Real-time monitoring and anomaly detection method for AI independent keys in all-in-one product

Abstract

The invention relates to the technical field of electronic information, and provides a real-time monitoring and anomaly detection method for an AI independent key in an all-in-one machine product, which comprises the following steps of: firstly, acquiring a to-be-detected all-in-one machine, identifying the AI independent key in the all-in-one machine, monitoring a key trigger signal, inquiring a signal feature code, performing spectrum analysis to obtain a feature spectrum sequence, and extracting a spectrum key value; detecting a key response track corresponding to the frequency spectrum key value, performing dynamic regularization analysis to obtain a regularized track sequence, identifying track sequence parameters, calculating track response values, collecting the change trend of the track response values, determining a fluctuation amplitude value to determine an adjustment mode, identifying a key state, comparing state data to obtain state comparison parameters, and performing dynamic regularization analysis to obtain a state comparison result; and inquiring potential abnormal factors, determining an abnormal level and an analysis index, constructing an abnormal detection signal and generating an abnormal detection scheme. The key use stability of the all-in-one machine product can be improved.

Description

Technical Field

[0001] The present invention relates to a method and a device for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product, and belongs to the field of electronic information technology. Background Art

[0002] Acquire the all-in-one device to be detected, identify the AI ​​independent key in the all-in-one device to be detected, monitor the key trigger signal corresponding to the AI ​​independent key, query the signal feature code corresponding to the key trigger signal, perform spectrum analysis on the signal feature code to obtain a feature spectrum sequence, and extract spectrum key values ​​in the feature spectrum sequence;

[0003] Detecting the key response trajectory corresponding to the spectrum key value, performing dynamic regularization analysis on the key response trajectory to obtain a regularized trajectory sequence, identifying trajectory sequence parameters in the regularized trajectory sequence, and calculating the trajectory response value corresponding to the trajectory sequence parameters;

[0004] Collecting a change trend corresponding to the trajectory response value, identifying a change interval of the change trend within a specific time period, calculating a fluctuation amplitude value corresponding to the change trend based on the change interval, and determining an adjustment mode corresponding to the all-in-one device to be detected based on the fluctuation amplitude value;

[0005] Identify the key state corresponding to the AI ​​independent key in the adjustment mode, monitor the state data corresponding to the key state in real time, compare the state data with the preset state, and obtain the state comparison parameter;

[0006] Query the potential abnormal factors in the state comparison parameters, determine the abnormal level corresponding to the potential abnormal factors, analyze the abnormal level index corresponding to the abnormal level, construct the abnormal detection signal corresponding to the all-in-one machine to be detected based on the abnormal level index, and generate the abnormal detection scheme corresponding to the AI ​​independent button in the all-in-one machine to be detected based on the abnormal detection signal. Summary of the invention

[0007] The present invention provides a method and system for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product, the main purpose of which is to improve the stability of button use of the all-in-one product.

[0008] To achieve the above-mentioned purpose, the present invention provides a method for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product, comprising:

[0009] Acquire the all-in-one device to be detected, identify the AI ​​independent key in the all-in-one device to be detected, monitor the key trigger signal corresponding to the AI ​​independent key, query the signal feature code corresponding to the key trigger signal, perform spectrum analysis on the signal feature code to obtain a feature spectrum sequence, and extract spectrum key values ​​in the feature spectrum sequence;

[0010] Detecting the key response trajectory corresponding to the spectrum key value, performing dynamic regularization analysis on the key response trajectory to obtain a regularized trajectory sequence, identifying trajectory sequence parameters in the regularized trajectory sequence, and calculating the trajectory response value corresponding to the trajectory sequence parameters;

[0011] Collecting a change trend corresponding to the trajectory response value, identifying a change interval of the change trend within a specific time period, calculating a fluctuation amplitude value corresponding to the change trend based on the change interval, and determining an adjustment mode corresponding to the all-in-one device to be detected based on the fluctuation amplitude value;

[0012] Identify the key state corresponding to the AI ​​independent key in the adjustment mode, monitor the state data corresponding to the key state in real time, compare the state data with the preset state, and obtain the state comparison parameter;

[0013] Query the potential abnormal factors in the state comparison parameters, determine the abnormal level corresponding to the potential abnormal factors, analyze the abnormal level index corresponding to the abnormal level, construct the abnormal detection signal corresponding to the all-in-one machine to be detected based on the abnormal level index, and generate the abnormal detection scheme corresponding to the AI ​​independent button in the all-in-one machine to be detected based on the abnormal detection signal.

[0014] Optionally, performing spectrum analysis on the signal characteristic code to obtain a characteristic spectrum sequence includes:

[0015] Performing denoising processing on the signal characteristic code to obtain a denoised signal characteristic code;

[0016] Querying the time domain signal corresponding to the denoised signal feature code;

[0017] Converting the time domain signal into a frequency domain signal;

[0018] Extracting main frequency components from the frequency domain signal;

[0019] Normalizing the amplitude of the main frequency component to obtain a normalized frequency feature;

[0020] Performing spectrum analysis on the normalized frequency features to obtain a characteristic spectrum sequence.

[0021] Optionally, the dynamically regularizing and analyzing the key response trajectory to obtain a regularized trajectory sequence includes:

[0022] Performing trajectory preprocessing on the key response trajectory to obtain preprocessed trajectory data;

[0023] Locating a trajectory time axis in the preprocessed trajectory data;

[0024] Extracting the time distance parameter corresponding to the trajectory time axis;

[0025] Using the time distance parameter, constructing a trajectory distance matrix corresponding to the key response trajectory;

[0026] Searching for the best matching path between different trajectories in the distance matrix;

[0027] Based on the optimal matching path, the key response trajectory is dynamically regularized and analyzed to obtain a regular trajectory sequence.

[0028] Optionally, calculating the trajectory response value corresponding to the trajectory sequence parameter includes:

[0029] The trajectory response value corresponding to the trajectory sequence parameters is calculated using the following formula:

[0030]

[0031] Where R represents the trajectory response value corresponding to the trajectory sequence parameter, N represents the total number of sampling points in the time interval corresponding to the trajectory sequence parameter, i represents the sampling point index in the time interval corresponding to the trajectory sequence parameter, and A(t i ) represents the response amplitude of the i-th sampling point at time t, F(t i ) represents the trigger frequency of the i-th sampling point at time t, m represents the number of dimensions of the trajectory sequence parameter corresponding to the position parameter, j represents the dimension index of the trajectory sequence parameter corresponding to the position parameter, p j (t i ) represents the value of the i-th sampling point at time t in the j dimension, Represents the average value of the j dimension in the time interval.

[0032] Optionally, the calculating, based on the change interval, a fluctuation amplitude value corresponding to the change trend includes:

[0033] The fluctuation amplitude value corresponding to the change trend is calculated using the following formula:

[0034]

[0035] Among them, FB represents the fluctuation amplitude value corresponding to the change trend, V max Indicates the maximum value of the interval corresponding to the change interval, V min represents the minimum value of the interval corresponding to the change interval, Q represents the total number of sampling times within the change trend, p represents the sampling index within the change trend, R(t' p ) represents the trend response value of the pth sampling at time t', It represents the average value of the trend response of all sampling points within the change trend.

[0036] Optionally, determining the adjustment mode corresponding to the to-be-detected all-in-one device based on the fluctuation amplitude value includes:

[0037] Analyze the range of the fluctuation amplitude value;

[0038] Based on the size range, the to-be-detected all-in-one device is stably classified to obtain a preliminary classification level;

[0039] Determining a stability index corresponding to the preliminary classification level;

[0040] Querying the adjustment factor corresponding to the stability index;

[0041] Based on the adjustment factor, an adjustment mode corresponding to the all-in-one device to be detected is determined.

