor determines the state of the lock (e.g., whether the lock is fully closed) by monitoring the change in magnetic field intensity. In one way, the sensor is installed at a fixed point inside the lock body aligned with the movement path of the bolt, so that it can dynamically monitor the change in magnetic field intensity as the bolt moves in and out.

[0013] The lock core is directly or indirectly connected to the bolt through an internal mechanical structure (such as gears, linkages, etc.). This means that when the lock core rotates, it will drive the bolt to perform a corresponding linear movement (extend or retract). For example, after inserting the correct key and rotating it, the mechanism inside the lock core will be triggered, causing the bolt to retract from the lock catch on the door frame, thus opening the door.

[0014] The state of the bolt (whether it is extended or retracted) determines whether the door is locked. The function of the lock core is to provide a method to control the position of the bolt so that the position of the bolt can only be changed under specific conditions (such as correctly entering the password, using the correct key, or passing electronic authentication), thereby opening or closing the door lock.

[0015] The movement of the bolt can be used as part of the feedback signal. For example, the magnetic induction sensor can judge whether the bolt has been fully retracted or extended by detecting the position change of the permanent magnet connected to the bolt, so as to confirm the state of the lock.

[0016] When the locking operation is performed, the locking tongue moves outwards and inserts into the lock catch of the door frame. During this process, the permanent magnet connected to the locking tongue gradually approaches the magnetic induction sensor, resulting in a decrease in the gap between the two. Due to the shorter distance, the magnetic field strength detected by the magnetic induction sensor increases.

[0017] Conversely, when the unlocking operation is carried out, the locking tongue retracts inwards, away from the lock catch of the door frame, and at the same time drives the permanent magnet away from the magnetic induction sensor, causing an increase in the gap between the two. At this time, the magnetic field strength detected by the magnetic induction sensor weakens.

[0018] In addition, long-term mechanical wear will cause the gap of the lock core to expand, resulting in an increase in the distortion degree of the signal waveform (the proportion of harmonic components) and a non-linear attenuation characteristic of the signal intensity. The lock core gap refers to the distance between the lock core and the door body.

[0019] In the embodiment of the present application, the input source of the magnetic induction signal is: the original magnetic induction signal is a sequence of original signals collected in real time by the magnetic induction sensor at a sampling interval of 10 milliseconds, including the amplitude of the magnetic induction intensity, the time stamp, and the frequency parameter. For example, the signal amplitude in the closing stage drops from 2.8 millitesla to 0.6 millitesla and lasts for about 300 milliseconds; The steps for denoising the original magnetic induction signal to obtain the denoised signal sequence are as follows: For filtering and denoising, a Butterworth digital filter is used for third-order high-pass filtering (filtering out low-frequency drift below 5 Hz, temperature drift) and second-order low-pass filtering (filtering out high-frequency noise above 500 Hz, electromagnetic interference) to generate the denoised signal sequence.

[0020] Alternatively, for spectrum analysis, a 512-point discrete Fourier transform (DFT) is performed on the denoised signal sequence, the amplitude ratio of the fundamental wave to the harmonic components is calculated, and the waveform distortion degree (the third harmonic does not exceed 8% of the fundamental wave) is extracted; For attenuation modeling, based on the signal amplitude attenuation curve, a hyperbolic fitting method is used to establish a signal intensity attenuation model, and the parameters include the initial intensity coefficient (2.5 millitesla) and the attenuation rate coefficient (0.008 / ms); For trend fitting, the signal-to-noise ratio is calculated in segments according to a 100-ms time window (retained when SNR > 15 dB), the peak and valley data points are extracted (peak value 2.8 mT, valley value 2.2 mT), and a cubic spline interpolation is used to generate a change trend curve, including the time stamp, the magnetic induction intensity, and the waveform distortion degree value; Output data: a set of structured data, representing the time evolution law of the magnetic induction signal, for subsequent wear analysis and parameter optimization.

[0021] Such as Figure 1 - Figure 2As shown, in step S102, based on the change trend sequence, the change in the magnetic induction signal characteristics caused by the wear of the intelligent lock body is extracted, the waveform distortion increment and the signal intensity attenuation rate are quantified, and the grading index of the lock body wear degree is determined.

