Method capable of enhancing anti-interference capability of intelligent lock

Abstract

The invention relates to the technical field of intelligent lock safety protection, in particular to a method capable of enhancing the anti-interference capability of an intelligent lock, and the method comprises the steps: collecting electromagnetic induction signals around the intelligent lock in real time, and intercepting a continuous discrete time sequence as a to-be-analyzed sample sequence when the amplitude of the electromagnetic induction signals exceeds a preset ground noise threshold value; calculating a signal transient impact degree index of the to-be-analyzed sample sequence, determining an adaptive penalty factor of variational mode decomposition based on the signal transient impact degree index, and performing variational mode decomposition on the to-be-analyzed sample sequence by using the adaptive penalty factor, obtaining a preset number of intrinsic mode components and center frequencies corresponding to the intrinsic mode components; by means of the technical scheme, high-frequency attack feature amplification and low-frequency interference feature suppression can be achieved, accurate distinguishing of electromagnetic attack signals and electrostatic interference signals from physical underlying features is achieved, and the problem of false alarm or crash of the intelligent lock caused by static electricity is effectively solved.

Description

A method to enhance the anti-interference capability of smart locks Technical Field

[0001] This application relates to the field of smart lock security technology, and in particular to a method for enhancing the anti-interference capability of smart locks. Background Technology

[0002] With the rapid development of the smart home industry, smart door locks, as core terminal devices for home security, are directly related to the safety of users' property and residence, and have become an indispensable part of modern home life. However, the core components of smart locks, such as electronic control units and communication modules, are sensitive to the external electromagnetic environment and face increasingly severe threats of malicious electromagnetic interference. Existing technologies, such as high-intensity electromagnetic pulse generators known as "black boxes" and Tesla coils, can generate high-frequency, high-voltage electromagnetic pulses that can directly intrude into the electronic circuits inside smart locks, interfering with the normal operating timing of chips and damaging the stability of communication protocols. This may cause the main control MCU of the smart lock to restart or the program to crash, and in extreme cases, it may cause the door lock to misoperate and lock, posing a serious threat to user safety.

[0003] To address the aforementioned electromagnetic interference threats, existing technologies typically employ two methods: hardware shielding and software monitoring. Hardware shielding usually involves installing a metal shield or applying shielding paint inside the smart lock to create a physical isolation structure, blocking external electromagnetic fields from radiating to the core circuitry. However, this method significantly increases the material cost and manufacturing complexity of the smart lock. Software monitoring primarily relies on a watchdog circuit to monitor the program's running status or a voltage sampling module to monitor changes in the smart lock's power supply voltage. When a program crash or abnormal voltage drop is detected, the main control MCU is triggered to reset and restore normal operation. However, in actual use, static electricity carried by the human body in winter can generate instantaneous high-voltage pulses, producing instantaneous high-voltage signals similar to malicious electromagnetic attacks. Because the amplitude and instantaneous voltage change trends of such unintentional interference are similar to those of malicious electromagnetic attacks, and existing monitoring methods struggle to distinguish between this unintentional instantaneous static interference and malicious continuous high-frequency electromagnetic attacks based on waveform characteristics, smart locks frequently trigger false alarms or malfunction due to static electricity generated by user touches, severely impacting the user experience.

[0004] However, the frequency of electromagnetic waves generated by malicious electromagnetic attack devices is not a fixed value. Affected by factors such as the stability of the power supply voltage and the number of turns of the coil winding, its frequency fluctuation range can cover a wide frequency band from tens of kilohertz to hundreds of megahertz. The filtering algorithms used in existing software protection methods are all designed based on fixed cutoff frequency parameters, which are difficult to adapt to such dynamically changing attack frequencies. This results in a low accuracy rate in identifying wide-range frequency-converting electromagnetic attacks, and it is easy to miss or misjudge cases. Summary of the Invention

[0005] To address the problems of existing technologies that struggle to distinguish between instantaneous electrostatic interference and continuous electromagnetic attacks, and the inadequacy of fixed-parameter filtering for frequency-dependent electromagnetic attacks, this application provides a method to enhance the anti-interference capability of smart locks.