[0042] Optionally, comparing the state data with a preset state to obtain a state comparison parameter includes:

[0043] identifying key status features in the status data;

[0044] Determine a difference metric between the key state feature and the feature in the preset state;

[0045] Calculate the difference between the state data and the preset state according to the difference metric standard;

[0046] Performing evaluation and analysis on the difference value to obtain an evaluation and analysis result;

[0047] Based on the preliminary evaluation result, the state data is compared with a preset state to obtain a state comparison parameter.

[0048] Optionally, the querying of potential abnormal factors in the state comparison parameters includes:

[0049] Analyzing parameter constituent elements corresponding to the state comparison parameters;

[0050] Based on the parameter constituent elements, query the state comparison parameters for abnormal characteristic indicators;

[0051] According to the characteristic index, the state comparison parameter is screened to obtain a screened parameter item;

[0052] Identifying abnormal correlation items between the screening parameter items;

[0053] Query the potential abnormal factor corresponding to the abnormal association item.

[0054] Optionally, constructing an abnormality detection signal corresponding to the to-be-detected integrated device based on the abnormality level indicator includes:

[0055] Query the indicator attribute corresponding to the abnormal level indicator;

[0056] Analyze the key attribute elements corresponding to the indicator attributes;

[0057] According to the key attribute elements, design a signal trigger rule corresponding to the all-in-one device to be detected;

[0058] Optimizing and adjusting the signal triggering rule to obtain an optimized signal rule;

[0059] Extracting anomaly detection entries in the optimized signal rule;

[0060] Based on the abnormality detection entry, an abnormality detection signal corresponding to the all-in-one device to be detected is constructed.

[0061] Optionally, in order to solve the above problem, the present invention provides an AI independent button real-time monitoring and abnormality detection system in an all-in-one product, the system comprising:

[0062] A key value extraction module is used to obtain the integrated machine to be detected, identify the AI ​​independent key in the integrated machine to be detected, monitor the key trigger signal corresponding to the AI ​​independent key, query the signal feature code corresponding to the key trigger signal, perform spectrum analysis on the signal feature code to obtain a feature spectrum sequence, and extract the spectrum key value in the feature spectrum sequence;

[0063] a response value calculation module, used for detecting the key response trajectory corresponding to the spectrum key value, performing dynamic regularization analysis on the key response trajectory to obtain a regularized trajectory sequence, identifying trajectory sequence parameters in the regularized trajectory sequence, and calculating the trajectory response value corresponding to the trajectory sequence parameters;

[0064] A mode adjustment module, used for collecting the change trend corresponding to the trajectory response value, identifying the change interval of the change trend within a specific time period, calculating the fluctuation amplitude value corresponding to the change trend based on the change interval, and determining the adjustment mode corresponding to the all-in-one device to be detected based on the fluctuation amplitude value;

[0065] A state comparison module, used to identify the key state corresponding to the AI ​​independent key in the adjustment mode, monitor the state data corresponding to the key state in real time, compare the state data with the preset state, and obtain a state comparison parameter;

[0066] A scheme generation module is used to query the potential abnormal factors in the state comparison parameters, determine the abnormal level corresponding to the potential abnormal factors, analyze the abnormal level index corresponding to the abnormal level, construct an abnormal detection signal corresponding to the all-in-one machine to be detected based on the abnormal level index, and generate an abnormal detection scheme corresponding to the AI ​​independent button in the all-in-one machine to be detected based on the abnormal detection signal.

[0067] First of all, the present invention obtains the all-in-one machine to be detected and identifies the AI ​​independent button in the all-in-one machine to be detected, so as to concentrate on the precise analysis of the specific device and avoid unnecessary waste of resources and detection errors. Accurately identifying the AI ​​independent button is like finding the key entry point for the subsequent detection work. This process makes the detection more targeted and can quickly focus on the buttons that play an important role in starting the intelligent functions of the all-in-one machine. At the same time, the present invention can deeply understand the entire response process after the user presses the AI ​​independent button by detecting the button response trajectory corresponding to the spectrum key value, which is helpful to accurately evaluate the performance and stability of the button and timely discover the delays, freezes or abnormal fluctuations that may occur in the response process. The present invention collects the change trend corresponding to the trajectory response value and identifies the change range of the change trend within a specific time period, which is helpful to timely discover potential performance degradation or instability factors, so as to take corresponding maintenance or adjustment measures before the problem becomes serious, and is beneficial In order to predict the future situation of key response, a model can be established to predict the changing trend of trajectory response value in the future period of time, and provide a reference basis for equipment maintenance plan and resource allocation. The present invention can timely understand the working condition of the key by identifying the key state corresponding to the AI ​​independent key in the adjustment mode and monitoring the state data corresponding to the key state in real time, so as to ensure its normal operation in the adjustment mode, and can find some potential problem trends, such as the key response gradually slowing down, intermittent failures, etc., and perform maintenance and replacement in advance to improve the reliability and service life of the all-in-one machine. Furthermore, the present invention can find possible problems with the key in advance by querying the potential abnormal factors in the state comparison parameters, and can take corresponding preventive measures before the problem is obviously manifested, so as to avoid the key failure from affecting the normal use of the all-in-one machine, and can quickly locate the all-in-one machine that may have problems by monitoring and analyzing the state comparison parameters, arrange maintenance plans in advance, and improve management efficiency. Therefore, the real-time monitoring and abnormality detection method and system of AI independent keys in the all-in-one machine product proposed by the present invention can improve the key use stability of the all-in-one machine product. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 A flowchart of a method for real-time monitoring and abnormality detection of independent AI buttons in an all-in-one product provided by an embodiment of the present invention;

[0069] Figure 2 A schematic diagram of the structure of an electronic device for implementing the method for real-time monitoring and abnormality detection of independent AI buttons in the all-in-one product provided by one embodiment of the present invention.

[0070] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0071] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0072] The embodiment of the present application provides a method for real-time monitoring and abnormality detection of independent AI buttons in an all-in-one product. The execution subject of the method for real-time monitoring and abnormality detection of independent AI buttons in an all-in-one product includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided in the embodiment of the present application. In other words, the method for real-time monitoring and abnormality detection of independent AI buttons in an all-in-one product can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0073] Embodiment 1:

[0074] Reference Figure 1 FIG. 1 is a flow chart of a method for real-time monitoring and abnormality detection of independent AI buttons in an all-in-one product provided by an embodiment of the present invention. In this embodiment, the method for real-time monitoring and abnormality detection of independent AI buttons in an all-in-one product includes:

[0075] S1. Obtain an all-in-one device to be detected, identify an AI independent key in the all-in-one device to be detected, monitor a key trigger signal corresponding to the AI ​​independent key, query a signal feature code corresponding to the key trigger signal, perform spectrum analysis on the signal feature code, obtain a feature spectrum sequence, and extract spectrum key values ​​in the feature spectrum sequence.

[0076] The present invention obtains the all-in-one machine to be detected and identifies the AI ​​independent buttons in the all-in-one machine to be detected, so as to concentrate on the accurate analysis of specific devices and avoid unnecessary waste of resources and detection errors. Accurately identifying the AI ​​independent buttons is like finding the key entry point for subsequent detection work. This process makes the detection more targeted and can quickly focus on the buttons that play an important role in starting the intelligent functions of the all-in-one machine.

[0077] Among them, the all-in-one machine to be tested refers to an integrated electronic device that is about to undergo comprehensive testing and evaluation, which can be widely used in multiple fields such as office, education, entertainment, etc., and has the characteristics of high integration and small space occupation; the AI ​​independent button refers to a button on the all-in-one product that is specifically used to quickly start a specific intelligent function, which is different from traditional mechanical buttons and capacitive touch buttons, etc. Optionally, the identification of the AI ​​independent button in the all-in-one machine to be tested can be achieved through a hardware detection tool, such as: using a special electronic equipment detection tool to connect to the interface of the all-in-one machine to be tested, read the hardware information of the device, and thus identify the AI ​​independent button.