[0022] Further, in step S102, in the embodiment of the present application, by analyzing the change trend sequence, the magnetic induction signal characteristics reflecting the wear state of the intelligent lock body are extracted, the waveform distortion increment and the signal intensity attenuation rate are quantified, and a grading index of the lock body wear degree is generated based on multi-dimensional feature analysis, providing a basis for parameter optimization; Obtain the signal sequence during the operation of the lock body from the magnetic induction sensor and calculate the waveform distortion increment parameter; In the embodiment of the present application, the magnetic induction sensor of the intelligent lock collects a continuous magnetic induction signal sequence during the operation of the lock body at a sampling frequency of 100 Hz, records the signal amplitude, sampling time, and frequency information. The system uses the mean value of the waveform distortion of the historical operation data as the benchmark. Among them, the amplitude ratio of the fundamental wave to the third harmonic is calculated as the waveform distortion degree. For example, the benchmark value is set to 5%. By comparing the difference between the waveform distortion degree of the current signal sequence and the benchmark value, the waveform distortion increment parameter is calculated. For example, in the slight wear state, the increment usually does not exceed 10%; in the moderate wear state, the increment is between 15% and 25%; in the severe wear state, the increment may exceed 30%. The embodiment of the present application does not overly limit the specific setting of the benchmark value, which can be adjusted according to the lock body type and usage scenario; Combine the lock body rotation angle data, fit the signal intensity change curve, and calculate the signal intensity attenuation rate; In the embodiment of the present application, the angle sensor of the lock body is used to record the angle data during the rotation of the lock core. The sampling frequency is 100 Hz, and the rotation angle range is 0 to 180 degrees. The amplitude data of the magnetic induction signal at different angles is extracted, and a continuous curve of the signal intensity changing with the angle is fitted by the cubic spline interpolation method. The system calculates the slope of the curve at each angle point to obtain the signal intensity attenuation rate parameter. In the state of a new lock body, the signal intensity changes smoothly, and the absolute value of the slope usually does not exceed 0.02 mT / degree; as the wear intensifies, the absolute value of the slope may increase to more than 0.05 mT / degree, reflecting the instability during the operation of the lock body; Extract the frequency domain characteristics of the waveform distortion increment and the signal intensity attenuation rate through wavelet decomposition; To further analyze the signal characteristics, the embodiment of the present application performs four-layer wavelet decomposition on the waveform distortion increment parameter and the signal strength attenuation rate parameter respectively, and selects the db4 wavelet basis function to decompose the signal into five frequency bands: 0-2 Hz, 2-4 Hz, 4-8 Hz, 8-16 Hz and 16-32 Hz. The system calculates the energy distribution characteristics of each frequency band and generates a feature set containing low-frequency and high-frequency coefficients. In the new lock body state, the energy proportion of the low-frequency band (0-4 Hz) is usually more than 80%, reflecting a stable magnetic field distribution; as wear increases, the energy proportion of the high-frequency band (8-32 Hz) increases, indicating that more unstable vibrations occur during operation; Construct a multi-dimensional feature vector and quantify the degree of wear to generate a grading index; Based on the energy distribution characteristics extracted by wavelet decomposition, the embodiment of the present application constructs an eight-dimensional wavelet feature, including the energy proportion of the waveform distortion increment and the signal strength attenuation rate in each frequency band. The system inputs the 8-dimensional wavelet feature and the historical wear data into the support vector regressor based on the radial basis kernel function for processing. The kernel function parameter σ is set to 0.8, and the relaxation factor C is set to 10. The regressor outputs the wear quantification score value of the wear degree of the lock body, ranging from 0 to 1. The system divides the wear level according to the wear quantification score value of the wear degree of the lock body: 0 to 0.3 is slight wear, 0.3 to 0.6 is moderate wear, and 0.6 to 1.0 is severe wear. For example, when the score value is 0.4, it is judged as moderate wear. The classification index is determined by experimental data statistics to ensure the reliability and consistency of the judgment; Continuously monitor wear classification indicators and analyze wear trend; In the embodiment of the present application, the system continuously monitors the lock body wear grading index, records the time point when the wear quantification score value of the lock body wear degree exceeds the threshold of each interval, such as the moment when the wear degree is upgraded from slight wear to moderate wear, and uses the exponential smoothing method to calculate the rate of change of the wear degree, and the smoothing coefficient is set to 0.3. If the wear degree rises rapidly from a slight state to a moderate state, or from a moderate state to a severe state within 30 days, it indicates that the lock body wear rate is abnormal and further maintenance warnings need to be triggered. The system generates a trend sequence containing timestamps and wear levels to provide data support for dynamically adjusting lock control parameters; In the embodiments of the present application, by extracting the waveform distortion increment and the signal intensity attenuation rate, and combining wavelet decomposition and support vector regression analysis, the system can accurately quantify the wear degree of the lock body and generate a grading index. Compared with the traditional static threshold method, the embodiments of the present application significantly improve the sensitivity and accuracy of wear detection through multi-dimensional feature analysis and dynamic trend monitoring. For example, in actual tests, when the lock body is in a moderately worn state, the system can identify potential instabilities in advance through the increase in the energy ratio of the high-frequency band, thereby providing a reliable basis for optimizing the lock control parameters. The embodiments of the present application do not overly limit the specific dimensions of the wavelet features or the parameter settings of the regressor, and can be optimized and adjusted by technicians according to the actual scenario to adapt to different types of intelligent lock bodies and operating environments. The wear grading index and trend sequence generated by the system lay a foundation for subsequent predictive maintenance and dynamic parameter adjustment, effectively extending the service life of the intelligent lock and reducing potential safety hazards.

[0023] As Figure 1 - Figure 2 shown, in step S103, a prediction model is constructed using the historical wear data of the intelligent lock body. Combining the wear degree grading index and the change trend sequence, the wear state of the intelligent lock body within a preset time is predicted, and a change curve of the lock core gap is generated through fitting analysis.

