DP interface connection state detection method and system and computer equipment

Abstract

The invention relates to the technical field of interface detection, and discloses a DP interface connection state detection method and system and computer equipment, and the method comprises the steps: collecting a historical level signal of an HPD pin in a DP interface, and carrying out the preprocessing of the historical level signal, and obtaining a level preprocessing signal and level fluctuation characteristic data; executing time domain decomposition processing to obtain a high-low level stable region identification result and stable region duration time data; calculating a high level threshold and a low level threshold to obtain a level double-threshold detection parameter; and performing state detection on the real-time level signal of the HPD pin based on the level double-threshold detection parameter to obtain a target connection state judgment result and a state judgment reliability value, thereby improving stability and anti-interference capability of DP interface connection state detection.

Description

Technical Field

[0001] The present application relates to the field of interface detection technology, and in particular to a DP interface connection status detection method, system, and computer equipment. Background Art

[0002] The DisplayPort interface has become the mainstream standard for connecting high-definition display devices, and the HPD pin is a key component for hot-plug detection. Traditionally, DP interface connection status detection uses a single threshold method: a simple comparison of the HPD pin voltage level with a preset fixed threshold is performed. When the voltage level is above the threshold, it is considered connected; when it is below the threshold, it is considered disconnected. However, this single-threshold detection method faces serious reliability issues in practical applications. Especially in harsh environments with significant external interference, the HPD signal is susceptible to electromagnetic interference, power supply fluctuations, and other factors, causing drastic voltage fluctuations and frequent misjudgments of the connection status.

[0003] More critically, the traditional single-threshold detection method cannot simultaneously meet the two conflicting technical requirements of connection status accuracy and disconnection response speed. Setting a higher threshold can reduce the probability of false connections but increase the risk of false disconnections. Setting a lower threshold can improve the sensitivity of disconnection detection but also lead to an increase in false connection judgments. Furthermore, the single-threshold detection method is unable to handle fluctuations in the HPD signal near the threshold. When the signal repeatedly oscillates near the threshold, the system will repeatedly judge the device as connected and disconnected, causing the display device to frequently restart, seriously affecting the user experience. Summary of the Invention

[0004] The present application provides a DP interface connection status detection method, system and computer device, thereby improving the stability and anti-interference capability of DP interface connection status detection.

[0005] A first aspect of the present application provides a method for detecting a DP interface connection state, the method comprising: Collecting historical level signals of the HPD pin in the DP interface and preprocessing the historical level signals to obtain level preprocessing signals and level fluctuation characteristic data; Performing time domain decomposition processing on the level preprocessing signal and the level fluctuation characteristic data to obtain high and low level stable area identification results and stable area duration data; Calculating a high level threshold and a low level threshold using the high and low level stable area identification result and the stable area duration data to obtain a level dual threshold detection parameter; The real-time level signal of the HPD pin is detected based on the level dual threshold detection parameter to obtain the target connection state judgment result and the state judgment reliability value of the DP interface.

[0006] A second aspect of the present application provides a DP interface connection status detection system, the DP interface connection status detection system comprising: An acquisition module is used to collect historical level signals of the HPD pin in the DP interface and preprocess the historical level signals to obtain level preprocessing signals and level fluctuation characteristic data; a processing module, configured to perform time domain decomposition processing on the level preprocessing signal and the level fluctuation characteristic data to obtain high and low level stable area identification results and stable area duration data; a calculation module, configured to calculate a high-level threshold and a low-level threshold using the high- and low-level stable region identification result and the stable region duration data, to obtain a level dual-threshold detection parameter; The status detection module is used to perform status detection on the real-time level signal of the HPD pin based on the level dual threshold detection parameter to obtain the target connection status judgment result and status judgment reliability value of the DP interface.

[0007] The third aspect of the present application provides a computer device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the computer device executes the above-mentioned DP interface connection status detection method.

[0008] Compared with existing technologies, this application has the following advantages: By establishing a high- and low-level dual-threshold collaborative detection mechanism and introducing hysteresis characteristics, it effectively solves the problem of frequent state switching caused by HPD signal fluctuations near a single threshold, thereby improving the stability and anti-interference capability of DP interface connection state detection. It uses time-domain adaptive decomposition processing technology to accurately identify the stable and transitional regions of the HPD signal, achieving precise characterization of HPD signal characteristics. It also implements differentiated time constraints, sets different time thresholds for connected and disconnected states, and solves the problem of balancing the conflicting indicators of connection judgment accuracy and disconnection response speed. Through HPD signal jitter characteristic analysis and dynamic parameter calculation, a signal jitter index evaluation model is established, enabling accurate judgment of critical connection states and significantly reducing the false positive rate. A three-way parallel multi-channel hybrid detection network is introduced, focusing on the low-frequency main characteristics, high-frequency variation characteristics, and sequential logic characteristics of the HPD signal, providing comprehensive signal characteristic analysis and enhancing the comprehensiveness and robustness of the system's judgment. An adaptive parameter adjustment mechanism is established, allowing the system to dynamically optimize judgment parameters based on actual signal characteristics to adapt to HPD signal fluctuations in different usage environments, thereby improving the environmental adaptability of the detection method. A reliability index evaluation mechanism is introduced to monitor the reliability of status judgment results in real time, promptly identify potential judgment risks, and provide trigger conditions for additional detection processes, thus ensuring the reliable operation of the entire system. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0010] The structures, proportions, sizes, etc. depicted in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with this technology. They are not intended to limit the conditions under which the present invention can be implemented and therefore have no substantive technical significance. Any structural modifications, changes in proportional relationships, or adjustments in size should still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and objectives that can be achieved by the present invention.

[0011] Figure 1 1 is a flow chart of a method for detecting the connection status of a DP interface provided by an embodiment of the present invention; Figure 2 1 is a schematic block diagram of the structure of a DP interface connection status detection system provided by an embodiment of the present invention; Figure 3It is a schematic block diagram of the structure of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0012] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0013] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined, or partially merged, so the actual execution order may vary depending on the actual situation.

[0014] It should also be understood that the terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit the present application. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0015] It should be further understood that the term "and / or" used in this specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations. Figure 1 In one embodiment of the present application, a method for detecting the connection status of a DP interface includes: Step 100: Collect historical level signals of the HPD pin in the DP interface, and preprocess the historical level signals to obtain level preprocessing signals and level fluctuation characteristic data; It is understandable that the execution subject of this application can be a DP interface connection status detection system, or a terminal or a server, which is not limited here. The embodiment of this application is described by taking the server as the execution subject as an example.

[0016] Specifically, the analog level signal output by the HPD pin of the DP interface is continuously and accurately sampled. A dedicated high-speed sampling circuit, combined with a stable clock control module, continuously samples the analog level on the HPD pin at a fixed sampling frequency of 10kHz to accurately capture short-term level transitions, slight fluctuations, and high-frequency interference. The collected analog level signal is digitized using an analog-to-digital converter (ADC) with a 12-bit conversion accuracy. The converted digitized level signal retains the quantized level value of each sample, forming a continuous historical level sequence. The digitized level signal is then subjected to median filtering, and a sliding window mechanism is used to perform local denoising on the digital level sequence, with the window size set to 5 sampling points. The median filter has excellent pulse suppression capabilities, effectively eliminating transient spike interference and producing a relatively smooth preliminary filtered signal. To remove high-frequency noise signals with periodic or nonlinear characteristics, a wavelet threshold denoising algorithm is introduced based on the preliminary filtered signal. A three-layer wavelet decomposition using the db4 wavelet basis function is used, and a soft thresholding mechanism is applied in the wavelet domain. An adaptive threshold selection strategy is used. This method effectively balances signal fidelity and noise suppression to produce a level preprocessed signal. The mean, standard deviation, kurtosis, and skewness of the HPD pin level signal are calculated based on the level preprocessed signal. The mean describes the overall signal level, the standard deviation reflects the amplitude of signal fluctuations, the kurtosis indicates whether the signal contains sharp pulse components, and the skewness reveals whether the signal distribution is symmetrical. These four indicators together constitute the level statistical characteristics of the HPD pin level signal. By combining these level statistical characteristics with the microscopic fluctuation parameters of the level preprocessed signal, such as the level change rate, change amplitude, or amplitude density over a short period of time, we comprehensively construct level fluctuation characteristic data that characterizes the dynamic and stability characteristics of the HPD signal.