[0006] This application provides a method to enhance the anti-interference capability of smart locks, comprising: real-time acquisition of electromagnetic induction signals around the smart lock; when the amplitude of the electromagnetic induction signal exceeds a preset noise floor threshold, extracting a continuous discrete-time sequence as a sample sequence to be analyzed; calculating the transient impact index of the sample sequence to be analyzed; determining an adaptive penalty factor for variational mode decomposition based on the transient impact index; performing variational mode decomposition on the sample sequence to be analyzed using the adaptive penalty factor to obtain a preset number of intrinsic mode components and their corresponding center frequencies; calculating the energy operator value of each intrinsic mode component; weighting the energy operator value in combination with its corresponding center frequency to calculate the current weighted attack strength index; acquiring the current real-time battery voltage of the smart lock; updating the current cumulative damage value according to the weighted attack strength index and the drop in real-time battery voltage relative to the standard voltage; and performing a locking protection operation on the smart lock if the cumulative damage value exceeds a preset locking threshold of the smart lock.

[0007] This application addresses the technical pain points of electromagnetic interference protection for smart locks by constructing a multi-dimensional collaborative protection system. It establishes a real-time ring buffer and a noise floor threshold triggering mechanism to continuously cache and accurately determine electromagnetic induction signals, sensitively capturing various sudden interferences and achieving timely response and acquisition of interference signals. Simultaneously, based on the signal statistics transient impact index of the sample sequence to be analyzed, it adaptively adjusts the variational mode decomposition bandwidth. By adapting the optimal adaptive penalty factor, it achieves optimal feature extraction results for both broadband electrostatic interference and narrowband malicious electromagnetic attacks, ensuring feature recognition accuracy. Furthermore, it constructs a coupled judgment mechanism based on signal energy and real-time battery voltage drop amplitude, which can accurately identify malicious attack characteristics and monitor the power supply operating limits of the smart lock in real time, forming a dual protection logic. This significantly improves the accuracy of anti-interference identification and the safety of use of smart locks, building a solid protective barrier.

[0008] In one embodiment, calculating the transient impact index of the sample sequence to be analyzed specifically includes: first, obtaining the fourth central moment of the sample sequence to be analyzed; then, calculating the variance of the sample sequence to be analyzed and taking the square of the variance; dividing the fourth central moment by the square of the variance of the sample sequence to be analyzed to obtain the transient impact index of the signal. A very small constant is added when calculating the variance of the sample sequence to be analyzed to prevent division by zero error.

[0009] In one embodiment, the adaptive penalty factor for variational mode decomposition is determined based on the signal transient impulse index, satisfying the following relationship: ;in, Indicates the adaptive penalty factor. Indicates the base bandwidth. This represents the adjustment range constant. Indicates the transient impact index of a signal. The threshold for distinguishing morphological features This indicates the sensitivity index.

[0010] This relationship enables intelligent adaptation to the environment based on the transient impact index of the electromagnetic induction signal. When a broadband electrostatic interference signal with a large transient impact is detected, the adaptive penalty factor of the variational mode decomposition is automatically lowered to relax the bandwidth limit and avoid misjudging oscillating signals of electrostatic interference. When a narrowband malicious attack signal with a small transient impact is detected, the adaptive penalty factor is automatically raised to tighten the bandwidth constraint and accurately lock the characteristic frequency of the attack signal. From the perspective of the underlying physical characteristics of the signal, this effectively solves the technical pain point of false alarms of electrostatic interference.

[0011] In one embodiment, the adaptive penalty factor is used to perform variational mode decomposition on the sample sequence to be analyzed. Specifically, the adaptive penalty factor is used as a bandwidth constraint parameter in the variational mode decomposition algorithm. The sum of the bandwidth estimates of each intrinsic mode component is minimized through iterative solution, thereby separating the sample sequence to be analyzed into a preset number of intrinsic mode components.

[0012] In one embodiment, the energy operator value of each intrinsic mode component is calculated separately, specifically by subtracting the product of the values ​​of the intrinsic mode component at the previous time and the next time from the square of the value of the intrinsic mode component at the current time, to obtain the energy operator value of the intrinsic mode component at the current time.