[0024] Furthermore, in step S103, in the embodiments of the present application, by analyzing the historical wear data of the intelligent lock body, an adaptive prediction model is constructed. The wear degree grading index and the change trend sequence are input to predict the wear state of the lock body within a certain period in the future, and a change curve of the lock core gap is generated through a fitting method to provide a basis for dynamically optimizing the lock control parameters; Extract wear features from the historical data and group them to construct a wear trend prediction model; In the embodiments of the present application, the system extracts a data set containing the wear degree grading index and the change trend sequence from the historical operation records of the intelligent lock body, and groups them according to the daily rotation frequency of the lock body. For example, the daily usage frequency of the lock body is divided into three groups: 10 - 20 times, 20 - 40 times, and 40 - 50 times. Each group of data covers the operation records of 30 consecutive days. A long short-term memory neural network is used to construct a wear trend predictor. The network consists of three layers: the input layer receives three feature parameters, namely the wear degree of the lock body, the change rate of the wear degree of the lock body, and the usage cycle of the lock body; the hidden layer contains 128 neurons, and dynamically controls the retention and update of information through the forget gate and the input gate; the output layer generates the predicted values of the lock core gap for the next 7 days. The training data contains 10,000 groups of historical records, of which 80% is used for model training and 20% is used for verification to ensure the generalization ability of the predictor. The embodiments of the present application do not overly limit the grouping method or the network structure, and can be adjusted according to the actual scenario; Quantify the wear acceleration feature by analyzing the wear trend through polynomial fitting; For the trend sequence of the predicted values of the key cylinder clearance for each group, the embodiments of the present application adopt a fourth-order polynomial fitting method to generate a fitting curve reflecting the characteristics of wear acceleration. The polynomial coefficients characterize the variation law of the wear rate. In the slight wear state, the absolute value of the quadratic term coefficient is usually less than 0.01, indicating a stable wear process; in the moderate wear state, the quadratic term coefficient increases to 0.02 to 0.05, showing an accelerating wear rate; in the severe wear state, the quadratic term coefficient exceeds 0.05, reflecting a significant acceleration trend. The quadratic term system calculates the duration distribution of the wear state in the three intervals of slight, moderate, and severe. For example, slight wear usually lasts for more than 90 days, moderate wear lasts for 30 - 60 days, and severe wear lasts for less than 30 days, providing a reference for subsequent prediction; Adopt a sliding time window to predict the wear state and generate a prediction sequence; In the embodiments of the present application, based on the predicted values of the key cylinder clearance output by the wear trend predictor, the system adopts a sliding time window method for short-term wear state prediction. The time window is set to 30 days, the sampling interval is 1 day, and the predictor updates the prediction result once a day to generate a sequence containing timestamps and predicted wear states. The system calculates the prediction error of the key cylinder clearance by using an exponentially weighted method by comparing the deviation between the predicted value of the key cylinder clearance and the actual measured value. The weight coefficient is set to 0.7 to emphasize the contribution of recent data. If the prediction deviation exceeds 10% for 5 consecutive days, the system triggers an online update mechanism and fine-tunes the network weights with a learning rate of 0.01 using the backpropagation algorithm to improve the prediction accuracy; Establish a change curve of the key cylinder clearance and set an early warning mechanism; According to the sequence of the predicted values of the key cylinder clearance, the embodiments of the present application construct a continuous change curve of the key cylinder clearance through cubic spline interpolation. The sampling points include timestamps, predicted values of the key cylinder clearance, and corresponding wear degrees. In the state of a new lock body, the key cylinder clearance is about 0.1 mm; in the slight wear state, the gap increase rate does not exceed 0.01 mm / month; in the moderate wear state, it increases to 0.02 to 0.03 mm / month; in the severe wear state, it can reach 0.05 mm / month. The system sets a gap early warning threshold of 0.3 mm. When the predicted gap value exceeds this threshold, the trigger time point is recorded, and based on the gap data in the recent 30 days, the time to reach the critical gap of 0.5 mm is estimated by using linear extrapolation. The actual measurement shows that the time interval from early warning to the critical state is usually 60 to 90 days, providing an adequate response window for maintenance intervention; In the embodiments of the present application, by constructing a long short-term memory neural network predictor and combining polynomial fitting and sliding window methods, the system can accurately predict the wear state of the intelligent lock body and the change trend of the lock core gap. Compared with the traditional static model, the embodiments of the present application significantly improve the sensitivity and reliability of prediction by dynamically analyzing multi-dimensional features. For example, in high-frequency usage scenarios, the system can identify the trend of accelerated wear in advance through the rapid increase of the quadratic coefficient, thereby triggering the warning mechanism. The embodiments of the present application do not overly limit the time window length or fitting method, which can be optimized and adjusted by technicians according to the characteristics of the lock body and the operating environment. The generated gap change curve provides accurate data support for subsequent threshold adjustment and maintenance decisions, effectively extending the service life of the intelligent lock and reducing the safety risks caused by wear.

[0025] As Figure 1 - Figure 2 shown, in step S104, by combining the change of the lock core gap and the historical door body mating state data, a dynamic mapping relationship between the magnetic induction signal intensity and the gap parameters is constructed, and the magnetic induction threshold range is dynamically optimized to improve the determination accuracy of the intelligent lock.