[0017] Step 200: performing time domain decomposition processing on the level preprocessing signal and the level fluctuation characteristic data to obtain high and low level stable area identification results and stable area duration data; Specifically, the level preprocessed signal is input into a wavelet decomposition system, where it is divided into components at different frequency levels using a three-layer discrete wavelet transform. A mother wavelet function suitable for the level signal's characteristics is selected to decompose the original level signal into a set of low-frequency approximate coefficients and a set of high-frequency detail coefficients. The low-frequency component reflects the signal's overall trend and stable changes and is the primary basis for assessing whether a stable connection exists. The high-frequency component, on the other hand, focuses on unstable characteristics such as sharp fluctuations, sudden changes, or jitter within the signal and is a key parameter for identifying external interference and signal switching. After the wavelet decomposition is completed, the energy distribution of the two types of coefficients is analyzed, and the ratio of low-frequency energy to high-frequency energy in each signal window is calculated. This stability assessment index for the level signal is constructed, reflecting the overall stability of the signal within a specific time period. If the value is significantly higher, the signal segment is relatively stable and serves as a candidate for subsequent identification of stable regions. The stability index is compared with the system's preset criteria to determine whether the current time window is a preliminary stable region. Regions that meet the criteria are marked, resulting in a preliminary stable region labeling result. The level values ​​in the preliminary stable area marking results are classified, and the areas with level values ​​higher than the first target value are marked as high-level stable areas, and the areas with level values ​​lower than the second target value are marked as low-level stable areas. A time constraint is imposed on the above level area division results, that is, it is determined whether the duration of each signal segment marked as a stable area exceeds the minimum valid duration set by the system. If the duration of a certain signal segment is too short, it is eliminated from the final result. Based on the remaining stable areas, the starting time point, ending time point and duration between the two time points of each high-level or low-level stable area are extracted one by one to form a stable area time data set.

[0018] Step 300: Calculate the high level threshold and the low level threshold using the high and low level stable region identification results and the stable region duration data to obtain the level dual threshold detection parameters; It should be noted that the average level values ​​of all areas identified as high-level stable areas and the average level values ​​of all areas identified as low-level stable areas are extracted from the high- and low-level stable area identification results. These statistical level parameters representing different stable time periods are integrated to form stable area level value distribution data. This data set contains multiple stable level values ​​collected in different time periods. Their distribution range, fluctuation trend, and boundary position play a decisive role in the subsequent adaptive setting of thresholds. Fluctuation amplitude analysis is performed on the level signals in the high- and low-level stable areas, and the standard deviation values ​​within each high- and low-level stable area are calculated separately to obtain two representative sets of standard deviations, namely, stable area fluctuation amplitude data. An adaptive algorithm with a safety margin compensation strategy is used to jointly determine the high and low level thresholds based on the level value distribution data and the fluctuation amplitude data. For the high-level threshold setting method, the minimum average level value is extracted from the high-level stable area, and the maximum standard deviation value is selected from the high-level standard deviation set. This maximum standard deviation is used as a correction factor, multiplied by a first preset multiple as a downward offset, and this offset is deducted from the minimum average value to obtain the high-level threshold. The setting of the low-level threshold adopts the opposite strategy. The maximum value is selected from the average level values ​​of all low-level stable areas, and then the maximum standard deviation is extracted from the low-level standard deviation set. After multiplying it by the second preset multiple, it is used as the upward offset and superimposed on the maximum average value to calculate the low-level threshold. Based on the extreme value correction method, it can effectively prevent misjudgments caused by local high noise or edge effects and improve the accuracy of connection and disconnection recognition. According to the stable area duration data, the time threshold for connection status judgment is set to M milliseconds, and the time threshold for disconnection status judgment is set to N milliseconds. M is greater than N to ensure the stability of connection identification and improve the response speed of disconnection. The high-level threshold, low-level threshold and time threshold parameters are unified and integrated to form a level dual-threshold detection parameter set.

[0019] Step 400: Perform status detection on the real-time level signal of the HPD pin based on the level dual-threshold detection parameters to obtain a target connection state judgment result and a state judgment reliability value of the DP interface.

[0020] Specifically, the real-time level signal from the HPD pin is collected continuously and at high frequency. A preprocessing mechanism consistent with the historical signal sampling stage is introduced, performing median filtering and wavelet denoising on the raw signal to generate a real-time level signal after interference suppression. This signal serves as the input for state detection and is compared point-by-point with the high and low thresholds in the previously calculated dual-threshold detection parameters. Based on the level at each moment, the level is determined to be in the high, low, or intermediate fuzzy region between the two, generating a real-time level comparison result. Based on this level comparison result, a set of clear state transition rules is applied to derive the current connection state trend. Specifically, when the real-time level signal gradually rises from below the low threshold and eventually exceeds the high threshold, the current state is considered connected. Conversely, when the signal drops from above the high threshold to below the low threshold, the disconnected state is considered, generating a preliminary state transition determination result. Because actual level signals can experience slight fluctuations or false thresholds within a short period of time, a time validation mechanism is introduced to prevent misjudgments. This imposes time constraints on the initial transition results. Specifically, a connection state transition must remain above the upper threshold for M milliseconds to be confirmed, while a disconnection state transition must remain below the lower threshold for at least N milliseconds to be considered a valid transition. This asymmetric time constraint strategy improves the stability of connection state determination while ensuring rapid response to disconnection events. Furthermore, in the transition region between the lower and upper thresholds, the current state preservation mechanism maintains the previously determined connection state, preventing frequent switching of connection states due to short-term fluctuations. An intermediate state sequence containing the preservation processing results is output. Combining these time constraint validation results with the state preservation processing logic, a preliminary state judgment sequence is generated, serving as the initial connection state judgment result for the current moment. Based on this initial judgment result and combined with previously extracted stable region duration data, the level sequence at the critical boundary is analyzed to identify potential state uncertainties or transition risks, thereby deriving a more accurate critical connection state judgment result. Multi-channel detection is performed based on the critical connection status judgment results and real-time level signals. Multiple independent detection channels that focus on the main level trend, high-frequency disturbance characteristics, and state timing characteristics are processed in parallel. Each channel draws its own state judgment opinion based on information from different dimensions. The decision fusion mechanism then uses comprehensive weight calculation to obtain the final target connection state judgment result, and calculates the current state judgment reliability value based on the credibility of each channel output.

[0021] The real-time level signal of the HPD pin is analyzed within different state intervals. By counting the amplitude of the small fluctuations in the level signal within the stable zone, the stable zone jitter index is calculated. This index reflects the degree of internal disturbance of the level signal in the apparent stable state. Simultaneously, the rate of change of the transition zone signal, which does not meet the stabilization time requirement or lies between the dual level thresholds, is analyzed. The level change rate and amplitude are extracted from this analysis to calculate the transition zone jitter index, which effectively characterizes the severity of signal fluctuations during critical switching. Together, these two constitute the signal jitter characteristic parameter that reflects the volatility of the HPD signal state. Based on these signal jitter parameters and the initial connection state, a critical connection state evaluation function is constructed. This function integrates the comprehensive relationship between the real-time level value, jitter intensity, and the previous state. A weighting strategy is used to assign different judgment contributions to different parameters, forming a dynamically adjusted multi-factor evaluation mechanism. The evaluation function is calculated in real time, outputting an evaluation function value used to determine the credibility of the current state. This function value is then compared with a preset evaluation threshold. When the evaluation function value exceeds the threshold, it indicates that the current state is approaching a stable and reliable connection. Conversely, if it falls below the threshold, it indicates that the system is disconnected or extremely unstable, generating a preliminary evaluation function judgment result. To improve the accuracy and interpretability of critical state judgment, the initial state judgment result and signal jitter characteristic parameters are used as input to construct a state transition decision tree with three-layer judgment logic. At the first layer, this decision tree determines the state trend by comparing the positional relationship between the level signal and the upper and lower thresholds. At the second layer, the jitter index is used to determine whether the signal has potential unstable behavior. At the third layer, the current state duration is combined with historical change records to assess whether the state meets the reliability standard. The three-layer logic refines the judgment step by step through conditional progression, forming a structured decision tree result. This ensures that state identification not only relies on the current signal value but also considers its changing trajectory and historical trends. Furthermore, to address the problem of frequent signal fluctuations in the critical level region, the state transition behavior of the real-time level signal is continuously statistically analyzed, focusing on whether the time intervals between multiple consecutive state transitions are stable and reliable. If a number of consecutive state transitions (e.g., E times) are detected, each with an interval shorter than the minimum stability time defined in the stability zone duration data, the system is deemed to be in a critical disturbance environment. To prevent misjudgment, the system proactively triggers a parameter adaptation mechanism, appropriately widening or narrowing the upper and lower intervals between the level dual thresholds and simultaneously adjusting the time threshold used to determine state validity, resulting in a set of dynamically revised target judgment parameters. The evaluation function results, decision tree results, and automatically adjusted target judgment parameters are integrated, and a critical connection state judgment result is generated through a logical weighting and credibility comprehensive calculation mechanism.