[0078] Furthermore, the present invention can understand the interaction between the user and the AI ​​independent key in real time by monitoring the key trigger signal corresponding to the AI ​​independent key and querying the signal feature code corresponding to the key trigger signal, ensuring that each key operation can be accurately captured, which helps to respond to user needs in a timely manner and improve the efficiency and convenience of the all-in-one machine.

[0079] Among them, the key trigger signal refers to the electrical signal or other detectable physical signal change generated when the user presses the AI ​​independent key on the all-in-one machine to be detected; the signal feature code refers to the code used to uniquely identify the specific attributes of the key trigger signal, which may include frequency characteristics, amplitude characteristics, time characteristics and other information of the key trigger signal. Optionally, the monitoring of the key trigger signal corresponding to the AI ​​independent key can be achieved through signal analysis tools, such as LabVIEW, MATLAB and other tools; the query of the signal feature code corresponding to the key trigger signal can be achieved through a convolutional neural network, such as using a convolutional neural network to process the image or waveform of the key trigger signal, automatically learn the characteristics of the signal and predict the feature code.

[0080] Furthermore, the present invention performs spectral analysis on the signal feature code to obtain a characteristic spectrum sequence, which helps to discover potential frequency interference or anomalies, can provide richer information for subsequent analysis and detection, and makes the performance evaluation of AI independent keys more comprehensive and accurate. Key performance indicators such as the response speed and stability of the keys can be judged based on specific frequency characteristics in the characteristic spectrum sequence.

[0081] The characteristic spectrum sequence refers to an ordered set of a series of spectrum features obtained by performing spectrum analysis on normalized frequency features, which includes the distribution and characteristic information of the signal at different frequencies.

[0082] As an embodiment of the present invention, the spectral analysis of the signal characteristic code to obtain a characteristic spectrum sequence includes: denoising the signal characteristic code to obtain a denoised signal characteristic code; querying a time domain signal corresponding to the denoised signal characteristic code; converting the time domain signal into a frequency domain signal; extracting the main frequency components in the frequency domain signal; normalizing the amplitude of the main frequency components to obtain a normalized frequency feature; and spectral analysis of the normalized frequency feature to obtain a characteristic spectrum sequence.

[0083] Among them, the denoised signal feature code refers to the signal feature code obtained after denoising the original signal feature code; the time domain signal refers to the signal that changes with time corresponding to the denoised signal feature code, which reflects the characteristics and changes of the signal in the time dimension; the main frequency component refers to the frequency component that occupies a dominant position in the frequency domain signal and has an important influence on the signal characteristics, usually the frequency part with large energy or specific meaning; the normalized frequency feature refers to the frequency feature obtained after amplitude normalization processing of the main frequency component, so that different frequency components are comparable in amplitude.

[0084] Furthermore, the denoising of the signal feature code can be achieved through a wavelet transform method, such as: using wavelet transform to perform multi-scale analysis on the signal to remove noise; the query of the time domain signal corresponding to the denoised signal feature code can be achieved through a signal reconstruction method, such as: based on the generation process of the denoised signal feature code, the time domain signal is reversely derived; the conversion of the time domain signal into a frequency domain signal can be achieved through a Fourier transform method, such as: fast Fourier transform, discrete cosine transform and other methods; the extraction of the main frequency components in the frequency domain signal can be achieved through a peak detection method, such as: looking for peak points with larger amplitudes in the frequency domain signal, the frequency components corresponding to these peak points are usually the main frequency components; the amplitude normalization of the main frequency components can be achieved through a normalization method, such as: maximum and minimum normalization, mean variance normalization; the spectrum analysis of the normalized frequency characteristics can be achieved through high-order spectrum analysis, such as: bispectral analysis, which can reveal the nonlinear characteristics of the signal.

[0085] The present invention helps to reduce the amount of data for subsequent analysis, improve processing efficiency, and avoid unnecessary waste of computing resources by extracting spectrum key values ​​from the characteristic spectrum sequence. Secondly, spectrum key values ​​can more intuitively reflect the core characteristics of the signal and can quickly grasp the essential attributes of the signal.

[0086] Among them, the spectral key value refers to a specific numerical value extracted from the characteristic spectrum sequence that can reflect the core characteristics and important attributes of the signal, which may be a key parameter such as the peak frequency in the characteristic spectrum sequence, the amplitude of a specific frequency component, the center of gravity of the frequency distribution, etc. Optionally, the extraction of the spectral key value in the characteristic spectrum sequence can be achieved through a principal component analysis method, such as: treating the characteristic spectrum sequence as high-dimensional data, and finding the main change direction of the data, i.e., the principal component, through principal component analysis.

[0087] S2. Detect the key response trajectory corresponding to the spectrum key value, perform dynamic regularization analysis on the key response trajectory to obtain a regularized trajectory sequence, identify trajectory sequence parameters in the regularized trajectory sequence, and calculate the trajectory response value corresponding to the trajectory sequence parameters.

[0088] By detecting the key response trajectory corresponding to the spectrum key value, the present invention can deeply understand the entire response process after the user presses the AI ​​independent key, which helps to accurately evaluate the performance and stability of the key and promptly discover problems such as delays, freezes or abnormal fluctuations that may occur in the response process.

[0089] Among them, the key response trajectory refers to the dynamic change path formed by a series of parameters such as key-related signal changes, time changes, state changes, etc. from the start of key triggering to the completion of system response after the user presses the AI ​​independent key. It may include information such as the change in the intensity of the key trigger signal over time, the time history of the system's response to the key, and the conversion of various intermediate states involved in the response process. Optionally, the key response trajectory corresponding to the key value of the spectrum can be detected through performance testing tools, such as: LoadRunner, JMeter, etc.

[0090] Furthermore, the present invention obtains a regular trajectory sequence by dynamically regularizing the key response trajectory, which can eliminate the trajectory differences between different key operations caused by factors such as time differences and different operating forces, making subsequent analysis more accurate and comparable, and helping to discover common problems and potential rules in the key response process, providing a clear direction for the optimization of the all-in-one machine.

[0091] The regularized trajectory sequence refers to a series of optimized and regularized trajectories obtained after dynamic regularization analysis of the key response trajectory.

[0092] As an embodiment of the present invention, the dynamic regularization analysis of the key response trajectory to obtain a regular trajectory sequence includes: performing trajectory preprocessing on the key response trajectory to obtain preprocessed trajectory data; locating the trajectory time axis in the preprocessed trajectory data; extracting the time distance parameters corresponding to the trajectory time axis; using the time distance parameters to construct a trajectory distance matrix corresponding to the key response trajectory; searching for the optimal matching path between different trajectories in the distance matrix; and based on the optimal matching path, performing dynamic regularization analysis on the key response trajectory to obtain a regular trajectory sequence.

[0093] Among them, the pre-processed trajectory data refers to the data obtained after preliminary processing of the key response trajectory such as removing outliers and noise, with the purpose of providing a more accurate and reliable data basis for subsequent analysis; the trajectory time axis refers to the axis used to represent time changes in the key response trajectory, which records the time points of each stage from triggering to completion of the key response; the time distance parameter refers to the parameter used to measure the distance between different time points, which can be a time interval, a time difference, etc., and a trajectory distance matrix can be constructed by extracting the parameter; the trajectory distance matrix refers to a matrix constructed according to the time distance parameters of the key response trajectory, which is used to represent the distance relationship between different trajectory points and provide a basis for finding the optimal matching path; the optimal matching path refers to the path that can achieve the best match between different trajectories searched by a specific algorithm in the trajectory distance matrix, and this path enables different key response trajectories to achieve optimal consistency after dynamic regularization.