[0026] Further, in step S104, in the embodiments of the present application, by analyzing the gap change in the lock core gap change curve and the historical door body mating state data, the system constructs a dynamic mapping model between the magnetic induction signal intensity and various gap parameters, and dynamically adjusts the magnetic induction threshold range to ensure the stability and reliability of the intelligent lock in complex environments; Extract and fit the door body mating state data to generate a time series change curve; In the embodiments of the present application, the system extracts the mating state data including the door body deformation amount, the door frame displacement amount, and the door lock mating gap from the historical records, groups them at a sampling frequency of once per hour. Under normal circumstances, the door body deformation amount remains within 0.2 mm, the door frame displacement amount does not exceed 0.3 mm, and the door lock mating gap is between 1.0 and 1.5 mm. These data are fitted using a cubic spline function to generate a continuous curve reflecting the change of the door body mating state over time. This curve can capture the periodic deformation caused by the day-night temperature difference, for example, the deformation amount reaches 0.15 mm during the day when the door body expands and drops to 0.05 mm at night, as well as the slow displacement trend of the door frame due to building settlement, with an average monthly increase of about 0.02 mm. The embodiments of the present application do not strictly limit the sampling frequency or fitting method, which can be adjusted according to the actual scenario; Extract the stability trend of the door body structure through moving average filtering; For the time-series change curve of the door body's mating state, the embodiment of the present application uses a 10-point moving average filter to remove short-term disturbances caused by temperature fluctuations or external vibrations, extracts the long-term change trends of the door body deformation and the door frame displacement, and generates a door body structure stability index from the filtered data, reflecting the deformation characteristics of the door body and the door frame during long-term operation. For example, when the door panel thickness is 40 mm, the door frame is fixed with expansion bolts, and the installation depth of the lock body is 25 mm, the stability index is relatively high, indicating that the influence of structural deformation on the magnetic induction signal is small. If the stability index decreases, it may indicate that the door body structure is loose or abnormally deformed, and it is necessary to further analyze its influence on the magnetic induction signal; Construct a mapping model between the gap and the magnetic induction signal intensity to generate dynamic mating parameters; Based on the door body structure stability index, the embodiment of the present application uses the random forest regression algorithm to establish a mapping function between the lock core gap and the magnetic induction signal intensity. The input features include door body structure parameters, such as the weight coefficient of the door panel thickness of 0.8, the weight coefficient of the door frame fixing strength of 0.9, and the weight coefficient of the lock body installation depth of 1.0. The random forest model is integrated and predicted by 100 decision trees, and outputs the dynamic parameters of the door lock mating, characterizing the non-linear influence of the gap change on the magnetic induction signal. For example, when the deformation of the door body is less than 0.1 mm, the attenuation amplitude of the signal intensity is lower than 5%; when the deformation reaches 0.2 mm, the attenuation increases to 15% - 20%; when it exceeds 0.3 mm, the attenuation significantly intensifies. This model is trained with a large amount of historical data to ensure the accuracy of the mapping relationship; Fit the signal intensity compensation curve and optimize the threshold range; For the dynamic parameter sequence of the door lock mating, the embodiment of the present application performs spectrum analysis to separate the low-frequency drift component and the high-frequency fluctuation component, and uses a second-order polynomial to fit the compensation curve of the magnetic induction signal intensity by the least squares method to generate the reference threshold range of the magnetic induction signal. The system divides the threshold interval into three gears: low, medium, and high, corresponding to 70% - 85%, 85% - 115%, and 115% - 130% of the nominal signal respectively. Calculate the signal fluctuation variance within each interval. Under normal conditions, the variance is lower than 3% of the mean value, while when there is magnetic field interference, the variance can increase to more than 10%. Based on the variance characteristics, the system uses an adaptive weighting method to perform segmented processing on the signal, reducing the influence of the interference interval on the threshold adjustment. The actual measurement shows that the adjusted signal fluctuation amplitude is reduced by about 40%, the false alarm rate is reduced from 12% to 3%, and the recognition accuracy rate is increased to more than 95%; Optimize the signal determination criterion through a deep neural network and Kalman filtering; In an embodiment of the present application, the system extracts the lock core clearance, door lock clearance, and door frame mating clearance from the historical door body mating state data, maps the data to the interval from 0 to 1 using maximum-minimum normalization, and trains the mapping function between the clearance parameters and the magnetic induction signal intensity through a three-layer deep neural network. The network structure includes hidden layers with 128, 64, and 32 nodes, and is trained using the stochastic gradient descent method with a batch size of 64 and a learning rate of 0.001. The training data contains 10,000 sets of records, of which 80% is used for training and 20% is used for verification. Combining the change trend of the lock core clearance output by the wear predictor and the real-time clearance data collected by the door body displacement sensor, the system generates a theoretical change curve of the magnetic induction signal intensity, segments the curve using a 60-minute sliding window, calculates the mean and standard deviation of each segment, sets the measurement noise covariance of the Kalman filter to the square of the standard deviation, and the state noise covariance to one-tenth of it. The upper and lower limits of the fluctuation range of the filtered signal are respectively expanded and reduced by 10% to form a dynamic threshold interval. According to the width of the fluctuation range, the system sets three levels of sampling frequencies: when the fluctuation is less than one-fourth of the interval width, 10-second low-frequency sampling is used; when it is between one-fourth and one-half, 5-second medium-frequency sampling is used; when it is greater than one-half, 2-second high-frequency sampling is used. Finally, combining the door body mating state parameters, the weighted average method is used to calculate the decision threshold, where the weight of the lock core clearance is 0.5, the weight of the door lock clearance is 0.3, and the weight of the door frame mating clearance is 0.2. The threshold stability index for 24 consecutive hours shows that the standard deviation is lower than 3% of the mean, verifying the optimization effect; In an embodiment of the present application, by integrating random forest regression, deep neural network, and Kalman filtering, the system constructs an accurate dynamic mapping model between the magnetic induction signal intensity and the clearance parameters, significantly improving the adaptability and stability of threshold adjustment. Compared with the traditional fixed threshold method, the present application can effectively cope with the influence of door body deformation and environmental interference. For example, when the signal standard deviation increases to 15% due to abnormal door body deformation, the system quickly adjusts the threshold through high-frequency sampling and filtering processing to maintain the decision accuracy. The embodiment of the present application does not strictly limit the neural network structure or filtering parameters, which can be optimized by technicians according to the actual operating environment. The optimized threshold range and sampling strategy provide strong support for the reliable operation of the intelligent lock, significantly reducing the misjudgment risk and extending the service life of the device.