[0022] Based on the real-time level signal, the overall trend and low-frequency stability characteristics are extracted. Low-frequency approximation coefficients are obtained through wavelet decomposition. Combined with the average level value within the identified stable regions, the corresponding standard deviation of the level, and the duration of each stable region, the feature information required for the first channel is constructed, forming a first eigenvector reflecting the signal's low-frequency behavior. Simultaneously, the high-frequency components of the level signal are analyzed, and high-frequency detail coefficients at each level are extracted through wavelet decomposition. The real-time level change rate is calculated to capture sudden changes. Combined with the jitter index, this forms the input feature data for the second channel, constructing a second eigenvector reflecting the signal's rapidly changing characteristics. Furthermore, a temporal logic feature vector is constructed using the critical connection state judgment results. By comprehensively considering the current judgment result and its state history from previous frames, these states are encoded as a time series, resulting in a third feature input representing the connection state evolution path. This third feature input complements the logical continuity and historical stability in state recognition, enhancing adaptability to complex state transitions. The feature vectors from these three different sources are fed into three pre-set weighted decision tree sub-detectors. These sub-detectors perform decision reasoning on the input data in their respective feature domains based on pre-trained weight parameter structures, outputting three independent connection state judgment results. The three independent channel state judgment results are then fused, using a weighted voting mechanism to dynamically adjust channel weights based on their importance and historical judgment accuracy. This fusion results in a single target connection state judgment. While outputting the target state, the reliability of this decision is calculated. The output of each channel in the fused judgment is multiplied by its corresponding weight, and the absolute value of the weighted sum is then calculated and compared with the maximum value of the current channel weight to determine the confidence level of this state judgment. A higher ratio indicates greater consistency and certainty in the system's current state judgment, indicating a more reliable judgment result. Conversely, if the ratio approaches the lower weight limit or if there is disagreement among the channel judgments, the system lacks confidence in the current state judgment, triggering additional judgments or alerting external systems to abnormal signal states.

[0023] The three types of eigenvectors are processed separately and structured modeled. In the low-frequency channel processing workflow, the first eigenvector is used as input. This vector includes features such as the low-frequency approximation coefficient, stable zone level, standard deviation, and stable time. A linear weighted algorithm is used to assign importance scores to each feature. The contribution of each feature to state discrimination is evaluated using prior training data, resulting in a set of weight vectors reflecting feature significance. The original eigenvectors are then rearranged according to this feature importance sequence, so that the sorted first weighted vector aligns with the discrimination priority, enhancing the model's ability to discriminate across key feature dimensions. The first weighted vector is then input into a weighted decision tree sub-detector corresponding to the low-frequency channel. A decision tree consisting of four layers of deep splitting is executed. Each layer makes decisions based on the current eigenvalue and the threshold boundaries established by the historical split nodes, gradually forming path splits and ultimately outputting the low-frequency channel state judgment result. This result is used to identify stable trends in connection states over long timescales and has strong fault tolerance for long-term signal characteristics. Meanwhile, the high-frequency channel processing path takes as input a second eigenvector, which contains information such as a set of high-frequency detail coefficients, level change rate, and jitter index that can sensitively reflect significant signal fluctuations. This vector undergoes nonlinear feature transformation, enhancing and reconstructing the original features using strategies such as function combination, cross-product, or exponential transformation. This generates a second weighted vector containing high-order feature information, enhancing its discriminability in identifying short-term sudden state changes. This vector is then fed into the weighted decision tree sub-detector corresponding to the high-frequency channel. A split calculation process is performed within a three-layer deep architecture, leveraging its sensitivity to high-frequency perturbations to achieve a rapid response mechanism. The resulting high-frequency channel state judgment is suitable for handling transient instabilities caused by poor connection and electromagnetic interference. Furthermore, to effectively capture the logical patterns of the DP interface connection state over time, a temporal logic eigenvector is processed. This vector consists of the initial judgment results and critical state markers at several preceding and subsequent moments. A state transition probability matrix is ​​constructed based on the discrete state sequence. This matrix characterizes the probability distribution of the system transitioning to the next state under different historical states by statistically analyzing the frequency of state transitions between consecutive time slices. By combining the transition characteristics of the Markov chain, the conditional probability distribution of the connection or disconnection state occurring at the current moment is calculated to form a time series state prediction model that can predict the state evolution trend. Based on this model, a time series decision tree with a weight control mechanism is constructed. The three aspects of information, transition probability, state duration, and state consistency, are used as the judgment basis. Through regularized processing, abnormal frequent switching or sudden changes are identified, and the state judgment results of the time series channel are output. The state judgment results of the low-frequency channel, the high-frequency channel, and the time series channel are integrated into the state judgment output of three independent channels. These results are weighted and integrated in the subsequent decision fusion module based on the historical accuracy and confidence assessment results of each channel to form the final DP interface connection state judgment result and its reliability score.

[0024] In the embodiment of the present application, by establishing a high-low level dual threshold collaborative detection mechanism and introducing a hysteresis characteristic, the problem of frequent state switching caused by HPD signal fluctuations near a single threshold is effectively solved, and the stability and anti-interference ability of the DP interface connection state detection are improved. The time domain adaptive decomposition processing technology is used to accurately identify the stable area and transition area of ​​the HPD signal, achieve accurate characterization of the HPD signal characteristics, implement differentiated time constraints, set different time thresholds for the connected state and disconnected state, and solve the problem of balancing the conflicting indicators of connection judgment accuracy and disconnection response speed. Through HPD signal jitter characteristic analysis and dynamic parameter calculation, a signal jitter index evaluation model is established, which realizes accurate judgment of critical connection state and significantly reduces the false positive rate. A three-way parallel multi-channel hybrid detection network is introduced, focusing on the low-frequency main characteristics, high-frequency change characteristics and timing logic characteristics of the HPD signal respectively, providing a comprehensive signal characteristic analysis, enhancing the comprehensiveness and robustness of the system judgment. An adaptive parameter adjustment mechanism is established, and the system can dynamically optimize the judgment parameters according to the actual signal characteristics, adapt to the HPD signal fluctuations in different usage environments, and improve the environmental adaptability of the detection method. A reliability index evaluation mechanism is introduced to monitor the reliability of status judgment results in real time, promptly identify potential judgment risks, and provide trigger conditions for additional detection processes, thus ensuring the reliable operation of the entire system.

[0025] In a specific embodiment, the process of executing step 100 may specifically include the following steps: Continuously sampling the analog level signal of the HPD pin in the DP interface to obtain a historical level signal, and digitizing the historical level signal to obtain a digitized level signal; Performing median filtering on the digitized level signal to obtain a preliminary filtered signal, and performing wavelet threshold denoising on the preliminary filtered signal to obtain a level preprocessed signal; Calculate the mean, standard deviation, kurtosis, and skewness statistics of the HPD pin level signal based on the level preprocessing signal, and combine the mean, standard deviation, kurtosis, and skewness statistics into level statistical features; The level statistical characteristics and the fluctuation parameters of the level preprocessing signal are combined to obtain the level fluctuation characteristic data.

[0026] Specifically, the analog voltage level signal from the HPD pin of the DP interface is continuously sampled. The sampling circuit should be equipped with a high-frequency sampling module and a timed sampling mechanism aligned with the system master clock to avoid non-uniformity or discontinuities in data acquisition. The sampling frequency should be set at or above 10kHz to effectively capture any bursts, weak level jitter, or high-frequency noise in the HPD signal. During the sampling process, the sampling circuit reads the analog voltage output from the HPD pin in real time and feeds the analog voltage level signal at each moment into the analog-to-digital conversion unit for digitization. During the digitization process, a high-precision analog-to-digital converter (ADC) with a 12-bit or higher resolution is used to discretize the analog voltage range (e.g., 0-3.3V) into 4096 levels, achieving a voltage resolution of approximately 0.8 millivolts per level. High-resolution sampling is critical for capturing subtle level changes in the HPD signal and significantly improves system recognition accuracy when determining the edge of a physical connection state or when the signal jitter amplitude approaches the false positive threshold. Through the continuous operation of the analog-to-digital converter, the sampling system converts the analog level sequence into a continuous sequence of digitized level signals, which are stored as a time-series sample stream. The digitized level signal undergoes median filtering, traversing the level data sequence in a sliding window format. The median of each sample point within the window is taken as the replacement value for the center point. Median filtering is suitable for eliminating outlier interference caused by transient pulses, switching power supply interference, or nonlinear response of analog circuits. A wavelet threshold denoising mechanism is introduced based on median filtering. The level signal is decomposed at multiple scales and converted into the wavelet domain. Noise signals are identified and suppressed in the wavelet coefficient space. A three-layer discrete wavelet decomposition is performed using the db4 wavelet basis function. A soft threshold compression operation is applied to the decomposed wavelet coefficients to attenuate insignificant high-frequency components. The threshold value is determined using an adaptive strategy. Typical algorithms dynamically calculate the appropriate compression limit based on the signal's noise standard deviation and signal length. This wavelet denoising method preserves the key characteristics of the level signal while reducing uncontrollable noise components, generating a level preprocessed signal with a clear structure and prominent fluctuation characteristics. The overall statistical characteristics of the level preprocessing signal are extracted to characterize the basic level distribution and fluctuation pattern of the signal. The mean of the preprocessed signal is calculated. This value reflects the overall level of the signal within the observation window and is an important basic parameter for determining connection stability. The standard deviation is then calculated to measure the intensity of the signal's fluctuation around the mean. A larger value indicates greater level instability. The kurtosis value is extracted to identify whether there are high-frequency sharp mutations or local extreme value clusters in the signal. A higher kurtosis indicates that the signal contains periodic interference or sharp level jumps. The skewness is calculated to measure the symmetry of the signal within the overall distribution. A positive or negative skewness indicates that the level signal is more prone to abnormal fluctuations in a certain direction.The above four statistics constitute the signal's level statistical characteristics. Combined in vector form, they reflect the central tendency, degree of dispersion, extreme value behavior, and distribution of the level signal. Combining the level statistical characteristics with the fluctuation parameters of the level preprocessing signal generates more discriminative level fluctuation characteristic data. The system calculates micro-indicators such as the signal's local rate of change, maximum rise speed, maximum fall amplitude, oscillation period, and short-term amplitude density within a short time window. These fluctuation parameters are then integrated with statistical values ​​such as mean, standard deviation, kurtosis, and skewness to construct a high-dimensional feature set reflecting both the global and local characteristics of the signal.