[0094] Furthermore, the trajectory preprocessing of the key response trajectory can be implemented through Python's data analysis library, such as: Pandas, NumPy and other tools; the positioning of the trajectory time axis in the preprocessed trajectory data can be implemented through a timestamp analysis method, such as: arranging the data points corresponding to different timestamps in chronological order to determine the time axis; the extraction of the time distance parameter corresponding to the trajectory time axis can be implemented through a time difference calculation method, such as: calculating the time difference between adjacent time points as the time distance parameter; the construction of the trajectory distance matrix corresponding to the key response trajectory can be implemented through the Euclidean distance method, such as: for each trajectory data point at each time point, calculating the Euclidean distance between them, and forming these distances into a matrix; the search for the optimal matching path between different trajectories in the distance matrix can be implemented through a dynamic programming algorithm, such as: using a dynamic programming algorithm to find the optimal path in the distance matrix; the dynamic regularization analysis of the key response trajectory can be implemented through a regularization analysis method, such as: DTW, PCA and other methods.

[0095] By identifying the trajectory sequence parameters in the regular trajectory sequence, the present invention can deeply understand the specific characteristics and laws of the key response trajectory, and can clarify key information such as the trajectory change range, trend and stability under different key operations, which is helpful to accurately evaluate the performance of the key system and timely discover potential problem areas.

[0096] Among them, the trajectory sequence parameters refer to a series of numerical indicators used to describe the characteristics of a regular trajectory sequence, which may include parameters such as trajectory length, average speed, maximum acceleration, response time, etc. that reflect the performance of the key response trajectory in time and space dimensions, and may also cover indicators such as the degree of trajectory fluctuation and symmetry that reflect the trajectory morphological characteristics. Optionally, the identification of trajectory sequence parameters in the regular trajectory sequence can be achieved through a curve fitting method, such as: using a curve fitting algorithm to fit the regular trajectory sequence, and then extracting parameters from the fitting curve.

[0097] Furthermore, by calculating the trajectory response values ​​corresponding to the trajectory sequence parameters, the present invention can convert the abstract trajectory sequence parameters into more intuitive and practical numerical values, so as to better understand the actual effect represented by the key response trajectory, and help to accurately evaluate the response performance of the keys in different situations. For example, it can be clearly known whether the response speed of the key is fast enough and how stable the response is.

[0098] Among them, the trajectory response value refers to the value used to evaluate the overall performance of the key response trajectory within a specific time interval. It is calculated by combining multiple factors such as response amplitude, trigger frequency, and position parameters, and reflects the timeliness, stability, accuracy and other characteristics of the system response caused by the key operation.

[0099] According to an embodiment of the present invention, the step of calculating the trajectory response value corresponding to the trajectory sequence parameter includes:

[0100] The trajectory response value corresponding to the trajectory sequence parameters is calculated using the following formula:

[0101]

[0102] Where R represents the trajectory response value corresponding to the trajectory sequence parameter, N represents the total number of sampling points in the time interval corresponding to the trajectory sequence parameter, i represents the sampling point index in the time interval corresponding to the trajectory sequence parameter, and A(t i ) represents the response amplitude of the i-th sampling point at time t, F(t i ) represents the trigger frequency of the i-th sampling point at time t, m represents the number of dimensions of the trajectory sequence parameter corresponding to the position parameter, j represents the dimension index of the trajectory sequence parameter corresponding to the position parameter, p j (t i) represents the value of the i-th sampling point at time t in the j dimension, Represents the average value of the j dimension in the time interval.

[0103] Specifically, the time interval refers to a specific time period selected for analyzing the key response trajectory, and within this time period, the key response is sampled and analyzed to obtain sufficient data to calculate the trajectory response value; the response amplitude refers to the value of the key response at a specific time point (such as t i The magnitude of the system response caused by the key operation can be measured in many ways, such as the signal strength generated when the key is pressed, the brightness change amplitude of the screen display, the volume of the sound, etc. The trigger frequency refers to the magnitude of the system response caused by the key operation at a specific time point (such as t i moment), the frequency of key triggering or the frequency of some periodic signal related to it. For example, for mechanical keys, the trigger frequency can be understood as the number of times the key is pressed and released per unit time; for keys triggered by electronic signals, the trigger frequency can be the frequency of a specific signal related to the key operation; the position parameter refers to the parameter used to describe the position information of the key response trajectory in different dimensions.

[0104] S3. Collect the change trend corresponding to the trajectory response value, identify the change interval of the change trend within a specific time period, calculate the fluctuation amplitude value corresponding to the change trend based on the change interval, and determine the adjustment mode corresponding to the all-in-one device to be detected based on the fluctuation amplitude value.

[0105] The present invention collects the change trend corresponding to the trajectory response value and identifies the change range of the change trend within a specific time period, which helps to timely discover potential performance degradation or unstable factors so that corresponding maintenance or adjustment measures can be taken before the problem becomes serious. It is also helpful to predict the future situation of key response. A model can be established to predict the change trend of the trajectory response value within a future period of time, providing a reference for equipment maintenance planning and resource allocation.

[0106] Among them, the change trend refers to the overall trend of the trajectory response value over time, which can be reflected in different modes such as gradual increase, gradual decrease, first increase and then decrease, first decrease and then increase, or maintaining relative stability; the specific time period refers to a clear time range specially delineated for the purpose of targeted analysis of the change of trajectory response value. The determination of this time period can be based on different research purposes and actual needs, such as a working cycle of equipment, a maintenance cycle, a product testing phase, etc.; the change interval refers to the numerical range limit of the fluctuation of the trajectory response value within a specific time period, which can be determined by determining the minimum and maximum values. It can be expressed by the span between the intervals, or it can be described by the center value of the interval and a certain deviation range around the center value, reflecting the stability and fluctuation range of the key response value within a specific time period. Optionally, the acquisition of the change trend corresponding to the trajectory response value can be achieved through a time series analysis method, such as: treating the trajectory response value as time series data, and determining the change trend by analyzing the trend, seasonality and periodicity of the time series; the identification of the change interval of the change trend within a specific time period can be achieved through a statistical analysis method, such as: calculating the mean, standard deviation and other statistics of the trajectory response value within a specific time period to determine the change interval.

[0107] Furthermore, the present invention calculates the fluctuation amplitude value corresponding to the change trend based on the change interval, which can intuitively reflect the stability of the key response trajectory in a specific time period, can clearly understand the severity of the change in key response, can evaluate the improvement effect of the equipment, and provide a basis for subsequent optimization and upgrading, thereby judging the stability and reliability of the system.

[0108] Among them, the fluctuation amplitude value refers to a quantitative indicator used to measure the degree of fluctuation of the change trend, which comprehensively considers the range of the change interval and the difference between the trend response value of each sampling point and the trend response average value. The larger the fluctuation amplitude value, the more drastic the fluctuation of the change trend, that is, the more unstable the change of the trajectory response value in a specific time period; conversely, the smaller the fluctuation amplitude value, the relatively stable change trend and the moderate change of the trajectory response value.

[0109] As an embodiment of the present invention, the calculating, based on the change interval, the fluctuation amplitude value corresponding to the change trend includes:

[0110] The fluctuation amplitude value corresponding to the change trend is calculated using the following formula:

[0111]

[0112] Among them, FB represents the fluctuation amplitude value corresponding to the change trend, V max Indicates the maximum value of the interval corresponding to the change interval, Vmin represents the minimum value of the interval corresponding to the change interval, Q represents the total number of sampling times within the change trend, p represents the sampling index within the change trend, R(t' p ) represents the trend response value of the pth sampling at time t', It represents the average value of the trend response of all sampling points within the change trend.

[0113] In detail, the total number of samplings refers to the total number of times the trajectory response values ​​are sampled within a specific time period in the process of analyzing the change trend. The number of samplings will affect the analysis accuracy of the change trend. A larger number of samplings can capture subtle changes in the change trend in more detail, but it will also increase the amount of calculation and the complexity of data processing; a smaller number of samplings may not accurately reflect the overall picture of the change trend; the trend response average value refers to the arithmetic mean of the trend response values ​​of all sampling points in a specific time period, which represents the overall level of the change trend and reflects the average performance of the trajectory response value in the time period. The trend response average value can be used as a benchmark value to measure the degree of deviation of the trend response value of each sampling point from the overall level.