[0027] As Figure 1 - Figure 2 shown, in step S105, according to the lock body wear degree grading index and the predicted value of the lock core clearance, when the adjusted magnetic induction threshold exceeds the preset upper limit, the lock core tolerance range is identified and the target locking decision criterion is generated to ensure the reliable operation of the intelligent lock.

[0028] Further, in step S105, in the embodiments of the present application, when the adjusted magnetic induction threshold exceeds the preset upper limit, the system combines the lock body wear degree and the predicted data of the lock core clearance to dynamically identify the lock core tolerance range, generate a target locking determination criterion, and improve the determination accuracy and stability of the intelligent lock in a complex operating environment; Obtain and smooth the magnetic induction threshold data sequence, and judge the threshold overrun situation; In the embodiments of the present application, the system obtains the adjusted magnetic induction threshold data sequence through the magnetic induction signal acquisition module at a sampling frequency of 10 times per second, records the signal amplitude and timestamp. Since the original data is affected by environmental interference and electronic noise, the fluctuation amplitude may reach 15% of the nominal value. To improve the data quality, the system uses a Kalman filter for smoothing processing, sets the measurement noise variance to one-tenth of the signal variance, such as 0.02, and the state noise variance to one-tenth of it. After filtering, the signal fluctuation amplitude is reduced to within 5%. The system compares the filtered threshold with the preset upper limit, such as 2.5 millitesla. If the threshold continuously exceeds the upper limit, record the overrun time point and trigger the subsequent tolerance adjustment process. The embodiments of the present application do not strictly limit the filtering parameters or the threshold upper limit, and can be optimized according to the actual operating environment; Use a deep neural network to predict the change trend of the lock core tolerance; For the three levels of slight, moderate, and severe in the lock body wear grading index, and the predicted value sequence of the lock core clearance, the embodiments of the present application construct a four-layer deep neural network model, including an input layer, two hidden layers, and an output layer. The input layer receives the wear grading index and the clearance prediction value. The hidden layers respectively contain 64 and 32 neurons, and use the ReLU activation function. The output layer generates the change trend of the lock core tolerance. The network training data set contains 1000 sets of historical records, among which slight wear accounts for 40%, moderate wear accounts for 35%, and severe wear accounts for 25%. The training uses the stochastic gradient descent algorithm with a batch size of 32, and the learning rate is set to 0.001. After 50 rounds of training, the prediction accuracy reaches 90%. For example, in the moderate wear state, the network can predict that the lock core tolerance will increase by 0.1 mm within the next 7 days, providing an accurate basis for threshold adjustment; Calculate the magnetic induction fluctuation compensation amount and generate tolerance compensation parameters; In the embodiment of the present application, the system uses an exponential weighting method to calculate the compensation amount of the magnetic induction signal fluctuation amplitude according to the tolerance increment and the signal strength compensation coefficient, and sets different signal strength compensation coefficients according to the wear degree: when the wear is slight, the signal strength compensation coefficient is 0.8, reducing the unnecessary compensation amplitude; when the wear is moderate, the signal strength compensation coefficient increases to 1.2 to adapt to the accelerating trend of the gap change; when the wear is severe, the signal strength compensation coefficient is 1.5, expanding the tolerance range to ensure the determination stability. The compensated tolerance values are 0.2 mm for slight wear, 0.3 mm for moderate wear, and 0.5 mm for severe wear. The system generates a tolerance compensation parameter sequence, records the compensation effect of each wear interval, and provides data support for the threshold segmentation calculation. By dynamically adjusting the signal strength compensation coefficient, this method effectively balances the requirements of sensitivity and stability; Generate the target determination standard curve through segmented calculation and interpolation; Based on the tolerance compensation parameter sequence, the embodiment of the present application performs segmented calculation on the upper limit threshold of the magnetic induction signal strength, and uses the segmented linear interpolation method to determine the threshold change slope of each interval. For example, the slope in the slight wear interval is 0.01 per hour, increasing to 0.02 per hour in moderate wear, and reaching 0.05 per hour in severe wear. The system calculates the dynamic change of the upper limit threshold according to the slope value, and sets the lower limit threshold to 80% of the upper limit to form a complete threshold interval. The cubic spline interpolation method is used to smooth the threshold interval to generate a continuous target determination standard curve, ensuring a smooth transition of the threshold boundary and avoiding misjudgment caused by sudden changes. The target determination standard includes the upper limit threshold, the lower limit threshold, and the tolerance range, comprehensively characterizing the operating state of the lock body; Verify the stability of the determination standard and trigger automatic calibration; In the embodiment of the present application, the system sets a 1-hour verification time window to continuously monitor the stability of the determination standard curve and calculate the standard deviation of the determination results. If the standard deviation of 5 consecutive determination results exceeds 10% of the mean, it indicates that the determination results are unstable, and the system triggers the automatic calibration process. The calibration process includes re-collecting 100 groups of magnetic induction signal samples, updating the parameters of the deep neural network, and adjusting the tolerance compensation coefficient until the standard deviation drops below 5% of the mean. The measured data shows that the determination accuracy after calibration is improved to over 95%, and the misjudgment rate is reduced to below 3%. Through the dynamic calibration mechanism, the embodiment of the present application ensures the reliability of the determination standard during long-term operation and significantly reduces the misjudgment risk caused by environmental changes or increased wear; In the embodiments of the present application, through the combination of Kalman filtering, deep neural network, and piecewise interpolation methods, the system can accurately identify the tolerance range of the lock core and generate a stable target determination criterion. Compared with the traditional fixed threshold method, the embodiments of the present application significantly improve the adaptability of the intelligent lock under increased wear or environmental interference through dynamic compensation and automatic calibration. For example, in the severely worn state, the system successfully avoids locking failure caused by increased clearance by increasing the tolerance range and slope adjustment. The embodiments of the present application do not strictly limit the network structure or calibration period, which can be optimized by technicians according to the characteristics of the lock body and the usage scenario. The generated target determination criterion provides strong support for the reliable operation of the intelligent lock, effectively extending the equipment life and improving the security.

[0029] In this embodiment, the noun explanation of the magnetic induction threshold is the critical determination range dynamically set based on the magnetic induction signal intensity and waveform characteristics during the operation of the intelligent lock, which is used to judge whether the lock body is in a normal locked state. Its core is to analyze parameters such as the intensity, waveform distortion degree, and attenuation rate of the magnetic induction signal, and combine data such as the lock core clearance, door body mating state, and wear degree to adjust the upper and lower threshold intervals in real time. The threshold range reflects the magnetic field stability and mechanical mating state of the lock body under specific working conditions, and is the core basis for judging the effectiveness of lock control. The triggering conditions in the application process of the magnetic induction threshold are specifically as follows: Dynamically determine the state of the lock body, triggering condition: The intensity of the magnetic induction signal collected in real time exceeds the reference threshold range (±20% of the nominal value). If the signal intensity exceeds the upper limit (2.5 millitesla), trigger the identification of the lock core tolerance range (step S105). If the signal intensity is lower than the lower limit (1.2 millitesla), it is determined that the lock body closing is abnormal, and trigger the adjustment of lock control parameters (step S106). Adapt to lock body wear and environmental interference, triggering condition: The lock body wear degree is classified (slight, moderate, severe) or the door body structure is deformed (door frame displacement > 0.3 mm). Dynamically map the relationship between the clearance and the magnetic induction signal through random forest regression and deep neural network, adjust the threshold interval (step S104), and calculate the threshold slope in segments according to the wear level (severe wear compensation coefficient 1.5) to expand the tolerance range (step S105). Optimize the control strategy, triggering condition: The fluctuation variance of the magnetic induction signal > 10% (normal is 3%) or the misjudgment probability > 5%. If the risk level is "high risk", exponentially adjust the weight factor of the wear parameter and re-optimize the threshold (step S107).

[0030] As Figure 1 - Figure 2 shown, in step S106, continuously monitor whether the intensity of the magnetic induction signal meets the target locking determination criterion, and dynamically adjust the lock control parameters to optimize the operation state of the intelligent lock.