[0027] In a specific embodiment, the process of executing step 200 may specifically include the following steps: Perform three-layer discrete wavelet transform on the level preprocessed signal to obtain a low-frequency approximate coefficient set and a high-frequency detail coefficient set; Calculating the energy distribution proportion value based on the low-frequency approximation coefficient set and the high-frequency detail coefficient set to obtain a level signal stability evaluation index, and comparing the level signal stability evaluation index with a preset threshold to obtain a preliminary stable area marking result; Classifying the level values ​​in the preliminary stable area marking result, marking the area greater than the first target value as a high-level stable area, and marking the area less than the second target value as a low-level stable area, to obtain a level partition marking result; Perform minimum duration constraint filtering on the level partition marking results to obtain the high and low level stable area identification results; According to the high and low level stable area identification results, the start time, end time and duration of each stable area are calculated to obtain the stable area duration data.

[0028] Specifically, the preprocessed signal is decomposed using a discrete wavelet transform (DWT) to obtain a multi-scale coefficient set that encompasses both level variation trends and local disturbance components, thereby distinguishing the signal's energy distribution and variation characteristics at different time scales. The db4 wavelet basis function, which is well-suited to the structural characteristics of the level signal, is selected. It features tight support and smoothness, balancing the local expression of high-frequency features with the global resolution of low-frequency trends. The preprocessed level signal is then subjected to a three-layer discrete wavelet decomposition operation. The decomposition results include a set of low-frequency approximation coefficients in the third layer and high-frequency detail coefficients from the first to third layers. The low-frequency approximation coefficients represent the signal's slow-changing trend over the entire time range, while the high-frequency detail coefficients reflect rapid transitions, short-period disturbances, or high-frequency interference. Energy analysis is used to calculate the total energy ratio of the low-frequency and high-frequency coefficients, measuring the energy concentration of the signal in different frequency bands. Since a stable level signal has a smooth waveform with most of its energy concentrated in the low-frequency components, the ratio of low-frequency energy to total high-frequency energy is used as a key indicator of stability, namely, the level signal stability assessment index. The higher the index is, the closer the signal is to a stable state; conversely, if the high-frequency component occupies a larger amount of energy, it means that there are more disturbances and discontinuous changes in the signal. The evaluation index is then compared with the preset stability threshold. When the stability index exceeds the threshold, the current signal segment is marked as a potential stable area. If it does not exceed the threshold, it is considered unstable or in a transitional state, and the marking result of the preliminary stable area is generated. The level signal values ​​contained in the potential stable area are analyzed and classified by comparing their mean levels with the preset target level threshold. If the level value in a certain stable area is higher than the first target value as a whole, the area is classified as a high-level stable area; if the level value is lower than the second target value as a whole, it is classified as a low-level stable area; and the area between the two target values ​​is not marked as a valid stable area for the time being because its level characteristics are not obvious and cannot be accurately attributed to a fixed level state. The aforementioned level partitioning strategy is based on modeling the physical characteristics of real connection states. Since the DP interface HPD pin typically exhibits a high level when connected and a distinct low level output when disconnected, a preliminary judgment and distinction of the physical connection state is achieved by comparing the thresholds of the average level of the stable zone, resulting in a level partitioning result. Based on the level partitioning results, the duration of each segment marked as a high-level or low-level stable zone is determined to determine whether it meets the minimum duration requirement. This duration threshold parameter, derived from a statistical analysis of HPD signal behavior under real connection states, is set in the order of milliseconds to over ten milliseconds to eliminate short-period fluctuating signal segments that are mistakenly marked as stable zones due to transient disturbances or misjudgments. If a signal segment meets the level classification criteria but its duration falls below the time threshold, it is deemed a non-real stable state and excluded from the identification results. Otherwise, it is retained as a valid stable zone.After completing the above filtering process, the stable zone results that meet the time domain, frequency domain, level characteristics and duration constraints are obtained, that is, the high and low level stable zone identification results. On this basis, the time attributes of each stable zone are analyzed and extracted, including the starting time point, end time point and stable duration between the two of each stable zone. These time data are used to reflect the persistence and signal stability of the current level state, and serve as the core reference data for subsequent state judgment. For example, when judging whether the connection state is confirmed or the disconnection state is established, the time a certain state is maintained is compared with the preset connection threshold or disconnection threshold to determine whether the state has sufficient persistence. At the same time, these time features are used to dynamically adjust the judgment parameters of the system under different operating environments to achieve adaptive response to the influence of ambient temperature, electromagnetic interference, cable quality, etc.

[0029] In a specific embodiment, the process of executing step 300 may specifically include the following steps: Extracting the average level value set of the high-level stable area and the average level value set of the low-level stable area from the high-level and low-level stable area identification results to obtain stable area level value distribution data; Calculate the respective standard deviation sets based on the level signals in the high-level stable area and the low-level stable area to obtain the fluctuation amplitude data of the stable area; Obtain a high level threshold based on the minimum high level value of the stable area level value distribution data minus the maximum high level standard deviation value of the stable area fluctuation amplitude data of a first preset multiple; A low level threshold is obtained based on the maximum low level value of the stable area level value distribution data plus the maximum low level standard deviation value of the stable area fluctuation amplitude data of a second preset multiple; According to the stable zone duration data, the time threshold for connection state judgment is set to M milliseconds, and the time threshold for disconnection state judgment is set to N milliseconds to obtain the dual-threshold time constraint parameters; The high level threshold, low level threshold and dual-threshold time constraint parameters are integrated to obtain the level dual-threshold detection parameters.

[0030] Specifically, the high- and low-level stable zone identification results are analyzed and data extracted. This identification result marks the stable state of the DP interface HPD pin signal in multiple time periods, including stable segments belonging to high levels and stable segments belonging to low levels. By traversing these stable regions, the level signals within each stable region are statistically averaged, and the average level values ​​of each segment in the high-level stable region are extracted. This is then formed into an ordered set, which serves as reference data representing the actual level level in the high-level state. Similarly, the average value of each level signal segment in the low-level stable region is extracted, and a corresponding low-level average value set is constructed. These two sets reflect the level distribution of the DP interface in the connected and disconnected states, respectively, and are used to describe the center position of the level in the stable state. Together, they constitute the stable zone level value distribution data. Based on the level signals in the high-level stable region and the low-level stable region, the respective standard deviation sets are calculated to obtain the stable zone fluctuation amplitude data, which effectively reveals whether there are potential instability factors such as obvious jitter, interference, or transient disturbances in the high-level and low-level regions. Based on the stable zone level value distribution data and the fluctuation amplitude data, high and low level judgment thresholds are constructed. To calculate the high-level threshold, the minimum average level value is found from the set of average values ​​in the high-level stable region, serving as the lower boundary reference value for the level in that region. The maximum value is then selected from the set of high-level standard deviations, representing the amplitude of the most volatile segment among all high-level segments. A first preset multiple parameter is then introduced to expand the influence of the standard deviation on the threshold calculation, preventing the high-level judgment boundary from being too compact due to an accidental increase in the amplitude of fluctuation during actual operation. The maximum standard deviation value is multiplied by the multiple and then deducted from the minimum high-level value to obtain the high-level threshold. This high threshold is adaptive and automatically shifts according to the stability of the level signal during actual operation. The more volatile the signal fluctuations, the lower the high threshold is set to increase the fault tolerance for determining the connection status. In cases of smaller fluctuations, the system tends to set a higher high-level threshold that is closer to the actual signal level to improve the accuracy of the judgment. The low-level threshold is constructed using the opposite strategy. The maximum value is extracted from the set of average level values ​​in the low-level stable region, serving as the upper boundary reference for the low-level signal. The maximum value is then extracted from the set of low-level standard deviations to reflect the maximum fluctuation amplitude in the low-level state. A second preset multiplier parameter is introduced to weight this maximum standard deviation and then add it to the low-level maximum value to form the low-level threshold. The low-level threshold is positioned slightly above the low-level signal, effectively accommodating slight deviations in normal signals while preventing slight upward fluctuations from being mistakenly identified as a connection state due to being too close to the actual level.By dynamically constructing upper and lower boundaries, the high- and low-level detection regions are effectively isolated, with sufficient hysteresis space between them. This prevents frequent state switching during small signal fluctuations near the boundary, enhancing the system's stability and reliability under boundary conditions. Based on previously extracted stable zone duration data, the typical duration of stable states in historical samples is analyzed to set time thresholds for connected and disconnected states. A longer confirmation time, such as M milliseconds, is set for the connected state, ensuring that the HPD signal must remain stable in the high-level region for a period exceeding this time before the connection is confirmed valid. A shorter time, such as N milliseconds, is set for the disconnected state, requiring a faster response time. The disconnected state is confirmed valid only after the low-level signal reaches this threshold. This asymmetric time setting logic reflects the system's rational differentiation of the dynamic response characteristics of physical connection events, emphasizing robustness in connection confirmation and speed in disconnection responses. These two complement each other, thereby improving the real-time and reliability of state detection. The three core judgment elements, high-level threshold, low-level threshold, and connection and disconnection time thresholds, are integrated to form a level dual-threshold detection parameter structure.