[0114] Furthermore, the present invention determines the adjustment mode corresponding to the all-in-one machine to be tested based on the fluctuation amplitude value, can accurately understand the stability and degree of change of the key response trajectory of the all-in-one machine, can determine the corresponding adjustment mode according to different fluctuation amplitude values, and can optimize and adjust the all-in-one machine in a targeted manner to improve its performance.

[0115] The adjustment mode refers to a specific method and strategy combination for adjusting the integrated device to be detected based on the adjustment factor. The adjustment mode may include a variety of different operations, such as hardware adjustment, software optimization, system setting change, etc.

[0116] As an embodiment of the present invention, determining the adjustment mode corresponding to the integrated machine to be detected based on the fluctuation amplitude value includes: analyzing the size range of the fluctuation amplitude value; based on the size range, stability grading the integrated machine to be detected to obtain a preliminary classification level; determining a stability index corresponding to the preliminary classification level; querying the adjustment factor corresponding to the stability index; and determining the adjustment mode corresponding to the integrated machine to be detected based on the adjustment factor.

[0117] Among them, the size range refers to the numerical interval in which the fluctuation amplitude value is distributed, which reflects the fluctuation degree range of the key response trajectory of the all-in-one machine to be tested within a specific time period, and by analyzing the size range of the fluctuation amplitude value, the stability of the all-in-one machine can be preliminarily judged. For example, the size range can be divided into different intervals such as a small fluctuation range (such as a fluctuation amplitude value between 0 and 10), a medium fluctuation range (such as a fluctuation amplitude value between 10 and 20) and a large fluctuation range (such as a fluctuation amplitude value greater than 20); the preliminary classification level refers to the preliminary classification of the stability of the all-in-one machine to be tested according to the size range of the fluctuation amplitude value. This classification can intuitively reflect the stability level of the all-in-one machine. For example, it can be divided into a high The stability level is a medium stability level and a low stability level. The high stability level indicates that the fluctuation amplitude of the key response trajectory of the all-in-one machine is small, and the system is relatively stable. The medium stability level indicates that the fluctuation amplitude is at a medium level. The low stability level indicates that the fluctuation amplitude is large, and the system stability is poor. The stability index refers to a quantitative indicator used to further accurately describe the degree of stability corresponding to the preliminary classification level. It comprehensively considers factors such as the fluctuation amplitude value, the use time of the all-in-one machine, and the operating environment, and can more comprehensively reflect the stability status of the all-in-one machine. The adjustment factor refers to a parameter or strategy determined according to the stability index for adjusting the all-in-one machine to be tested. The adjustment factor can be a hardware parameter adjustment value, a software algorithm optimization parameter, a system setting adjustment value, etc.

[0118] Furthermore, the analysis of the size range of the fluctuation amplitude value can be achieved through a statistical analysis method, such as: calculating the mean, median, standard deviation and other statistical quantities of the fluctuation amplitude value, and determining the size range of the fluctuation amplitude value by comparing with a preset threshold value; the stability classification of the integrated machine to be detected can be achieved through a clustering analysis method, such as: taking the fluctuation amplitude value of the integrated machine to be detected as a feature, and using a clustering algorithm to divide the integrated machine into different stability levels; the determination of the stability index corresponding to the preliminary classification level can be achieved through a weighted average method, such as: assigning different weights to each indicator according to the degree of influence of different indicators on stability, and then calculating the stability index through weighted average; the query of the adjustment factor corresponding to the stability index can be achieved through a rule reasoning method, such as: building a rule library, and determining the adjustment factor through rule reasoning according to the stability index and other relevant conditions; the determination of the adjustment mode corresponding to the integrated machine to be detected can be achieved through a pattern matching method, such as: different adjustment modes can be pre-defined, such as hardware adjustment mode, software optimization mode, comprehensive adjustment mode, etc., and then the corresponding adjustment mode is selected according to the specific situation of the adjustment factor.

[0119] S4. Identify the button state corresponding to the AI ​​independent button in the adjustment mode, monitor the status data corresponding to the button state in real time, compare the status data with the preset state, and obtain the status comparison parameter.

[0120] The present invention identifies the key state corresponding to the AI ​​independent key in the adjustment mode and monitors the status data corresponding to the key state in real time, so as to timely understand the working condition of the key and ensure its normal operation in the adjustment mode. It can also discover some potential problem trends, such as the key response gradually becoming slower and intermittent failures, so as to perform maintenance and replacement in advance and improve the reliability and service life of the all-in-one machine.

[0121] Among them, the button state refers to the working performance of the AI ​​independent button in a specific adjustment mode, which may include dynamic state information such as whether the button is pressed, the duration of pressing, the frequency of pressing, etc., and may also cover the physical state of the button, such as whether the button is worn, deformed, loose, etc.; the state data refers to a data set that quantifies and records the button state, which may include the electrical parameters of the button, such as the resistance value, current value, voltage value, etc. when the button is triggered; it may also include performance indicator data such as the response time of the button, the number of presses, and the continuous working time. Optionally, the identification of the button state corresponding to the AI ​​independent button in the adjustment mode can be achieved through a sensor monitoring method, such as: by installing a pressure sensor, a displacement sensor, etc. on the button, and monitoring the physical changes of the button in real time to determine the button state; the real-time monitoring of the state data corresponding to the button state can be achieved through an event-driven monitoring method, such as: when the button state changes, an event is triggered and the corresponding state data is recorded.

[0122] The present invention compares the state data with the preset state to obtain state comparison parameters, which can promptly discover the difference between the current state of the button and the ideal state, help intervene in the early stage of the problem, avoid small problems from gradually evolving into serious failures, and adjust the settings of the all-in-one machine to achieve better performance and stability. For example, if it is found that the state data of the button in the current adjustment mode is significantly different from the preset state, you can try to switch to other adjustment modes to find the most suitable working state.

[0123] Among them, the preset state refers to the ideal state or standard state set for the AI ​​independent button under specific conditions. The preset state is usually determined based on the design requirements, performance indicators, usage scenarios and other factors of the all-in-one machine. It may include the response time range, pressing force threshold, trigger frequency range, etc. that the button should have when working normally; the state comparison parameter refers to a quantitative parameter obtained by integrating various difference information between the state data and the preset state. The state comparison parameter can be used to intuitively reflect the degree of closeness or deviation between the current state of the button and the preset state.

[0124] As an embodiment of the present invention, the state data is compared with the preset state to obtain state comparison parameters, including: identifying key state features in the state data; determining a difference metric between the key state features and features in the preset state; calculating the difference value between the state data and the preset state according to the difference metric; evaluating and analyzing the difference value to obtain an evaluation and analysis result; based on the preliminary evaluation result, comparing the state data with the preset state to obtain state comparison parameters.

[0125] Among them, the key state feature refers to an important feature extracted from the state data that can reflect the essential attributes of the current state of the AI ​​independent key, which may include the response time, pressing force, trigger frequency, continuous working time, etc. of the key; the difference metric standard refers to a specific criterion for measuring the degree of difference between the key state feature in the state data and the corresponding feature in the preset state. The difference metric standard may vary according to different key state features. For example, the absolute time difference may be used as a metric for the response time, and the percentage deviation may be used for the pressing force; the difference value refers to the specific numerical difference between the key state feature in the state data and the corresponding feature in the preset state calculated according to the difference metric standard. The difference value may be a positive number, a negative number or zero, respectively indicating that the actual state exceeds, is insufficient or is equal to the preset state; the evaluation and analysis result refers to the conclusion drawn after comprehensive evaluation and analysis of the difference value between the state data and the preset state. The evaluation and analysis result may include a qualitative description of the key state, such as good, general, poor, etc., and may also include specific suggestions and measures, such as whether it is necessary to adjust the adjustment mode, perform maintenance or replace the key.