[0013] In one embodiment, the weighted attack strength index satisfies the following relationship: ;in, This represents the weighted attack strength index. This represents the total number of intrinsic modal components. Indicates the first The center frequency of each intrinsic modal component Indicates the length of the sample sequence to be analyzed. Indicates the first Each intrinsic mode component in Energy operator value at time t. This indicates taking the absolute value.

[0014] By introducing a logarithmic frequency weighting term, a feature enhancement mechanism targeting high-frequency attack signals is constructed, enabling the smart lock's main control MCU to selectively perceive the frequency of high-frequency attack signals. This feature enhancement mechanism can significantly amplify the characteristic values ​​of the megahertz-level high-frequency signals unique to the black box, while effectively suppressing the feature weights of low-frequency environmental interference signals. This greatly improves the signal-to-noise ratio between malicious attack signals and background interference, providing highly recognizable feature support for the accurate determination of subsequent attack signals and ensuring the accuracy and reliability of attack identification.

[0015] In one embodiment, the cumulative damage value satisfies the following relationship: ;in, This represents the cumulative damage value at the current moment. This represents the cumulative damage value at the previous moment. Indicates the forgetting factor, This represents the weighted attack strength index at the current moment. Indicates standard voltage. This indicates the current real-time battery voltage of the smart lock. This represents the voltage coupling sensitivity coefficient.

[0016] By using the real-time battery voltage drop of the smart lock as an acceleration factor for attack judgment, the accumulated damage value rises rapidly under continuous malicious attacks, while the accumulated damage value drops rapidly due to the decay effect of the forgetting factor in occasional electrostatic interference, thus avoiding ineffective accumulation.

[0017] In one embodiment, after extracting a continuous discrete-time sequence as the sample sequence to be analyzed, the mean of the sample sequence to be analyzed is calculated, and the mean is subtracted from each data point in the sample sequence to be analyzed, thus completing the DC component removal preprocessing.

[0018] In one embodiment, the execution of the smart lock's locking protection operation specifically includes: when the updated cumulative damage value exceeds the smart lock's preset locking threshold, the smart lock's main control MCU automatically cuts off the power supply to the smart lock's internal motor drive circuit, writes the current attack event into the log, and the main control MCU controls the smart lock to enter the locking protection state.

[0019] In one embodiment, the real-time acquisition of electromagnetic induction signals around the smart lock specifically includes: using a radio frequency receiving module or a ring monitoring coil installed on the circuit board inside the front panel of the smart lock, and utilizing the analog-to-digital converter interface of the microcontroller, continuously sampling the voltage signal sensed by the radio frequency receiving module at a preset sampling rate to obtain the electromagnetic induction signal.

[0020] The technical solution of this application has the following beneficial technical effects: This application can adaptively adjust the variational mode decomposition penalty factor through the signal transient impact index, adapt to the optimal feature extraction of different signals, and combine the logarithmic frequency weighting term to amplify high-frequency attack features and suppress low-frequency interference, so as to achieve accurate low-level differentiation between electromagnetic attack signals and electrostatic interference signals, effectively solve the pain point of smart locks false alarms or crashes due to external static electricity, and greatly improve the accuracy of attack identification and signal-to-noise ratio.

[0021] Furthermore, the battery voltage drop amplitude is introduced as a decision acceleration factor. A dynamic update mechanism for the cumulative damage value is constructed by combining the leaky bucket principle and the forgetting factor. This allows the cumulative damage value to accumulate rapidly under continuous malicious attacks, while the cumulative damage value of occasional electrostatic interference decays rapidly. At the same time, when the attack intensity threatens the stability of the smart lock's power supply, the accumulation of the cumulative damage value is accelerated exponentially, ensuring that the main control MCU completes the locking protection before the smart lock's power supply voltage is too low and causes it to crash and fail. Attached Figure Description

[0022] Figure 1 is a flowchart of a method for enhancing the anti-interference capability of a smart lock according to an embodiment of this application; Figure 2 shows the distribution of adaptive parameter adjustment based on signal features; Figure 3 shows the distribution of fixed threshold decision in the prior art; Figure 4 shows the distribution of multi-dimensional feature coupling decision in this invention. Detailed Implementation

[0023] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0024] As shown in Figure 1, a method to enhance the anti-interference capability of a smart lock includes steps S101 to S104, which are described in detail below.