[0031] In a specific embodiment, the process of executing step 400 may specifically include the following steps: The real-time level signal of the HPD pin is collected and preprocessed to obtain the real-time level signal; Compare the real-time level signal with the high level threshold and the low level threshold in the level dual threshold detection parameters to obtain a level state comparison result; Applying a state transition rule according to the level state comparison result, determining that a connection state transition occurs when the real-time level signal rises from below a low level threshold to above a high level threshold, and determining that a disconnection state transition occurs when the real-time level signal drops from above a high level threshold to below a low level threshold, thereby obtaining a preliminary state transition determination result; Applying the dual-threshold time constraint parameter in the level dual-threshold detection parameter to the preliminary state transition determination result for time verification, applying the M millisecond time threshold to the connected state and the N millisecond time threshold to the disconnected state, to obtain a time constraint verification result; Applying a state holding mechanism to an area where the real-time level signal is between a low level threshold and a high level threshold, keeping the previous state judgment unchanged, and obtaining a state holding processing result; Generate a state judgment sequence according to the time constraint verification result and the state preservation processing result to obtain the initial connection state judgment result; Combine the initial connection state judgment result and the stable zone duration data to perform critical state analysis and obtain the critical connection state judgment result; Multi-channel detection is performed based on the critical connection status judgment result and the real-time level signal to obtain the target connection status judgment result and status judgment reliability value of the DP interface.

[0032] Specifically, the HPD pin voltage signal is collected in real time. This signal directly reflects the DP interface connection status, and its voltage level changes can describe the device's connection and disconnection status. To ensure real-time sampling and integrity, the sampling frequency should be set to 10kHz or higher. A high-speed analog-to-digital conversion module converts the continuous analog voltage signal into a high-resolution digital voltage value. The collected signal undergoes preprocessing, using a median filter to eliminate sudden noise points and a wavelet denoising algorithm for signal smoothing and noise suppression, resulting in a real-time voltage signal curve. The real-time voltage signal is compared with the high and low thresholds in the dual-threshold detection parameters to generate a voltage state comparison result, identifying the basic voltage state range of the current HPD signal. Based on the voltage state label, whether the voltage value changes across different ranges, that is, the state transition trend, is identified. When the voltage signal rises from below the low threshold and then crosses the high threshold within a short period of time, the predefined state transition rule determines this as a state transition from disconnected to connected. When the voltage signal falls from above the high threshold to below the low threshold, it is determined as a state transition from connected to disconnected. This judgment mechanism, based on a typical hysteresis control model, effectively avoids repeated signal transitions near critical levels. By introducing upper and lower thresholds, it enhances the system's stability and anti-interference capabilities within level boundary states. Time constraint parameters are introduced to continuously verify state transition trends. This time verification process uses the pre-set connection confirmation time (M milliseconds) and disconnection confirmation time (N milliseconds) in the dual-threshold detection parameters as the timing basis for determining whether a state change has occurred. The system continuously monitors the level maintenance after a state transition. If a connection state transition is identified, the level signal must remain stable in the high level region for at least M milliseconds to be considered a true connection. Conversely, if a disconnection state transition is identified, it only needs to remain in the low level region for at least N milliseconds to be considered a valid disconnection. The asymmetric time setting takes into account the DP interface's practical requirements of requiring a longer confirmation period during physical connection and a faster response during disconnection. This optimizes the bidirectional control logic for state changes and outputs time constraint verification results. For the ambiguous region between the high and low thresholds, a state retention mechanism is introduced to stabilize the judgment logic. Under this mechanism, if the signal level fails to meet the state transition conditions and remains in the middle range between the upper and lower thresholds, the current state remains unchanged, and the system automatically maintains the previously determined connected or disconnected state. This mechanism effectively prevents frequent state reversals caused by slight voltage fluctuations, improving the stability and continuity of the judgment process. Combining the state retention logic with the aforementioned time verification mechanism, a continuous state judgment sequence is generated, forming the initial connection state judgment result. This initial judgment result is integrated with the stable zone duration data to perform higher-level critical state analysis.This analysis primarily addresses misjudgments that occur when the HPD signal is in a borderline state, jittering, or undergoing rapid and frequent state changes. During this phase, the system reassesses the persistence and credibility of the state transition trends near the threshold boundary in the judgment results. If the interval between multiple consecutive state changes is less than the minimum duration threshold in the stable zone, the system triggers a parameter adaptation mechanism to dynamically adjust the distance between the upper and lower thresholds and the state confirmation time value to improve judgment stability in critical disturbance environments and output a critical connection state judgment result. Multi-channel detection is performed based on the critical connection state judgment result and the real-time level signal. The multi-channel detection mechanism constructs three feature channels to process the low-frequency stability trend of the level, the high-frequency disturbance characteristics, and the state temporal evolution logic. Low-frequency approximate features, high-frequency detailed features, and temporal logic features are extracted and fed into three weighted decision submodules. Each submodule generates a single-channel state recognition result based on its independent judgment model. The decision fusion module integrates the judgment outputs of the three channels and performs a weighted voting based on the channel weights to produce the final target connection state judgment result. At the same time, the confidence consistency between each channel weight and the current voting result is analyzed, and the ratio of the decision confidence to the maximum channel weight is calculated to obtain the reliability value of the state judgment. This value effectively reflects the trust level of the current state recognition. If the value is low, the system triggers additional judgment strategies or marks the current state as low reliability, thus building a connection state detection mechanism with self-sensing, self-adjusting, and multi-dimensional fault tolerance capabilities.

[0033] In a specific embodiment, the step of performing critical state analysis in combination with the initial connection state determination result and the stable zone duration data to obtain the critical connection state determination result may specifically include the following steps: The jitter index in the stable region and the jitter index in the transition region are calculated based on the real-time level signal of the HPD pin, and the jitter index in the stable region and the jitter index in the transition region are used as signal jitter characteristic parameters; Constructing a critical connection state evaluation function, and calculating an evaluation function value based on the critical connection state evaluation function, comparing the evaluation function value with a preset evaluation threshold, determining a connected state when the value is greater than the preset evaluation threshold, and determining a disconnected state when the value is less than the preset evaluation threshold, and obtaining an evaluation function determination result; A state transition decision tree with three-layer judgment logic is constructed based on the initial connection state judgment result and signal jitter characteristic parameters, and a decision tree judgment result is generated through the state transition decision tree; The real-time level signal of the HPD pin is counted for the number of consecutive state transitions. When the interval between E consecutive state transitions is less than the minimum stable time in the stable zone duration data, the dual threshold interval and time threshold are automatically adjusted to obtain the target judgment parameters. The critical connection state judgment result is generated according to the evaluation function judgment result, the decision tree judgment result and the target judgment parameter.