[0126] Furthermore, the identification of key state features in the state data can be achieved through a feature extraction algorithm, such as: using principal component analysis (PCA), linear discriminant analysis (LDA) and other algorithms to reduce the dimension and extract features of the state data to determine the key state features; the determination of the difference metric between the key state features and the features in the preset state can be achieved through statistical tools, such as: R language, MATLAB, etc. can perform data distribution analysis and statistical modeling to help determine the difference metric; the calculation of the difference value between the state data and the preset state can be achieved through a programming language algorithm, such as: Python, Java, etc. can use basic arithmetic operations and data type conversion to calculate the difference value; the evaluation and analysis of the difference value can be achieved through a data analysis tool, such as: Excel, SPSS, etc. can perform data statistical analysis and chart drawing to help evaluate and analyze the difference value; the comparison of the state data with the preset state can be achieved through data visualization tools, such as: Tableau, PowerBI.

[0127] S5. Query the potential abnormal factors in the state comparison parameters, determine the abnormal level corresponding to the potential abnormal factors, analyze the abnormal level index corresponding to the abnormal level, construct an abnormal detection signal corresponding to the all-in-one device to be detected based on the abnormal level index, and generate an abnormal detection scheme corresponding to the AI ​​independent button in the all-in-one device to be detected based on the abnormal detection signal.

[0128] By querying the potential abnormal factors in the status comparison parameters, the present invention can discover possible problems with buttons in advance, and can take corresponding preventive measures before the problems are obviously manifested, so as to avoid the impact of button failures on the normal use of the all-in-one machine. In addition, by monitoring and analyzing the status comparison parameters, the present invention can quickly locate the all-in-one machine that may have problems, arrange maintenance plans in advance, and improve management efficiency.

[0129] Among them, the potential abnormal factors refer to factors that may cause abnormal state comparison parameters after analyzing abnormal related items. Potential abnormal factors can be hardware failures, software setting problems, environmental factors, etc. For example, if the abnormal related items indicate that both the key response time and the pressing force are abnormal, after analysis it may be found that this is caused by hardware aging of the key or software conflicts, then hardware aging or software conflicts are potential abnormal factors.

[0130] As an embodiment of the present invention, the query of potential abnormal factors in the state comparison parameters includes: analyzing parameter constituent elements corresponding to the state comparison parameters; based on the parameter constituent elements, querying the state comparison parameters for abnormal characteristic indicators; according to the characteristic indicators, screening parameter items of the state comparison parameters to obtain screened parameter items; identifying abnormal correlation items between the screened parameter items; and querying the potential abnormal factors corresponding to the abnormal correlation items.

[0131] Among them, the parameter constituent elements refer to the specific components contained in the state comparison parameters. It can include indicators in multiple aspects, such as the difference value of the response time of the key, the difference value of the pressing force, the difference value of the trigger frequency, the difference value of the continuous working time, etc.; the characteristic indicators refer to specific indicators that can indicate potential abnormalities in the state comparison parameters, which are usually determined based on experience, data analysis or preset rules; the screened parameter items refer to parameter items obtained by screening the state comparison parameters based on characteristic indicators. These parameter items are considered to be abnormal and need to be further analyzed and processed. The screened parameter items can be a single state comparison parameter constituent element or a combination of multiple constituent elements; the abnormal association items refer to the combination of parameter items with abnormal association relationships identified in the screened parameter items. These association items may indicate the existence of common potential abnormal causes. For example, if it is found that the key response time difference value and the pressing force difference value are both large, and similar associations are also shown in historical data, then these two parameter items can be considered as abnormal association items.

[0132] Furthermore, the analysis of the parameter constituent elements corresponding to the state comparison parameters can be achieved through a data deconstruction analysis method, such as: regarding the state comparison parameters as a complex data structure, and determining the parameter constituent elements by decomposing and analyzing each component; the query of the state comparison parameters for abnormal characteristic indicators can be achieved through a threshold setting method, such as: setting one or more thresholds based on experience or historical data, and when the state comparison parameters exceed these thresholds, it is considered that there are abnormal characteristic indicators; the parameter item screening of the state comparison parameters can be achieved through a data analysis tool, such as: Excel, SPSS, etc. can use the screening function to screen the data for parameter items; the identification of abnormal correlation items between the screened parameter items can be achieved through an association rule mining algorithm, such as: Apriori algorithm, FP-Growth algorithm, etc., to mine the correlation relationship between the screened parameter items; the query of the potential abnormal factors corresponding to the abnormal correlation items can be achieved through a causal analysis method, such as: analyzing the causal relationship between abnormal correlation items and potential abnormal factors through causal analysis methods, such as causal diagrams, fishbone diagrams, etc.

[0133] The present invention can intuitively understand the urgency of the key problem by determining the abnormality level corresponding to the potential abnormality factor and analyzing the abnormality level index corresponding to the abnormality level, so as to give priority to handling serious abnormal situations. For example, a potential abnormality factor with a high abnormality level may mean that the key function is about to fail completely and emergency repair measures need to be taken immediately, while a low-level abnormality may only need to be handled at the next maintenance, and technical personnel, maintenance equipment, time and other resources can be reasonably arranged according to the abnormality level to ensure that important problems are solved in a timely manner.

[0134] Among them, the abnormality level refers to the classification level of the abnormal degree of the key state represented by the potential abnormal factor, which reflects the severity and urgency of the key problem. The abnormality level can be divided according to different standards, for example, according to the degree of impact on the function of the all-in-one machine, the number and importance of potential abnormal factors, the duration of the abnormality and other factors; the abnormality level index refers to the specific characteristics and standards used to measure and distinguish different abnormality levels. These indicators can be quantitative values ​​or qualitative descriptions. For example, for high-level abnormalities, indicators may include key response time exceeding twice the preset value, large difference in pressing force leading to frequent false triggering, etc.; for medium-level abnormalities, indicators may be key response time slightly longer than the preset value but still within an acceptable range, certain deviation in pressing force but not affecting normal use, etc.; for low-level abnormalities, indicators may be slight changes in key response time, slight differences in pressing force but not significantly affecting user experience, etc. Optionally, determining the abnormality level corresponding to the potential abnormal factor can be achieved through a scoring system method, such as: setting the score range according to factors such as the degree of influence of the potential abnormal factor on the function of the all-in-one machine, the frequency of occurrence, and the duration, and finally adding the scores of each factor to obtain a total score, and dividing the abnormality level according to the total score; analyzing the abnormality level indicator corresponding to the abnormality level can be achieved through an indicator decomposition method, such as decomposing the abnormality level indicator and analyzing the specific meaning and influencing factors of each indicator.

[0135] Furthermore, the present invention constructs an abnormal detection signal corresponding to the all-in-one machine to be detected based on the abnormal level index, which can realize real-time monitoring and early warning, and can continuously monitor the operating status of the all-in-one machine. Once it is found that the abnormal level index exceeds the normal range, the detection signal is immediately triggered to remind relevant personnel to take timely measures, which helps to intervene when the problem just occurs to avoid further deterioration of the problem and ensure the stable operation of the all-in-one machine.

[0136] The abnormal detection signal refers to the signal output finally constructed to indicate that an abnormal situation occurs in the integrated device to be detected. The abnormal detection signal can be in various forms, such as sound and light alarms, digital signals, email notifications, etc.

[0137] As an embodiment of the present invention, the abnormality detection signal corresponding to the integrated machine to be detected is constructed based on the abnormality level index, including: querying the indicator attribute corresponding to the abnormality level index; analyzing the key attribute elements corresponding to the indicator attribute; designing the signal triggering rule corresponding to the integrated machine to be detected according to the key attribute elements; optimizing and adjusting the signal triggering rule to obtain an optimized signal rule; extracting the abnormality detection entry in the optimized signal rule; and constructing the abnormality detection signal corresponding to the integrated machine to be detected based on the abnormality detection entry.