[0031] Further, in step S106, in the embodiment of the present application, the system dynamically optimizes the locking control parameters by analyzing the matching situation between the magnetic induction signal intensity and the target determination criteria in real time, ensuring that the intelligent lock maintains an efficient and stable locking control state under different operating conditions; Collect magnetic induction signals and divide threshold intervals to generate initial optimized values of locking control parameters; In the embodiment of the present application, the system uses a high-speed differential comparator to collect magnetic induction signal intensity data at a sampling frequency of 1000 Hz, records the signal amplitude in real time, and based on the target locking determination criteria, the system divides the signal intensity into three hierarchical intervals: high intensity interval is greater than 2.0 mT, medium intensity interval is 1.2 to 2.0 mT, and low intensity interval is less than 1.2 mT. Each interval generates corresponding initial optimized values of locking control parameters through a mapping function. For example, the high intensity interval corresponds to a motor torque reference value of 0.8 Nm, the medium intensity interval is 0.6 Nm, and the low intensity interval is 0.4 Nm. This hierarchical mapping ensures the dynamic matching of the locking control parameters with the signal intensity and improves the control accuracy. The embodiment of the present application does not strictly limit the interval division or the form of the mapping function, and can be optimized according to the characteristics of the lock body and the application scenario; Combine the lock body operating state data to construct a recurrent neural network model and establish a dynamic mapping between the motor torque and the speed; For the initial optimized values of the locking control parameters, the system combines the lock body operating state data to construct a three-layer recurrent neural network model, including an input layer, a hidden layer with 64 neurons, and an output layer. The network adopts a cyclic structure with a time step of 10 ms to capture the temporal characteristics of the motor torque and speed. The input layer receives the torque reference value and the current speed of the intelligent lock motor driving the lock core or the lock tongue in the operating interval of the intelligent lock, and the output layer outputs the optimized intelligent lock motor control signal, which controls the torque of the lock core or the lock tongue. The training data set contains 1000 sets of historical records, covering the operating data under different load conditions, and is trained using the stochastic gradient descent method with a learning rate set to 0.001. After training, the network can accurately predict the torque adjustment requirements. For example, in the high intensity interval, the torque output deviation is controlled within 5%. This method significantly improves the response speed and stability of the motor control through temporal modeling; Set fuzzy control rules to dynamically adjust the motor speed; Based on the theoretical torque curve output by the recurrent neural network, the embodiments of the present application design seven fuzzy control rules for finely adjusting the motor speed. The rules use the speed deviation and the rate of change of the deviation as input variables, which are respectively divided into five fuzzy subsets: the speed deviation includes large negative, small negative, zero, small positive, and large positive; the rate of change of the deviation includes rapid decrease, slow decrease, stable, slow increase, and rapid increase. The output variable is the speed correction amount, which is also divided into five levels. The fuzzy rules achieve dynamic control through 25 combinatorial logics. For example, when the speed deviation is small positive and the rate of change is slow increase, a moderate positive correction amount is output. The actual measurement shows that the fuzzy control controls the speed fluctuation within 5% of the rated value, significantly improving the smoothness of the lock body operation. The embodiments of the present application do not strictly limit the number of rules or the division of fuzzy subsets, which can be adjusted according to actual needs; Use a Kalman state observer to monitor the lock control state and trigger calibration; In the embodiments of the present application, the system collects the lock tongue position data with a resolution of 0.01 mm through a high-precision position sensor, and records the displacement curve during the unlocking and locking processes. For example, the displacement of the lock tongue from locked to retracted is 20 mm. The system calculates the smoothness index of the displacement curve and evaluates the motion stability through the slope change rate. When the smoothness is less than 0.1, it indicates that the lock tongue moves smoothly. Based on this, the system constructs a Kalman state observer to simultaneously monitor three state variables: motor torque, speed, and lock tongue position. The observer sets the standard deviation of the torque measurement noise to 0.05 N·m, the speed to 10 revolutions per minute, the position to 0.02 mm, and the state noise covariance to one-tenth of the measurement noise. The system uses a 5-minute calibration window to record the standard deviation of the state parameters. If the torque standard deviation exceeds 10% of the mean, the speed exceeds 8%, or the position exceeds 5%, the automatic calibration process is triggered. The calibration includes collecting 100 new samples, updating the neural network parameters and fuzzy rules until the parameters are stable. This method ensures the long-term reliability of the lock control state through multi-variable observation and dynamic calibration; In the embodiments of the present application, through the integration of high-speed signal acquisition, recurrent neural network, fuzzy control, and Kalman state observation, the system realizes the real-time optimization of lock control parameters. Compared with traditional static control methods, the present application can quickly adapt to the changes in magnetic induction signals and the fluctuations in the operating state of the lock body. For example, in the low-intensity interval, the system avoids the lock tongue jamming caused by overload by reducing the torque output and finely adjusting the speed. The embodiments of the present application do not strictly limit the sampling frequency or the observer parameters, which can be optimized by technicians according to the actual scenario. The optimized lock control state data significantly improves the operating efficiency and stability of the intelligent lock, providing a reliable guarantee for the safe operation of power facilities.

[0032] As Figure 1 - Figure 2As shown, in step S107, combining the optimized lock control state data and the change trend sequence, the misjudgment probability is evaluated and it is determined whether the security risk is lower than the preset threshold. If the risk exceeds the standard, the weight factors of the magnetic induction threshold and the wear parameters of the determination tolerance are adjusted, and the lock control parameters are re-optimized to ensure the safe operation of the intelligent lock.

[0033] Further, in step S107, in the embodiment of the present application, the system analyzes the optimized lock control state data and the change trend sequence, quantifies the misjudgment probability and evaluates the security risk, dynamically adjusts the weight factors of the magnetic induction threshold and the wear parameters of the determination tolerance, and optimizes the lock control parameters to improve the security and stability of the intelligent lock; Extract the lock control state data and calculate the misjudgment probability through a deep neural network; In the embodiment of the present application, the system extracts the intelligent lock motor torque curve, the speed change curve of the intelligent lock motor, and the lock tongue position curve of the intelligent lock from the optimized lock control state data, constructs a three-layer deep neural network model, and the input layer receives 9 state parameters, including the intelligent lock motor torque, the speed of the intelligent lock motor, the current of the intelligent lock motor, the lock tongue position of the intelligent lock, the speed of the intelligent lock motor, and the acceleration of the intelligent lock motor, etc.; the two hidden layers respectively contain 128 and 64 neurons, and the ReLU activation function is used; the output layer generates a misjudgment score of the intelligent lock operation state. The network training data set contains 10,000 groups of historical records, of which 2,000 groups are misjudgment samples. The training uses the stochastic gradient descent method, and the learning rate is set to 0.001. After training, the network can accurately identify abnormal states. For example, when the deviation of the lock tongue position exceeds 0.1 mm, the misjudgment probability may increase to 0.1. This method significantly improves the accuracy of misjudgment prediction through multi-dimensional parameter modeling, providing a reliable basis for risk assessment; Use the Bayesian probability method to evaluate the security risk and divide the risk levels; For the misjudgment score of the intelligent lock operation state, the embodiment of the present application uses the Bayesian probability calculation method to evaluate the security risk of the intelligent lock operation state, and divides three security risk level intervals for the low, medium, and high intelligent lock operation states: the misjudgment probability is lower than 0.05 for low risk, 0.05 to 0.15 for medium risk, and higher than 0.15 for high risk. The system determines the probability distribution of each interval through historical data statistics. For example, 95% of the data is in the low risk interval during normal operation, and the proportion of the medium and high risk intervals increases significantly in abnormal states. The Bayesian method combines the prior probability and real-time data to dynamically update the risk assessment result. For example, when the misjudgment probability exceeding 0.1 is detected continuously for 5 times, the system determines it as medium risk and triggers subsequent adjustments. The embodiment of the present application does not strictly limit the risk interval division and can be optimized according to the actual scenario; Set an adaptive weight update rule and dynamically compensate the magnetic induction determination standard; Based on the risk assessment results, the system designs an adaptive weight update rule: among them, the weight factor of the wear parameter is determined according to the above risk level, and the magnetic induction threshold and determination tolerance are dynamically compensated by adjusting the weight factor of the wear parameter. The weight factor of the wear parameter can be understood as a parameter for compensating the magnetic induction threshold and determination tolerance; the weight factor of the wear parameter is a parameter dynamically adjusted according to the security risk assessment results of the intelligent lock. It is used to reflect the sensitivity adjustment requirements of the system under different risk levels.