[0034] Specifically, the HPD pin level signal is divided into a stable region and a transition region. The standard deviation and mean of the level signal within the stable region are calculated, and the ratio of the standard deviation to the mean level is used as the jitter index for that stable region. A higher value indicates a more unstable stable state. In the transition region, the level change rate is calculated and combined with the change duration to quantify the severity of the level signal during the transition state. This transition region jitter index reflects the intensity and speed of level fluctuations during state transitions. The stable and transition region jitter indices are used as signal jitter characteristic parameters. Based on these jitter parameters, a critical connection state evaluation function is constructed. This function comprehensively considers three factors: the current level value, signal jitter intensity, and historical state labels in a multi-factor weighted manner. The level value reflects the static state distribution, the jitter parameter characterizes the degree of transient disturbances, and the historical state represents the trend inertia of state switching. These three dimensions are weighted and summed by setting weight coefficients to produce a continuously changing evaluation function value that dynamically reflects the current system's "degree of deviation" between connected and disconnected states. The evaluation function value is compared with a pre-set evaluation threshold. When its value exceeds the threshold, it indicates that the overall system state is leaning toward the connected end, resulting in a connected state. When its value is below the threshold, the current state is considered to be disconnected, resulting in the evaluation function's judgment result. To enhance the hierarchy and stability of the judgment logic, a state transition decision tree with three layers of judgment logic is constructed, combining the initial connection state judgment result and the aforementioned signal jitter characteristic parameters. At the first layer, the basic state type is determined based on the position of the current level relative to the dual thresholds, that is, whether the current level is high, low, or in the intermediate range. At the second layer, a jitter index comparison mechanism is introduced. By comparing the jitter index in the stable or transitional zone with a preset jitter threshold, the current state is determined to be stable or experiencing frequent disturbances. At the third layer, a comprehensive judgment is made based on the duration of historical states and the duration of the most recent state change, identifying the trend and persistence of the state. The three layers of logic are progressive and complementary, constructing a comprehensive state judgment framework from static levels to dynamic fluctuations to state evolution, and outputting the final decision tree judgment result. To accommodate the frequent state transitions of level signals under special conditions, a state transition frequency statistics mechanism is introduced to continuously analyze the real-time level signal of the HPD pin on a timeline. If several consecutive state transitions are detected within a short period of time, and the intervals between these transitions are all less than the minimum stability time threshold defined in the stability zone duration data, the signal is deemed to be experiencing a critical disturbance trend, indicating frequent misjudgments due to factors such as unstable cable contact and external interference.To address this situation, an adaptive parameter adjustment mechanism is automatically triggered, dynamically increasing the dual-threshold interval to expand the safety margin for level judgment. It also appropriately raises the time confirmation thresholds for connected and disconnected states to ensure greater stability and continuity in state changes. This mechanism adaptively adjusts key parameters based on environmental changes, constructing target judgment parameters and making the judgment logic more robust and resilient to interference. The evaluation function results, the three-layer decision tree results, and the adaptively adjusted target judgment parameters are combined and integrated to form the final critical connection state judgment result through a priority judgment mechanism and credibility assessment strategy.

[0035] In a specific embodiment, the step of performing multi-channel detection based on the critical connection state judgment result and the real-time level signal to obtain the target connection state judgment result and the state judgment reliability value of the DP interface may specifically include the following steps: Extracting a low-frequency approximation coefficient, a stable region level value, a standard deviation, and a stable time according to the real-time level signal, and constructing a first eigenvector according to the low-frequency approximation coefficient, the stable region level value, the standard deviation, and the stable time; extracting a high-frequency detail coefficient set, a level change rate, and a jitter index according to the real-time level signal, and constructing a second eigenvector according to the high-frequency detail coefficient set, the level change rate, and the jitter index; Constructing a temporal logic feature vector based on the critical connection state judgment result; Input the first eigenvector, the second eigenvector, and the temporal logic eigenvector into the corresponding weighted decision tree sub-detector for state judgment, and obtain the state judgment results of three independent channels; The state judgment results of the three independent channels are fused to obtain the target connection state judgment result of the DP interface. The ratio of the decision confidence of the target connection state judgment result to the maximum weight is calculated to obtain the state judgment reliability value.

[0036] Specifically, the system extracts highly structural low-frequency features from the real-time level signal. The real-time HPD level signal is subjected to a discrete wavelet transform (DWT). The wavelet decomposition results are used to extract the third-level low-frequency approximate coefficients. These coefficients represent the overall stability of the signal over time and reflect the trend of the connection state over a large time window. Furthermore, combined with the results of the stable region analysis, the average level, standard deviation, and stability duration of the high-level or low-level stable state segments are extracted from the current signal. The average level reflects the voltage level during the connection state, the standard deviation measures the level fluctuation, and the stability duration indicates the duration of the current state. These four features together constitute the first eigenvector of the system, a quantitative expression of the signal's macroscopic and temporal stability, which is used to identify stable states in the low-frequency sub-detector channel. The system captures the transient characteristics of the real-time level signal and extracts a set of high-frequency detail coefficients from the first, second, and third layers through a wavelet transform process. These coefficients capture rapid changes in the signal, such as short-term jumps, edge perturbations, and spikes, reflecting local changes in the connection state boundary when it is unstable. At the same time, the signal's rate of change per unit time is calculated to capture the level slope and transient transition speed. Combined with the previously obtained jitter index in the stable or transition regions, this constructs a comprehensive description of the dynamic characteristics of the level signal under high-frequency perturbations. These characteristics are combined to form a second eigenvector, representing the short-term instability and response sensitivity of the HPD level signal. This is suitable for quickly detecting abnormal conditions in scenarios with frequent physical connection changes, severe electromagnetic interference, or intermittent cable contact. Based on the critical connection state judgment results, a temporal logic feature vector is constructed. This vector records the initial connection state and critical judgment results at the current and previous moments in time, using time windows as units. Sequential features such as state hold time, number of state flips, state persistence, and transition direction are extracted based on the state transition path. This approach constructs a temporal logic trajectory for predicting the direction of the connection state. This trajectory can determine whether a stable connection exists and infer whether the system is transitioning from connected to disconnected or disconnected to connected. This type of temporal feature is suitable for scenarios where boundary states are difficult to define. When the signal itself lacks sufficient information, observing the state evolution trend can still enable reasonable judgment, forming the third eigenvector. After obtaining the three eigenvectors, they are fed into three corresponding weighted decision tree sub-detectors. Each sub-detector processes the eigenvectors of its specific dimension using an independent decision model structure. The low-frequency channel outputs macroscopic state judgments based on level trends and stability. The high-frequency channel identifies short-term abnormal states based on signal perturbations and change rates. The temporal logic channel uses the state evolution path as a clue to infer future state trends.Each sub-detector outputs an independent state judgment result based on its internally trained feature selection path and weight evaluation criteria. This result is typically expressed as a binary value, indicating whether the connection is currently active. The outputs of these three channels are independent of each other, but a unified fusion mechanism is used at the decision-making level to achieve a comprehensive judgment. The judgment results of the three sub-detectors are then fused together for decision-making. This fusion mechanism uses a weighted voting strategy, with initial weights set based on parameters such as accuracy, stability, and information distribution of each channel during historical evaluation. Each judgment result is multiplied by its weight and summed, and all results are combined. The final target connection state judgment result is confirmed by a positive or negative sign, and the absolute value of the fusion score is used as a judgment strength indicator. To assess the confidence of the judgment, the current weighted fusion result is compared with the maximum weight in the weight set, resulting in a state judgment reliability value ranging from 0 to 1. If the ratio is high, it means that the judgment result is highly consistent with a certain information channel, with strong tendency and consistency, and the system considers that the current state recognition result is reliable; on the contrary, if the ratio is low, it means that there are large differences between the channels. The system marks the current state as a low reliability result and triggers additional judgment or prompts the upper-level system for processing when necessary.

[0037] In a specific embodiment, the step of inputting the first feature vector, the second feature vector, and the temporal logic feature vector into corresponding weighted decision tree sub-detectors for state judgment to obtain state judgment results for three independent channels may specifically include the following steps: Performing linear weighting and feature rearrangement on the first eigenvector to obtain a low-frequency channel feature importance sequence, and sorting the first eigenvector according to the low-frequency channel feature importance sequence to obtain a first weighted vector; Input the first weighted vector into the weighted decision tree sub-detector corresponding to the low-frequency channel to perform a four-layer deep decision tree splitting calculation to obtain a state judgment result of the low-frequency channel; Performing feature extraction and nonlinear feature combination on the second eigenvector to obtain a second weighted vector; Input the second weighted vector into the weighted decision tree sub-detector corresponding to the high-frequency channel to perform three-layer deep decision making to obtain a state judgment result of the high-frequency channel; The state transition probability matrix is ​​constructed from the temporal logic feature vector, and the conditional probability distribution of the next state is calculated based on the Markov transition characteristics to obtain the temporal state prediction model; Based on the time series state prediction model, a weighted time series decision tree is constructed to regularize the consistency and conversion characteristics of the previous and next states to obtain the state judgment result of the time series channel; The state judgment result of the low-frequency channel, the state judgment result of the high-frequency channel, and the state judgment result of the timing channel are taken as the state judgment results of the three independent channels.