[0025] S101 collects electromagnetic induction signals around the smart lock in real time. When the amplitude of the electromagnetic induction signal exceeds the preset noise floor threshold, a continuous discrete time series is extracted as the sample sequence to be analyzed.

[0026] In one embodiment, a radio frequency (RF) receiver module is first pre-installed on a circuit board inside the front panel of the smart lock. The RF receiver module can be a multiplexed NFC / RFID antenna or a specially designed PCB loop monitoring coil. The RF receiver module is then connected to the analog-to-digital converter (ADC) input pin of the smart lock's main control MCU. The main control MCU continuously samples the voltage signal sensed by the RF receiver module at a preset sampling rate through the ADC and converts it into an electromagnetic induction signal. For example, the preset sampling rate can be set to 100kHz to ensure that a wide frequency range of electromagnetic induction signals can be captured completely.

[0027] Furthermore, the main control MCU maintains a length of [missing information] in memory. The circular buffer, specifically, The value can be set to 1024, which allows the continuous data sampled by the main control MCU to be written to the circular buffer in real time. Newly written data overwrites the oldest stored data according to the first-in-first-out principle, realizing real-time data updating and caching, while preset noise floor threshold. The preset noise floor threshold can be set to 0.5. When the amplitude of the electromagnetic induction signal detected by the main control MCU is greater than the preset noise floor threshold, it determines that an environmental anomaly has occurred and immediately locks the data in the current circular buffer, treating it as a continuous discrete time series, thereby obtaining the sample sequence to be analyzed. Then, the mean of the sample sequence to be analyzed is calculated, and the corresponding mean is subtracted from each data point in the sample sequence to complete the DC component removal process, so as to eliminate the influence of DC interference on the subsequent electromagnetic induction signal decomposition and feature analysis.

[0028] Thus, by constructing a real-time ring buffer and a noise floor threshold triggering mechanism, the main control MCU can sensitively capture sudden electromagnetic interference events and obtain complete electromagnetic induction signal waveform data, providing an accurate data foundation for subsequent refined analysis.

[0029] S102, calculate the transient impact index of the signal of the sample sequence to be analyzed, determine the adaptive penalty factor of variational mode decomposition based on the transient impact index, and decompose the sample sequence to be analyzed using the adaptive penalty factor to obtain a preset number of intrinsic mode components and their corresponding center frequencies.

[0030] In one embodiment, to distinguish between broadband pulses caused by electrostatic discharge and narrowband oscillations caused by an attack, the sequence of the sample to be analyzed is calculated. Transient Impulse Index of Signal The following relation is satisfied: Where N is the length of the sample sequence to be analyzed. Let i be the i-th data point in the sample sequence to be analyzed. The mean value of the sample sequence to be analyzed after DC component removal preprocessing. It is a minimal constant, and can take a minimum value. This is used to prevent division by zero errors when the variance of the sample sequence to be analyzed is 0 due to the sample sequence being completely flat.

[0031] For example, a segment of the sample sequence to be analyzed after DC component removal preprocessing is extracted. ; Calculate the mean of the sample sequence to be analyzed. Subtract the mean from each data point in the sample sequence to obtain the mean-free sample sequence. ; Calculate the variance of the sample sequence to be analyzed. Calculate the fourth central moments of the sample sequence to be analyzed. Substitute the values ​​into the relational expression to calculate the transient impact index of the signal: The transient impact index of this signal can reflect the shape of the electromagnetic induction signal. If the electromagnetic induction signal is an extremely short pulse electrostatic interference, the transient impact index value will be significantly larger, reaching more than 20; if the electromagnetic induction signal is a narrowband oscillating electromagnetic attack, the transient impact index value will be relatively smaller.

[0032] Subsequently, the adaptive penalty factor is calculated. : Set the base bandwidth. Its adjustment range constant is 500. The morphological differentiation threshold is 2000. The sensitivity index is 3. Substituting the value into the above formula, the calculated transient impulse index of the signal is 2.5. The denominator term... The entire denominator fractional terms Finally, the adaptive penalty factor is calculated. The answer is 500 + 1183 = 1683.