[0138] Among them, the indicator attributes refer to the specific properties and characteristics of the abnormal level indicators, which may include the numerical range, change trend, stability, correlation with other indicators and other aspects of the indicators; the key attribute elements refer to the factors extracted from the indicator attributes that are crucial to constructing the abnormal detection signal. These elements can be specific numerical ranges, change rate thresholds, combinations of related indicators, etc.; the signal triggering rules refer to the specific criteria for determining when to trigger the abnormal detection signal. These rules are designed based on the key attribute elements and stipulate that abnormal detection signals should be generated under specific indicator conditions. The signal triggering rules may include threshold judgments for a single indicator, combination condition judgments for multiple indicators, etc.; the signal triggering rules refer to a more complete set of rules obtained by adjusting and improving the initial signal triggering rules; the abnormal detection entries refer to the specific entries and contents extracted from the optimized signal rules for constructing the abnormal detection signal, which may include specific indicator conditions, trigger conditions, signal types and other information.

[0139] Furthermore, the query of the indicator attributes corresponding to the abnormal level indicator can be achieved through a data analysis method, such as: analyzing the historical data of the abnormal level indicator, observing its numerical distribution, change trend, correlation with other indicators, etc., to determine the indicator attributes; the design of the signal trigger rule corresponding to the integrated machine to be detected can be achieved through an automated testing tool, such as: Selenium, Appium, etc., which can be used to simulate user operations and monitor system status, and design signal trigger rules according to the monitoring results; the optimization and adjustment of the signal trigger rules can be achieved through a machine learning algorithm, such as: decision tree, random forest, support vector machine, etc., to train the abnormal level indicator data and automatically optimize the signal trigger rules; the extraction of the abnormal detection entries in the optimized signal rules can be achieved through a rule parsing method, such as: parsing the optimized signal trigger rules, extracting the key conditions and judgment statements therein as abnormal detection entries; the construction of the abnormal detection signal corresponding to the integrated machine to be detected can be achieved through a monitoring tool, such as: Zabbix, Nagios and other tools.

[0140] Based on the abnormal detection signal, the present invention generates an abnormal detection scheme corresponding to the AI ​​independent button in the all-in-one machine to be detected, which can promptly and effectively respond to abnormal button situations, take quick action to avoid further deterioration of the abnormal situation, and ensure the normal operation of the all-in-one machine. This helps technicians quickly find the root cause of the fault and save time and cost for troubleshooting. At the same time, a standardized fault handling process can be implemented.

[0141] Among them, the abnormal detection scheme refers to a series of response measures and processing methods formulated for abnormal situations occurring in the AI ​​independent keys in the all-in-one machine to be detected, which includes a detailed description of the abnormal phenomenon, possible cause analysis, specific diagnostic steps, repair suggestions and preventive measures. For example, when an abnormal key response is detected, the abnormal detection scheme may point out that it may be caused by hardware failure, software conflict or environmental factors, and give specific diagnosis and repair suggestions such as checking hardware connections, updating software versions or adjusting the usage environment. At the same time, preventive measures such as regular maintenance and optimizing system settings may be proposed to avoid similar abnormal situations from happening again. Optionally, the generation of the abnormal detection scheme corresponding to the AI ​​independent key in the all-in-one machine to be detected can be implemented through a data analysis platform, such as: using the data analysis platform to analyze a large amount of all-in-one machine operation data, and when an abnormal signal is detected, an abnormal detection scheme is generated through data mining and machine learning algorithms.

[0142] First of all, the present invention obtains the all-in-one machine to be detected and identifies the AI ​​independent button in the all-in-one machine to be detected, so as to concentrate on the precise analysis of the specific device and avoid unnecessary waste of resources and detection errors. Accurately identifying the AI ​​independent button is like finding the key entry point for the subsequent detection work. This process makes the detection more targeted and can quickly focus on the buttons that play an important role in starting the intelligent functions of the all-in-one machine. At the same time, the present invention can deeply understand the entire response process after the user presses the AI ​​independent button by detecting the button response trajectory corresponding to the spectrum key value, which is helpful to accurately evaluate the performance and stability of the button and timely discover the delays, freezes or abnormal fluctuations that may occur in the response process. The present invention collects the change trend corresponding to the trajectory response value and identifies the change range of the change trend within a specific time period, which is helpful to timely discover potential performance degradation or instability factors, so as to take corresponding maintenance or adjustment measures before the problem becomes serious, and is beneficial In order to predict the future situation of key response, a model can be established to predict the changing trend of trajectory response value in the future period of time, and provide a reference basis for equipment maintenance plan and resource allocation. The present invention can timely understand the working condition of the key by identifying the key state corresponding to the AI ​​independent key in the adjustment mode and monitoring the state data corresponding to the key state in real time, so as to ensure its normal operation in the adjustment mode, and can find some potential problem trends, such as the key response gradually slowing down, intermittent failures, etc., and perform maintenance and replacement in advance to improve the reliability and service life of the all-in-one machine. Furthermore, the present invention can find possible problems with the key in advance by querying the potential abnormal factors in the state comparison parameters, and can take corresponding preventive measures before the problem is obviously manifested, so as to avoid the key failure from affecting the normal use of the all-in-one machine, and can quickly locate the all-in-one machine that may have problems by monitoring and analyzing the state comparison parameters, arrange maintenance plans in advance, and improve management efficiency. Therefore, the real-time monitoring and abnormality detection method and system of AI independent keys in the all-in-one machine product proposed by the present invention can improve the key use stability of the all-in-one machine product.

[0143] Embodiment 2:

[0144] like Figure 2 , which is a functional module diagram of an AI independent button real-time monitoring and anomaly detection system in an all-in-one product provided by an embodiment of the present invention.

[0145] The AI ​​independent key real-time monitoring and anomaly detection system 200 in the all-in-one product of the present invention can be installed in an electronic device. According to the functions implemented, the AI ​​independent key real-time monitoring and anomaly detection system 200 in the all-in-one product can include a key value extraction module 201, a response value calculation module 202, a mode adjustment module 203, a state comparison module 204 and a solution generation module 205. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0146] In this embodiment, the functions of each module / unit are as follows:

[0147] The key value extraction module 201 is used to obtain the all-in-one device to be detected, identify the AI ​​independent key in the all-in-one device to be detected, monitor the key trigger signal corresponding to the AI ​​independent key, query the signal feature code corresponding to the key trigger signal, perform spectrum analysis on the signal feature code to obtain a feature spectrum sequence, and extract the spectrum key value in the feature spectrum sequence;

[0148] A response value calculation module 202 is used to detect the key response trajectory corresponding to the spectrum key value, perform dynamic regularization analysis on the key response trajectory to obtain a regularized trajectory sequence, identify trajectory sequence parameters in the regularized trajectory sequence, and calculate the trajectory response value corresponding to the trajectory sequence parameters;

[0149] A mode adjustment module 203 is used to collect the change trend corresponding to the trajectory response value, identify the change interval of the change trend within a specific time period, calculate the fluctuation amplitude value corresponding to the change trend based on the change interval, and determine the adjustment mode corresponding to the all-in-one device to be detected based on the fluctuation amplitude value;

[0150] A state comparison module 204 is used to identify the key state corresponding to the AI ​​independent key in the adjustment mode, monitor the state data corresponding to the key state in real time, compare the state data with a preset state, and obtain a state comparison parameter;

[0151] The solution generation module 205 is used to query the potential abnormal factors in the state comparison parameters, determine the abnormal level corresponding to the potential abnormal factors, analyze the abnormal level index corresponding to the abnormal level, construct the abnormal detection signal corresponding to the all-in-one machine to be detected based on the abnormal level index, and generate the abnormal detection solution corresponding to the AI ​​independent button in the all-in-one machine to be detected based on the abnormal detection signal.

[0152] In detail, the modules described in the AI ​​independent button real-time monitoring and anomaly detection system 200 in the all-in-one product described in the embodiment of the present invention adopt the same technical means as the method for intelligent position error correction of SMT patches based on automation described in the accompanying drawings when used, and can produce the same technical effects, which will not be repeated here.