[0038] Specifically, for the first eigenvector, the feature set used by the low-frequency channel, this vector originally consists of key indicators reflecting the macroscopic characteristics of the signal, such as the low-frequency approximation coefficient, the stable region level value, the level standard deviation, and the state duration. To explore the relative contribution of these features to the current judgment task, linear weighting and feature reordering are performed. A set of linear weighting coefficients is constructed based on the correlation coefficients between each low-frequency feature and the actual state label during the historical training phase. Each dimension in the original eigenvector is then multiplied by the corresponding weight to obtain a feature weight value. The weighted results are then sorted according to their numerical values. The sorted results reflect the order of importance of each dimension of the current input sample to the judgment result, thereby generating a feature importance sequence for the low-frequency channel. Based on this feature importance sequence, the first eigenvector is reordered according to this feature importance sequence, with high-weight features placed first and low-weight features placed last, forming a first weighted vector with an orderly structure and enhanced semantics. The first weighted vector is input into the weighted decision tree sub-detector corresponding to the low-frequency channel. This detector employs a four-layer deep tree structure for state inference. The first layer performs preliminary screening to determine whether the level exceeds a certain threshold range. The second layer determines signal stability based on the strength of the low-frequency trend. The third layer combines standard deviation and stabilization time to identify typical connection states. The fourth layer integrates the results of the previous layers' judgment paths and outputs the final judgment conclusion. This four-layer split structure implements hierarchical discrimination logic for low-frequency features through path combination, outputting the low-frequency channel's state judgment result. This structure exhibits excellent stability and anti-interference capabilities, making it suitable for connection judgment scenarios in stable states. Simultaneously, nonlinear feature extraction and combination operations are performed on the second feature vector, representing the feature set of the high-frequency channel. Because the second feature vector contains highly sensitive and rapidly changing parameters such as high-frequency detail coefficients, level change rates, and jitter index, directly inputting them into the decision model in a linear form would not adequately represent its actual discrimination capabilities under complex perturbations. Therefore, multiple high-frequency features are cross-combined, for example, by constructing nonlinear combination features using square terms, difference terms, and sliding window local change rates to improve adaptability to abnormal signal patterns. The nonlinearly expanded features are then normalized to prevent scale differences from affecting the decision tree splitting criteria. A second weighted vector is constructed to reflect the behavioral characteristics of the current signal in the instantaneous disturbance dimension. The second weighted vector is input into the weighted decision tree sub-detector corresponding to the high-frequency channel. Considering the high sensitivity and easy overfitting of high-frequency features, the decision tree depth of the high-frequency channel is set to three layers. The first layer mainly identifies whether the jitter index exceeds the threshold range, the second layer determines whether there is a large jump in the rate of change, and the third layer performs judgment operations on the feature subspace of multiple high-frequency sub-feature combinations to comprehensively obtain the state recognition results. The state judgment results of the high-frequency channel can quickly respond to abnormal disturbances in the HPD signal and are suitable for identifying complex dynamic scenarios such as poor cable contact, temporary disconnection, and boundary jitter.In the temporal logic channel, the system constructs a state transition probability model based on the combined behavior of historical and current states. By statistically analyzing the transition paths between critical state judgments and the final confirmed state at several past moments, a state transition probability matrix is ​​constructed. In this matrix, each cell records the probability of a state transitioning to the target state under a given previous state, forming a state transition structure based on Markov process theory. Using this structure, the system predicts the conditional probability distribution of the next state based on the combined information of the current and previous states, forming a sequence-based temporal state prediction model. This model can still make reasonable predictions based on state evolution patterns even when level signal boundaries are unclear, thus compensating for the lack of expressiveness inherent in the level signal itself. The temporal state prediction model is converted into an executable temporal decision structure. By embedding the state transition probability matrix into a weighted temporal decision tree, a state reasoning model with the ability to correlate previous and subsequent states is formed. In this model, nodes not only make judgments based on the characteristics of the current state but also combine the duration of the previous state, transition frequency, and probability score to regularize the validity of the state. The state judgment results output by the above three channels are aggregated. These results come from the structural stability judgment of low-frequency characteristics, the disturbance response judgment of high-frequency characteristics, and the evolution path judgment of timing logic, forming the state judgment results of three independent channels. These results complement each other in the data dimension and form complementary characteristics in the decision logic: the low-frequency channel provides basic stability judgment, the high-frequency channel provides sensitive and rapid response, and the timing channel provides overall evolution trend.

[0039] The above describes the DP interface connection status detection method in the embodiment of the present application. The following describes the DP interface connection status detection system 10 in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of the DP interface connection status detection system 10 includes: The acquisition module 11 is used to collect the historical level signal of the HPD pin in the DP interface and pre-process the historical level signal to obtain a level pre-processing signal and level fluctuation characteristic data; The processing module 12 is used to perform time domain decomposition processing on the level preprocessing signal and the level fluctuation characteristic data to obtain high and low level stable area identification results and stable area duration data; A calculation module 13 is used to calculate a high-level threshold and a low-level threshold using the high- and low-level stable region identification results and the stable region duration data to obtain a level dual-threshold detection parameter; The status detection module 14 is used to perform status detection on the real-time level signal of the HPD pin based on the level dual threshold detection parameters to obtain the target connection status judgment result and status judgment reliability value of the DP interface.

[0040] Through the collaborative efforts of the aforementioned components, a dual-threshold detection mechanism with high and low levels, combined with hysteresis, effectively addresses the frequent state switching caused by HPD signal fluctuations around a single threshold, improving the stability and anti-interference capabilities of DP interface connection status detection. Time-domain adaptive decomposition processing technology accurately identifies the stable and transitional regions of the HPD signal, enabling precise characterization of HPD signal characteristics. Differentiated time constraints are implemented, allowing different time thresholds to be set for connected and disconnected states, thus balancing the conflicting requirements of connection judgment accuracy and disconnection response speed. By analyzing HPD signal jitter characteristics and calculating dynamic parameters, a signal jitter index evaluation model is established, enabling accurate judgment of critical connection states and significantly reducing the false positive rate. A three-way parallel multi-channel hybrid detection network focuses on the HPD signal's low-frequency main characteristics, high-frequency variation characteristics, and sequential logic characteristics, providing comprehensive signal characteristic analysis and enhancing the comprehensiveness and robustness of the system's judgment. An adaptive parameter adjustment mechanism dynamically optimizes judgment parameters based on actual signal characteristics, adapting to HPD signal fluctuations in different environments and improving the detection method's environmental adaptability. A reliability index evaluation mechanism is introduced to monitor the reliability of status judgment results in real time, promptly identify potential judgment risks, and provide trigger conditions for additional detection processes, thus ensuring the reliable operation of the entire system.

[0041] See also Figure 3 , Figure 3 This is a schematic block diagram of the structure of a computer device 300 provided in an embodiment of the present application. The computer device 300 includes a processor 301 and a memory 302. The processor 301 and the memory 302 are connected via a device bus 303, wherein the memory 302 may include a non-volatile storage medium and an internal memory.

[0042] The non-volatile storage medium may store a computer program. The computer program includes program instructions, and when the program instructions are executed by the processor 301, the processor 301 may execute any of the above-mentioned DP interface connection status detection methods.

[0043] The processor 301 is used to provide computing and control capabilities and support the operation of the entire computer device 300.

[0044] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor 301, the processor 301 can execute any of the above-mentioned DP interface connection status detection methods.

[0045] Those skilled in the art will understand that Figure 3The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 300 involved in the solution of the present application. The specific computer device 300 may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0046] It should be understood that the processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0047] It should be noted that, those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the computer device 300 described above can refer to the corresponding process of the aforementioned DP interface connection status detection method, and will not be repeated here.

[0048] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0049] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program code.

[0050] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for detecting the connection status of a DP interface, characterized in that: include: Collecting historical level signals of the HPD pin in the DP interface and preprocessing the historical level signals to obtain level preprocessing signals and level fluctuation characteristic data; Performing time domain decomposition processing on the level preprocessing signal and the level fluctuation characteristic data to obtain high and low level stable area identification results and stable area duration data; Calculating a high level threshold and a low level threshold using the high and low level stable area identification result and the stable area duration data to obtain a level dual threshold detection parameter; The real-time level signal of the HPD pin is detected based on the level dual threshold detection parameter to obtain the target connection state judgment result and the state judgment reliability value of the DP interface.

2. The DP interface connection status detection method according to claim 1, characterized in that: The collecting of historical level signals of the HPD pin in the DP interface and preprocessing of the historical level signals to obtain level preprocessing signals and level fluctuation characteristic data include: Continuously sampling the analog level signal of the HPD pin in the DP interface to obtain a historical level signal, and digitizing the historical level signal to obtain a digitized level signal; Performing median filtering on the digitized level signal to obtain a preliminary filtered signal, and performing wavelet threshold denoising on the preliminary filtered signal to obtain a level preprocessed signal; Calculating a mean value, a standard deviation, a kurtosis, and a skewness statistical value of the HPD pin level signal according to the level preprocessing signal, and combining the mean value, the standard deviation, the kurtosis, and the skewness statistical value into a level statistical feature; The level statistical feature and the fluctuation parameter of the level preprocessing signal are combined to obtain level fluctuation feature data.