[0033] At this point, the adaptive penalty factor is relatively large, while the transient impact index of the signal is less than 3, indicating that the algorithm judges that the electromagnetic induction signal is biased towards narrowband oscillation. Therefore, the bandwidth is tightened to extract the characteristics of the attack signal. If it is an electrostatic signal, its transient impact index is usually as high as 20 or more, which makes the denominator in the relational expression increase sharply, making the adaptive penalty factor close to 500, thereby relaxing the bandwidth limit and adapting to the broadband characteristics of the electrostatic signal.

[0034] Finally, the adaptive penalty factor is used as the bandwidth constraint parameter in the variational mode decomposition algorithm. Through iterative solution, the sum of the estimated bandwidths of each intrinsic mode component is minimized, thereby separating the sample sequence to be analyzed into a preset number of intrinsic mode components, and simultaneously outputting the center frequency corresponding to each intrinsic mode component. .

[0035] Thus, by using an adaptive penalty factor, the most suitable variational mode decomposition strategy can be adopted for different forms of electromagnetic induction signals, effectively avoiding the distortion of electromagnetic induction signal feature extraction caused by fixed parameters, ensuring that each intrinsic mode component has a single, concentrated frequency characteristic, and greatly improving the accuracy and stability of electromagnetic induction signal identification.

[0036] S103, calculate the energy operator value of each intrinsic mode component, and weight the energy operator value with its corresponding center frequency to calculate the current weighted attack strength index.

[0037] In one embodiment, the weighted attack strength index satisfies the following relationship: Among them, the energy operator .

[0038] For example, a major high-frequency intrinsic mode component was decomposed using the variational mode decomposition algorithm, and its center frequency was... The time-domain average energy operator value for this intrinsic mode component is calculated to be 4, given a Hz frequency of 1,000,000. Therefore, the center frequency weighting term... The square root of the energy term of this intrinsic mode component yields 2. The weighted attack strength exponent of this intrinsic mode component is calculated as follows: .

[0039] In contrast, if it is a common, non-malicious interference signal in the environment, its center frequency... If the frequency is 50Hz and the energy operator value is also set to 4, then the center frequency weighting term... The weighted attack strength exponent of this intrinsic mode component is calculated as follows: .

[0040] It can be seen that, under the premise that the time-domain average energy operator values ​​of the intrinsic mode components corresponding to the high-frequency attack signal and the intrinsic mode components corresponding to the power frequency interference are the same, the weighted attack intensity index of the intrinsic mode components corresponding to the high-frequency attack signal is amplified by more than 3 times compared to the contribution intensity of the intrinsic mode components corresponding to the power frequency interference.

[0041] Thus, by combining the center frequency weighting term with the energy operator value, an evaluation system that is extremely sensitive to high-frequency modulated signals is constructed. This system can effectively highlight potential black box electromagnetic attack signals from ordinary background noise, and achieve accurate differentiation between high-frequency attack signals and ordinary background interference.

[0042] S104: Obtain the current real-time battery voltage of the smart lock, update the current cumulative damage value based on the weighted attack strength index and the drop in real-time battery voltage relative to the standard voltage, and if the cumulative damage value exceeds the smart lock's preset locking threshold, execute the smart lock's locking protection operation.

[0043] In one embodiment, the cumulative damage value is calculated. The following relation is satisfied: In the formula, This represents the cumulative damage value at the current moment. This represents the cumulative damage value at the previous moment. Indicates the forgetting factor, This represents the weighted attack strength index at the current moment. Indicates standard voltage. This indicates the current real-time battery voltage of the smart lock. This represents the voltage coupling sensitivity coefficient.

[0044] Example: Cumulative damage value at the previous time step The forgetting factor is 10. The value is 0.9, which calculates the current weighted attack strength index. The voltage sensitivity coefficient is 27.6. 10, standard voltage It is 6V.

[0045] Scenario 1: The attack did not affect the smart lock's power supply. The real-time battery voltage of the smart lock's power supply is consistent with the standard voltage of 6V. In this case, the drop in real-time battery voltage relative to the standard voltage is: In the relation = The updated .

[0046] The second scenario: A severe attack causes the real-time battery voltage of the smart lock's power supply to drop. V, the drop in real-time battery voltage relative to the standard voltage is In the relation = The updated .