[0034] In the low-risk interval, the weight factor for adjusting the wear parameter maintains the initial value of 1.0; in the medium-risk interval, the weight factor for adjusting the wear parameter linearly increases by 0.1 for every 0.01 increase in the misjudgment probability; in the high-risk interval, the weight factor for adjusting the wear parameter is adjusted exponentially, with a base of 1.5 and an exponent that is the multiple by which the misjudgment probability exceeds 0.15. Using the updated weight factor for adjusting the wear parameter, the system dynamically compensates the magnetic induction threshold and determination tolerance. Assuming the original magnetic induction threshold is set to 2.0 millitesla (mT), the threshold can be adjusted according to the new weight factor. For example, in the medium-risk case, if the weight factor becomes 1.05, the new magnetic induction threshold can be adjusted to 2.0 * 1.05 = 2.1 mT; in the high-risk case, if the weight factor becomes 1.074, the new magnetic induction threshold is adjusted to 2.0 * 1.074 ≈ 2.148 mT; Similarly, the same weight factor can be applied to adjust the determination tolerance. For example, if the original determination tolerance is ±0.2 mT, then in the medium-risk case, the new determination tolerance can be adjusted to ±0.2 * 1.05 = ±0.21 mT; in the high-risk case, the new determination tolerance is adjusted to ±0.2 * 1.074 ≈ ±0.2148 mT; The magnetic induction threshold compensation coefficient is used to directly adjust the upper and lower threshold ranges of the magnetic induction signal intensity, so as to ensure that the intelligent lock can accurately judge its locked state. The magnetic induction threshold compensation coefficient is positively correlated with the misjudgment probability. For example, when the misjudgment probability is 0.1, the magnetic induction threshold compensation coefficient is 1.2, and the magnetic induction threshold compensation coefficient for the determination tolerance is 1.3; when the probability increases to 0.2, the magnetic induction threshold coefficients increase to 1.4 and 1.6 respectively. After compensation, the recognition sensitivity of the abnormal state is improved. For example, in the high-risk interval, the threshold range is narrowed by 10%, effectively reducing the misjudgment risk; Optimize the lock control parameters through fuzzy control and state feedback; In the embodiments of the present application, the system designs seven fuzzy control rules to update the locking control parameters based on the compensated magnetic induction determination criterion. The rules use the misjudgment probability and the risk level as input variables, and output the parameter adjustment amount, which is divided into five levels: when both are low, the adjustment amount is zero; when either index increases, the adjustment amount increases proportionally; when both are high, the maximum adjustment amount is adopted. The system further constructs a state feedback compensator and uses the proportional-integral control method to adjust the motor torque and speed in real time. The proportional coefficient is set to 0.8, and the integral time constant is 0.5 seconds. During the motor startup stage, the torque compensation amount can reach 20% of the nominal value; during stable operation, the speed compensation amount is controlled within 5%. This method ensures the accuracy and stability of parameter adjustment through fuzzy control and real-time feedback. The embodiments of the present application do not strictly limit the number of rules or control parameters, and can be adjusted according to actual needs; Verify the stability of the locking control state and trigger parameter optimization; For the optimized locking control parameters, the system sets a 1-hour parameter verification period, continuously records the stability index of the locking control state data, and uses the standard deviation to evaluate the degree of fluctuation. If the torque standard deviation exceeds 8% of the mean, the speed exceeds 6%, or the bolt position exceeds 4%, it is determined that the optimization effect is not ideal, and a new round of parameter optimization process is triggered. The optimization process includes re-collecting 100 groups of state data, updating the parameters of the deep neural network, and adjusting the compensation coefficients of the fuzzy control rules and state parameters. The actual measurement shows that after 3 to 4 rounds of optimization, the fluctuation of the state parameters can be controlled within the ideal range. For example, the torque standard deviation is reduced to less than 5% of the mean. This method significantly improves the long-term reliability of the locking control system through periodic verification and dynamic optimization; In the embodiments of the present application, accurate evaluation of the misjudgment probability and dynamic optimization of the locking control parameters are realized. Compared with the traditional static threshold method, the present application can quickly respond to abnormal states. For example, in the high-risk interval, through exponential weight adjustment and strict threshold compensation, the misjudgment rate is reduced to less than 3%. The embodiments of the present application do not strictly limit the network structure or verification period, and can be optimized by technicians according to the characteristics of the lock body and the operating environment. Finally, the optimized locking control parameters and control strategy provide strong support for the safe operation of the intelligent lock, significantly reducing potential safety hazards and extending the service life of the device.

[0035] For those skilled in the art, various corresponding changes and deformations can be made according to the technical solutions and concepts described above, and all these changes and deformations should fall within the protection scope of the claims of the present application.