3. The method for detecting the connection status of a DP interface according to claim 1, wherein: The performing time domain decomposition processing on the level preprocessing signal and the level fluctuation characteristic data to obtain high and low level stable area identification results and stable area duration data includes: Performing three-layer discrete wavelet transform processing on the level preprocessed signal to obtain a low-frequency approximation coefficient set and a high-frequency detail coefficient set; Calculating an energy distribution weight value based on the low-frequency approximation coefficient set and the high-frequency detail coefficient set to obtain a level signal stability evaluation index, and comparing the level signal stability evaluation index with a preset threshold to obtain a preliminary stable area marking result; Classifying the level values ​​in the preliminary stable area marking result, marking an area greater than a first target value as a high-level stable area, and marking an area less than a second target value as a low-level stable area, to obtain a level partition marking result; Performing minimum duration constraint filtering on the level partition marking result to obtain high and low level stable area identification results; The start time, end time and duration of each stable area are calculated according to the high and low level stable area identification results to obtain stable area duration data.

4. The method for detecting the connection status of a DP interface according to claim 1, wherein: The step of calculating a high level threshold and a low level threshold using the high and low level stable region identification result and the stable region duration data to obtain a level dual threshold detection parameter includes: Extracting an average level value set of a high-level stable area and an average level value set of a low-level stable area from the high- and low-level stable area identification results to obtain stable area level value distribution data; Calculating respective standard deviation sets based on the level signals of the high-level stable area and the low-level stable area to obtain stable area fluctuation amplitude data; Obtaining a high level threshold value based on a minimum high level value of the stable area level value distribution data minus a first preset multiple of the maximum high level standard deviation value of the stable area fluctuation amplitude data; A low level threshold is obtained based on the maximum low level value of the stable area level value distribution data plus the maximum low level standard deviation value of the stable area fluctuation amplitude data of a second preset multiple; According to the stable zone duration data, the time threshold for connection state judgment is set to M milliseconds, and the time threshold for disconnection state judgment is set to N milliseconds to obtain a dual-threshold time constraint parameter; The high level threshold, the low level threshold and the dual-threshold time constraint parameter are integrated to obtain a level dual-threshold detection parameter.

5. The method for detecting the connection status of a DP interface according to claim 1, wherein: The performing state detection on the real-time level signal of the HPD pin based on the level dual-threshold detection parameter to obtain the target connection state judgment result and the state judgment reliability value of the DP interface includes: Collecting and preprocessing the real-time level signal of the HPD pin to obtain a real-time level signal; Comparing the real-time level signal with the high level threshold and the low level threshold in the level dual-threshold detection parameter to obtain a level state comparison result; applying a state transition rule according to the level state comparison result, determining that a connection state transition occurs when the real-time level signal rises from below a low level threshold to above a high level threshold, and determining that a disconnection state transition occurs when the real-time level signal falls from above a high level threshold to below a low level threshold, thereby obtaining a preliminary state transition determination result; Applying the dual-threshold time constraint parameter in the level dual-threshold detection parameter to the preliminary state transition determination result to perform time verification, applying a time threshold of M milliseconds to the connected state and a time threshold of N milliseconds to the disconnected state, to obtain a time constraint verification result; Applying a state holding mechanism to a region where the real-time level signal is between a low level threshold and a high level threshold, keeping the previous state judgment unchanged, and obtaining a state holding processing result; Generate a state judgment sequence according to the time constraint verification result and the state preservation processing result to obtain an initial connection state judgment result; performing a critical state analysis based on the initial connection state judgment result and the stable zone duration data to obtain a critical connection state judgment result; Multi-channel detection is performed based on the critical connection state judgment result and the real-time level signal to obtain the target connection state judgment result and the state judgment reliability value of the DP interface.

6. The method for detecting the connection status of a DP interface according to claim 5, wherein: The performing critical state analysis in combination with the initial connection state judgment result and the stable zone duration data to obtain a critical connection state judgment result includes: Calculating a stable region jitter index and a transition region jitter index based on the real-time level signal of the HPD pin, and using the stable region jitter index and the transition region jitter index as signal jitter characteristic parameters; Constructing a critical connection state evaluation function, and calculating an evaluation function value based on the critical connection state evaluation function, comparing the evaluation function value with a preset evaluation threshold, determining a connected state when the value is greater than the preset evaluation threshold, and determining a disconnected state when the value is less than the preset evaluation threshold, to obtain an evaluation function determination result; Constructing a state transition decision tree with three-layer judgment logic according to the initial connection state judgment result and the signal jitter characteristic parameter, and generating a decision tree judgment result through the state transition decision tree; Counting the number of consecutive state transitions of the real-time level signal of the HPD pin, and automatically adjusting the dual-threshold interval and the time threshold when the interval between E consecutive state transitions is less than the minimum stable time in the stable zone duration data, to obtain a target judgment parameter; A critical connection state judgment result is generated according to the evaluation function judgment result, the decision tree judgment result and the target judgment parameter.

7. The method for detecting the connection status of a DP interface according to claim 6, wherein: The performing multi-channel detection based on the critical connection state judgment result and the real-time level signal to obtain the target connection state judgment result and the state judgment reliability value of the DP interface includes: Extracting a low-frequency approximation coefficient, a stable region level value, a standard deviation, and a stable time according to the real-time level signal, and constructing a first eigenvector according to the low-frequency approximation coefficient, the stable region level value, the standard deviation, and the stable time; Extracting a high-frequency detail coefficient set, a level change rate, and a jitter index according to the real-time level signal, and constructing a second eigenvector according to the high-frequency detail coefficient set, the level change rate, and the jitter index; Constructing a temporal logic feature vector according to the critical connection state judgment result; Inputting the first feature vector, the second feature vector, and the temporal logic feature vector into corresponding weighted decision tree sub-detectors for state judgment, respectively, to obtain state judgment results of three independent channels; The state judgment results of the three independent channels are fused to obtain the target connection state judgment result of the DP interface, and the ratio of the decision confidence of the target connection state judgment result to the maximum weight is calculated to obtain the state judgment reliability value.

8. The method for detecting the connection status of a DP interface according to claim 7, wherein: The first feature vector, the second feature vector, and the temporal logic feature vector are respectively input into corresponding weighted decision tree sub-detectors for state judgment to obtain state judgment results of three independent channels, including: Performing linear weighting and feature rearrangement on the first eigenvectors to obtain a low-frequency channel feature importance sequence, and sorting the first eigenvectors according to the low-frequency channel feature importance sequence to obtain a first weighted vector; Inputting the first weighted vector into the weighted decision tree sub-detector corresponding to the low-frequency channel to perform a four-layer deep decision tree splitting calculation to obtain a state judgment result of the low-frequency channel; Performing feature extraction and nonlinear feature combination on the second feature vector to obtain a second weighted vector; Inputting the second weighted vector into the weighted decision tree sub-detector corresponding to the high-frequency channel to perform three-layer deep decision making to obtain a state judgment result of the high-frequency channel; The temporal logic feature vector is used to construct a state transition probability matrix, and the conditional probability distribution of the next state is calculated according to the Markov transition characteristic to obtain a temporal state prediction model; Based on the time series state prediction model, a time series decision tree with weights is constructed, and the consistency and conversion characteristics of the previous and subsequent states are regularized to obtain the state judgment result of the time series channel; The state judgment result of the low-frequency channel, the state judgment result of the high-frequency channel, and the state judgment result of the timing channel are used as the state judgment results of three independent channels.

9. A DP interface connection status detection system, characterized in that: Used to execute the DP interface connection status detection method according to any one of claims 1 to 8, the DP interface connection status detection system includes: An acquisition module is used to collect historical level signals of the HPD pin in the DP interface and preprocess the historical level signals to obtain level preprocessing signals and level fluctuation characteristic data; a processing module, configured to perform time domain decomposition processing on the level preprocessing signal and the level fluctuation characteristic data to obtain high and low level stable area identification results and stable area duration data; a calculation module, configured to calculate a high-level threshold and a low-level threshold using the high- and low-level stable region identification result and the stable region duration data, to obtain a level dual-threshold detection parameter; The status detection module is used to perform status detection on the real-time level signal of the HPD pin based on the level dual threshold detection parameter to obtain the target connection status judgment result and status judgment reliability value of the DP interface.

10. A computer device, characterized in that: The computer device includes: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the computer device to execute the DP interface connection status detection method according to any one of claims 1 to 8.

Images