[0047] Therefore, it can be seen that when the real-time battery voltage of the smart lock power supply drops, the cumulative damage value accumulates more than twice as fast as under normal circumstances, enabling the smart lock to reach its preset locking threshold more quickly. For example, the preset locking threshold of the smart lock is 100. When the cumulative damage value at the current moment is greater than 100, the main control MCU automatically cuts off the power supply to the internal motor drive circuit of the smart lock and writes the current attack event into the log. The main control MCU then controls the smart lock to enter the locking protection state.

[0048] In this way, by linking the weighted attack strength index with the real-time battery voltage, when the main control MCU detects a dangerous signal of abnormal real-time battery voltage of the smart lock power supply, it can quickly increase the cumulative damage value growth rate, ensuring that the smart lock completes the safe locking action in priority before the main control MCU or actuator fails due to low real-time battery voltage, thus building the last line of defense for the smart lock against malicious electromagnetic attacks.

[0049] Referring to Figure 2, the distribution of adaptive parameter adjustment based on signal characteristics is shown, where the horizontal axis represents the transient impact of the signal and the vertical axis represents the adaptive penalty factor. As can be seen from the figure, electromagnetic attack signals are concentrated in the high-frequency oscillation feature region, corresponding to a higher adaptive penalty factor value, while electrostatic interference signals are concentrated in the transient pulse feature region, corresponding to a lower adaptive penalty factor value. This indicates that the present invention can automatically adjust the adaptive penalty factor value according to the difference in transient impact of electromagnetic induction signals. A high penalty factor is adapted for narrowband oscillating electromagnetic attack signals to tighten the decomposition bandwidth, while a low penalty factor is adapted for wideband pulse electrostatic interference signals to relax the bandwidth constraint. This achieves accurate adaptation and decomposition of signals of different forms, providing reliable support for subsequent signal feature extraction and interference type differentiation.

[0050] Referring to Figure 3, the distribution of fixed threshold judgments in the prior art is shown. As can be seen from the figure, the distribution of electromagnetic attack samples and electrostatic interference samples in the signal energy intensity dimension highly overlaps, and the energy intensity of both electromagnetic attack samples and electrostatic interference samples is significantly higher than the fixed judgment threshold set by the prior art. A large number of harmless electrostatic interference samples fall into the false triggering alarm area. The dashed line represents the single energy fixed judgment threshold used in the prior art, which uses only signal energy intensity as the sole judgment criterion. This figure intuitively demonstrates that the prior art, due to the judgment logic of using a single feature fixed threshold, cannot effectively distinguish between high-energy harmless electrostatic interference and high-energy harmful electromagnetic attacks, which easily leads to a large number of false triggering problems, interferes with the normal use of smart locks, and has significant technical defects.

[0051] Referring to Figure 4, the distribution of the multi-dimensional feature coupling decision of the present invention is shown. As can be seen in the figure, electromagnetic attack samples are concentrated in the high-risk locking response area above the dynamic decision boundary, which corresponds to the need to trigger protection operations. Electrostatic interference samples are all concentrated in the safety filtering area below the dynamic decision boundary, which corresponds to the need not to trigger protection. Electromagnetic attack signals and electrostatic interference signals are accurately distinguished. It can accurately identify harmful electromagnetic attacks to trigger safety protection, and effectively filter harmless electrostatic interference. It effectively solves the pain point of smart locks being falsely triggered due to external static electricity in the prior art, and greatly improves the accuracy, reliability and practicality of smart lock anti-interference protection.

[0052] It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the scope of protection of this application. Therefore, the scope of protection of this patent application shall be determined by the appended claims.

Claims

1. A method for enhancing the anti-interference capability of smart locks, characterized in that, include: The electromagnetic induction signal around the smart lock is collected in real time. When the amplitude of the electromagnetic induction signal exceeds the preset noise floor threshold, a continuous discrete time sequence is extracted as the sample sequence to be analyzed. The transient impact index of the signal of the sample sequence to be analyzed is calculated. Based on the transient impact index, an adaptive penalty factor for variational mode decomposition is determined. The adaptive penalty factor is used to perform variational mode decomposition on the sample sequence to be analyzed to obtain a preset number of intrinsic mode components and their corresponding center frequencies. The energy operator value of each intrinsic mode component is calculated respectively. The energy operator value is weighted in combination with its corresponding center frequency to calculate the current weighted attack strength index. The current real-time battery voltage of the smart lock is obtained. Based on the weighted attack strength index and the drop in real-time battery voltage relative to the standard voltage, the current cumulative damage value is updated. If the cumulative damage value exceeds the smart lock's preset locking threshold, the smart lock's locking protection operation is executed.