[0153] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product, characterized in that: The method comprises: Acquire the all-in-one device to be detected, identify the AI ​​independent key in the all-in-one device to be detected, monitor the key trigger signal corresponding to the AI ​​independent key, query the signal feature code corresponding to the key trigger signal, perform spectrum analysis on the signal feature code to obtain a feature spectrum sequence, and extract spectrum key values ​​in the feature spectrum sequence; Detecting the key response trajectory corresponding to the spectrum key value, performing dynamic regularization analysis on the key response trajectory to obtain a regularized trajectory sequence, identifying trajectory sequence parameters in the regularized trajectory sequence, and calculating the trajectory response value corresponding to the trajectory sequence parameters; Collecting a change trend corresponding to the trajectory response value, identifying a change interval of the change trend within a specific time period, calculating a fluctuation amplitude value corresponding to the change trend based on the change interval, and determining an adjustment mode corresponding to the all-in-one device to be detected based on the fluctuation amplitude value; Identify the key state corresponding to the AI ​​independent key in the adjustment mode, monitor the state data corresponding to the key state in real time, compare the state data with the preset state, and obtain the state comparison parameter; Query the potential abnormal factors in the state comparison parameters, determine the abnormal level corresponding to the potential abnormal factors, analyze the abnormal level index corresponding to the abnormal level, construct the abnormal detection signal corresponding to the all-in-one machine to be detected based on the abnormal level index, and generate the abnormal detection scheme corresponding to the AI ​​independent button in the all-in-one machine to be detected based on the abnormal detection signal.

2. The method for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product according to claim 1, characterized in that: The performing spectrum analysis on the signal characteristic code to obtain a characteristic spectrum sequence includes: Performing denoising processing on the signal characteristic code to obtain a denoised signal characteristic code; Querying the time domain signal corresponding to the denoised signal feature code; Converting the time domain signal into a frequency domain signal; Extracting main frequency components from the frequency domain signal; Normalizing the amplitude of the main frequency component to obtain a normalized frequency feature; Performing spectrum analysis on the normalized frequency features to obtain a characteristic spectrum sequence.

3. The method for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product according to claim 1, characterized in that: The step of dynamically regularizing and analyzing the key response trajectory to obtain a regular trajectory sequence includes: Performing trajectory preprocessing on the key response trajectory to obtain preprocessed trajectory data; Locating a trajectory time axis in the preprocessed trajectory data; Extracting the time distance parameter corresponding to the trajectory time axis; Using the time distance parameter, constructing a trajectory distance matrix corresponding to the key response trajectory; Searching for the best matching path between different trajectories in the distance matrix; Based on the optimal matching path, the key response trajectory is dynamically regularized and analyzed to obtain a regular trajectory sequence.

4. The method for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product according to claim 1, characterized in that: The calculating the trajectory response value corresponding to the trajectory sequence parameter includes: The trajectory response value corresponding to the trajectory sequence parameters is calculated using the following formula: Where R represents the trajectory response value corresponding to the trajectory sequence parameter, N represents the total number of sampling points in the time interval corresponding to the trajectory sequence parameter, i represents the sampling point index in the time interval corresponding to the trajectory sequence parameter, and A(t i ) represents the response amplitude of the i-th sampling point at time t, F(t i ) represents the trigger frequency of the i-th sampling point at time t, m represents the number of dimensions of the trajectory sequence parameter corresponding to the position parameter, j represents the dimension index of the trajectory sequence parameter corresponding to the position parameter, p j (t i ) represents the value of the i-th sampling point at time t in the j dimension, Represents the average value of the j dimension in the time interval.

5. The method for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product according to claim 1, characterized in that: The calculating, based on the change interval, a fluctuation amplitude value corresponding to the change trend includes: The fluctuation amplitude value corresponding to the change trend is calculated using the following formula: Wherein, FB represents the fluctuation amplitude value corresponding to the change trend, V max Indicates the maximum value of the interval corresponding to the change interval, V min represents the minimum value of the interval corresponding to the change interval, Q represents the total number of sampling times within the change trend, p represents the sampling index within the change trend, and R ( t' p ) represents the trend response value of the pth sampling at time t', It represents the average value of the trend response of all sampling points within the change trend.

6. The method for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product according to claim 1, characterized in that: The determining, based on the fluctuation amplitude value, a corresponding adjustment mode of the to-be-detected all-in-one device includes: Analyze the range of the fluctuation amplitude value; Based on the size range, the to-be-detected all-in-one device is stably classified to obtain a preliminary classification level; Determining a stability index corresponding to the preliminary classification level; Querying the adjustment factor corresponding to the stability index; Based on the adjustment factor, an adjustment mode corresponding to the all-in-one device to be detected is determined.

7. The method for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product according to claim 1, characterized in that: The step of comparing the state data with a preset state to obtain a state comparison parameter includes: identifying key status features in the status data; Determine a difference metric between the key state feature and the feature in the preset state; Calculate the difference between the state data and the preset state according to the difference metric standard; Performing evaluation and analysis on the difference value to obtain an evaluation and analysis result; Based on the preliminary evaluation result, the state data is compared with a preset state to obtain a state comparison parameter.

8. The method for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product according to claim 1, characterized in that: The querying of potential abnormal factors in the state comparison parameters includes: Analyzing parameter constituent elements corresponding to the state comparison parameters; Based on the parameter constituent elements, query the state comparison parameters for abnormal characteristic indicators; According to the characteristic index, the state comparison parameter is screened to obtain a screened parameter item; Identifying abnormal correlation items between the screening parameter items; Query the potential abnormal factor corresponding to the abnormal association item.

9. The method for real-time monitoring and abnormality detection of AI independent buttons in an all-in-one product according to claim 1, characterized in that: The constructing, based on the abnormality level indicator, an abnormality detection signal corresponding to the to-be-detected integrated device includes: Query the indicator attribute corresponding to the abnormal level indicator; Analyze the key attribute elements corresponding to the indicator attributes; According to the key attribute elements, design a signal trigger rule corresponding to the all-in-one device to be detected; Optimizing and adjusting the signal triggering rule to obtain an optimized signal rule; Extracting anomaly detection entries in the optimized signal rule; Based on the abnormality detection entry, an abnormality detection signal corresponding to the all-in-one device to be detected is constructed.

10. AI independent key real-time monitoring and anomaly detection system in all-in-one products, characterized by: The system is used to execute the AI ​​independent button real-time monitoring and abnormality detection method in the all-in-one product according to any one of claims 1 to 9, and the system comprises: A key value extraction module is used to obtain the integrated machine to be detected, identify the AI ​​independent key in the integrated machine to be detected, monitor the key trigger signal corresponding to the AI ​​independent key, query the signal feature code corresponding to the key trigger signal, perform spectrum analysis on the signal feature code to obtain a feature spectrum sequence, and extract the spectrum key value in the feature spectrum sequence; a response value calculation module, used for detecting the key response trajectory corresponding to the spectrum key value, performing dynamic regularization analysis on the key response trajectory to obtain a regularized trajectory sequence, identifying trajectory sequence parameters in the regularized trajectory sequence, and calculating the trajectory response value corresponding to the trajectory sequence parameters; A mode adjustment module, used for collecting the change trend corresponding to the trajectory response value, identifying the change interval of the change trend within a specific time period, calculating the fluctuation amplitude value corresponding to the change trend based on the change interval, and determining the adjustment mode corresponding to the all-in-one device to be detected based on the fluctuation amplitude value; A state comparison module, used to identify the key state corresponding to the AI ​​independent key in the adjustment mode, monitor the state data corresponding to the key state in real time, compare the state data with the preset state, and obtain a state comparison parameter; A scheme generation module is used to query the potential abnormal factors in the state comparison parameters, determine the abnormal level corresponding to the potential abnormal factors, analyze the abnormal level index corresponding to the abnormal level, construct an abnormal detection signal corresponding to the all-in-one machine to be detected based on the abnormal level index, and generate an abnormal detection scheme corresponding to the AI ​​independent button in the all-in-one machine to be detected based on the abnormal detection signal.

Images