Claims

1. An intelligent lock control method for power facilities based on multi-dimensional permission data processing, characterized in that Including: After denoising the magnetic induction signal sequence during the operation of the intelligent lock, a denoised signal sequence is generated; Calculate and extract the attenuation parameters of the hyperbolic attenuation model of the signal amplitude in the original magnetic induction signal sequence for the denoised signal sequence, and fit the change trend sequence; Taking the change trend sequence as the input, calculate the difference between the current waveform distortion degree and the historical reference value as the distortion increment and decompose it, construct a support vector regression model, output the wear quantification score of the lock body wear degree, and divide the wear index according to the score; After dividing the wear index, generate a gap change curve according to the judgment result of whether the predicted value of the lock core gap exceeds the preset threshold; Taking the predicted value of the lock core gap as a parameter, establish a database of the door body matching state; and construct a gap-magnetic induction intensity mapping model, output the signal intensity compensation coefficient and perform segmented optimization, and dynamically adjust the interval boundary of the segmented optimization; When the segmented optimization belongs to the high region threshold continuously exceeding the limit, output the tolerance increment of the lock core and door body gap in the next 7 days; calculate the magnetic induction fluctuation compensation amount according to the tolerance increment; generate the target judgment standard curve; Based on the target judgment standard of the target judgment standard curve, divide the magnetic induction signal intensity level, and dynamically adjust the torque output of the intelligent lock motor driving the lock core or the lock tongue; According to the lock control state data of the torque output, construct a deep neural network misjudgment probability model, output the misjudgment score, and dynamically adjust the weight factor of the wear parameter according to the risk level until the torque fluctuation of the intelligent lock motor driving the lock core or the lock tongue meets the preset parameters.

2. The method for controlling an intelligent lock of a power facility based on multi-dimensional permission data processing according to claim 1, wherein After obtaining the magnetic induction signal sequence during the operation of the intelligent lock and denoising it, a denoised signal sequence is generated, including; Using a Butterworth digital filter to perform third-order high-pass filtering on the sequence with a cut-off frequency of 5 Hz and second-order low-pass filtering with a cut-off frequency of 500 Hz to generate a denoised signal sequence.

3. The intelligent lock control method for power facilities based on multi-dimensional permission data processing according to claim 1, characterized in that, Taking the change trend sequence as the input, calculate the difference between the current waveform distortion degree and the historical reference value as the distortion increment and decompose it, construct a support vector regression model, output the wear quantification score of the lock body wear degree, and divide the wear index according to the score, including: Calculate the amplitude ratio of the fundamental wave to the third harmonic as the waveform distortion degree, calculate the difference between the current waveform distortion degree and the historical reference value as the distortion increment, and fit the signal intensity attenuation rate in combination with the lock core rotation angle data; Perform four-layer db4 wavelet decomposition on the distortion increment and the attenuation rate, and extract the energy distribution characteristics in the 0-32 Hz frequency band; Input 8-dimensional wavelet features and historical wear data, and through a support vector regression model, output a wear quantification score in the 0-1 interval, and divide the wear index into three levels: slight 0-0.3, moderate 0.3-0.6, and severe 0.6-1.0 according to the score.

4. The method for controlling an intelligent lock of a power facility based on multi-dimensional permission data processing according to claim 1, characterized in that After dividing the wear index, generating a gap change curve according to the judgment result of whether the predicted value of the lock core gap exceeds the preset threshold includes the following operations: Using a long short-term memory neural network to construct a prediction model, the input layer receives the wear quantification score, the change rate of the lock body wear degree, and the daily average usage frequency of the lock body, and outputs the predicted value of the lock core gap in the next 7 days; When the predicted value of the lock core gap exceeds the predicted threshold of the lock core gap, a maintenance warning is triggered and a lock core gap change curve including a timestamp is generated.

5. The intelligent lock control method for power facilities based on multi-dimensional permission data processing according to claim 1, characterized in that, The predicted value of the lock core gap is a core parameter, and the following operations are performed: A gap-magnetic induction intensity mapping model is constructed using the random forest regression algorithm. The weight of the door body parameters is input, and the signal intensity compensation coefficient is output. The reference threshold range of the magnetic induction signal is segmented and optimized based on the compensation coefficient: the nominal value in the low region is 70%-85%, the middle region is 85%-115%, and the high region is 115%-130%. The interval boundary is dynamically adjusted through a Kalman filter.

6. The intelligent lock control method for power facilities based on multi-dimensional permission data processing according to claim 1, wherein When the threshold in the high region is continuously exceeded, the following operations are performed: A deep neural network is used to predict the tolerance trend of the lock core. The wear level and the predicted value of the gap are input, and the tolerance increment of the gap between the lock core and the door body in the next 7 days is output. The magnetic induction fluctuation compensation amount is calculated based on the tolerance increment and the signal intensity compensation coefficient. The signal intensity compensation coefficient is set differently according to the wear level: slightly, moderately, and severely. The target determination standard curve is generated by cubic spline interpolation.

7. The method for controlling and managing an intelligent lock of a power facility based on multi-dimensional permission data processing according to claim 1, wherein, Based on the target determination standard of the target determination standard curve, the magnetic induction signal intensity level is divided, and the torque output of the intelligent lock motor driving the lock core or the lock tongue is dynamically adjusted, including the following operations: The magnetic induction signal intensity is collected in real time, and the signal intensity level is divided into three regions: high > 2.0 mT, medium 1.2 - 2.0 mT, and low < 1.2 mT. Combined with the lock body operation state data, a recurrent neural network model is constructed. The torque reference value and the rotational speed of the intelligent lock motor driving the lock core or the lock tongue in the operation interval of the intelligent lock are input, and the optimized intelligent lock motor control signal is output. A fuzzy controller is used with the rotational speed deviation and the deviation change rate as input variables to dynamically adjust the torque output.

8. The method for controlling and managing an intelligent lock of a power facility based on multi-dimensional permission data processing according to claim 1, wherein, Based on the lock control state data of the torque output, a deep neural network misjudgment probability model is constructed, and the misjudgment score is output. The weight factor for adjusting the wear parameters is dynamically adjusted according to the risk level until the torque fluctuation of the intelligent lock motor driving the lock core or the lock tongue meets the preset parameters, including the following operations: A deep neural network misjudgment probability model is constructed. The torque of the intelligent lock motor, the rotational speed of the intelligent lock motor, and the position parameter of the intelligent lock tongue are input, and the misjudgment score of the operation state of the intelligent lock in the 0-1 interval is output. The Bayesian method is used to divide the safety risk level of the operation state of the intelligent lock. The weight factor for adjusting the wear parameters is dynamically adjusted according to the risk level: maintained at 1.0 for low risk, linearly increasing for medium risk, and exponentially compensated with a base of 1.5 for high risk. Trigger the parameter re-optimization process until the torque fluctuation of the intelligent lock motor driving the lock core or the lock tongue meets the preset parameters.

Images