2. The method for enhancing the anti-interference capability of a smart lock according to claim 1, characterized in that, The calculation of the transient impact index of the sample sequence to be analyzed specifically includes: first, obtaining the fourth central moment of the sample sequence to be analyzed, then calculating the variance of the sample sequence to be analyzed, and taking the square of the variance. The fourth central moment is divided by the square of the variance of the sample sequence to be analyzed to obtain the transient impact index of the signal. A very small constant is added when calculating the variance of the sample sequence to be analyzed to prevent division by zero error.

3. The method for enhancing the anti-interference capability of a smart lock according to claim 1, characterized in that, The adaptive penalty factor satisfies the following relationship: ;in, Indicates the adaptive penalty factor. Indicates the base bandwidth. This represents the adjustment range constant. Indicates the transient impact index of a signal. The threshold for distinguishing morphological features This indicates the sensitivity index.

4. The method for enhancing the anti-interference capability of a smart lock according to claim 1, characterized in that, The variational mode decomposition of the sample sequence to be analyzed is performed using the adaptive penalty factor, specifically including: using the adaptive penalty factor as the bandwidth constraint parameter in the variational mode decomposition algorithm, and solving iteratively to minimize the sum of the bandwidth estimates of each intrinsic mode component, thereby separating the sample sequence to be analyzed into a preset number of intrinsic mode components.

5. The method for enhancing the anti-interference capability of a smart lock according to claim 1, characterized in that, The energy operator value of each intrinsic mode component is calculated separately. Specifically, the energy operator value of each intrinsic mode component at the current time is obtained by subtracting the product of the values ​​of the previous time and the next time from the square of the value of the intrinsic mode component at the current time.

6. The method for enhancing the anti-interference capability of a smart lock according to claim 1, characterized in that, The weighted attack strength index satisfies the following relationship: ;in, This represents the weighted attack strength index. This represents the total number of intrinsic modal components. Indicates the first The center frequency of each intrinsic modal component Indicates the length of the sample sequence to be analyzed. Indicates the first Each intrinsic mode component in Energy operator value at time t. This represents the value of the k-th intrinsic mode component at time t. This indicates taking the absolute value.

7. The method for enhancing the anti-interference capability of a smart lock according to claim 1, characterized in that, The cumulative damage value satisfies the following relationship: ;in, This displays the cumulative damage value at the current moment. This represents the cumulative damage value at the previous moment. Indicates the forgetting factor, This represents the weighted attack strength index at the current moment. Indicates standard voltage. This indicates the current real-time battery voltage of the smart lock. This represents the voltage coupling sensitivity coefficient.

8. A method for enhancing the anti-interference capability of a smart lock according to claim 1, characterized in that, After extracting a continuous discrete-time sequence as the sample sequence to be analyzed, the mean of the sample sequence to be analyzed is calculated, and the mean is subtracted from each data point in the sample sequence to be analyzed, thus completing the DC component removal preprocessing.

9. A method for enhancing the anti-interference capability of a smart lock according to claim 1, characterized in that, The specific steps of performing the smart lock's locking protection operation include: when the updated cumulative damage value exceeds the smart lock's preset locking threshold, the smart lock's main control MCU automatically cuts off the power supply to the smart lock's internal motor drive circuit, writes the current attack event into the log, and the main control MCU controls the smart lock to enter the locking protection state.

10. A method for enhancing the anti-interference capability of a smart lock according to claim 1, characterized in that, The real-time acquisition of electromagnetic induction signals around the smart lock specifically includes: using a radio frequency receiving module or a ring monitoring coil installed on the circuit board inside the front panel of the smart lock, and utilizing the analog-to-digital converter interface of the microcontroller, continuously sampling the voltage signal sensed by the radio frequency receiving module at a preset sampling rate to obtain the electromagnetic induction signal.

Images