A multifunctional measurement and detection system

Abstract

The present invention relates to the field of multifunctional detection technology, specifically a multifunctional measurement detection system, which includes: a time alignment module, an analog-to-digital conversion module, a data coupling module, a vector construction module, and an event encapsulation module. In the present invention, by time-aligning the measurement data in multiple channels, the synchronization consistency of heterogeneous data is improved, and combined with the dynamic analysis of amplitude fluctuation characteristics, the real-time adjustment of the analog-to-digital conversion resolution is achieved, and the response sensitivity of the system to non-steady-state signals is enhanced. The cross-channel amplitude and trend relationship is used to construct a composite parameter sequence, and the coupling expression capability between multiple parameters is optimized. Through state vector modeling and direction difference judgment, the recognition accuracy of multi-node linkage trends is improved. The time series is used to reconstruct the abnormal event path and encapsulate the process record, and the detection data is converted from raw sampling to structured event chain data, thereby enhancing the organization and traceability of the abnormal detection results.

Description

Technical Field

[0001] The present invention relates to the field of multifunctional detection technology, and in particular to a multifunctional measurement and detection system. Background Art

[0002] The field of multifunctional detection technology includes comprehensive testing and measurement of multiple physical quantities or performance parameters of the object under test. The core content of this technology is the use of combined sensors, signal acquisition modules and control systems to synchronously measure and process multiple parameters such as temperature, pressure, flow, voltage, current, displacement, vibration, etc. The technical field mainly includes multi-parameter perception mechanism, collaborative acquisition of detection signals, unified analysis mechanism of electrical and non-electrical quantities, and automated control logic of the test process. The coordinated work between multiple measurement modules is achieved through integrated structural design, which is suitable for various application scenarios such as industrial manufacturing, equipment commissioning, operation and maintenance, and experimental testing.

[0003] Among them, a multifunctional measurement and detection system refers to a device system with multiple detection units that can collect and measure comprehensive parameters of different types of objects under test. The patent subject is aimed at complex detection needs involving multiple test items at the same time, covering the measurement of physical quantities such as length detection, current detection, voltage detection, and temperature detection. Specifically, different signals are independently collected through a combination module equipped with a displacement sensor, voltage sensor, current transformer, thermocouple or thermistor, and then the detection signal is converted into a processable data form through the channel switching mechanism and analog-to-digital conversion module in the main control unit. With the help of the embedded sampling control circuit and storage management circuit, multiple channels of data are synchronously collected and measured in sub-items, completing the structured detection process of various parameters on the same platform, and building a functional detection platform with integrated features.

[0004] In the comprehensive measurement process of traditional multifunctional metrology and detection technology, the synchronous acquisition of multiple physical quantities usually relies on a unified sampling period and a fixed resolution mode, and cannot make real-time adjustments to the dynamic characteristics of the data itself, resulting in poor sampling accuracy of non-steady-state data, especially insufficient capture of short-term mutations and high-dynamic characteristic signals. The data is prone to blurring or distortion. The data processing process lacks in-depth analysis of the coupling relationship between cross-channel data, resulting in insufficient sensitivity to identify potential abnormal states and events, which limits the ability to identify complex faults early and delays the detection time. There is a lack of means to perform multi-dimensional vector processing on the measurement data change process. Data correlation analysis is limited to the change trend of a single parameter itself, and fails to effectively utilize the dynamic linkage relationship between parameters for in-depth mining, which limits the depth and comprehensiveness of analysis of abnormal events and significantly affects detection efficiency and accuracy. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a multifunctional measurement and detection system.

[0006] In order to achieve the above-mentioned object, the present invention adopts the following technical solution: a multifunctional measurement and detection system includes:

[0007] The time alignment module obtains a sequence of continuous sampling values from each measurement channel, calculates the rate of change between adjacent sampling points, extracts the time position of the peak node, compares the time offset of the peak node in each channel, and performs time alignment on the data to generate a heterogeneous data alignment sequence.

[0008] The analog-to-digital conversion module calls the heterogeneous data alignment sequence, extracts the change amplitude of the measurement data in a continuous period and calculates the rate fluctuation characteristics, adjusts the resolution parameters of the analog-to-digital conversion in real time according to the rate change, and generates a dynamic sampling configuration interval;

[0009] The data coupling module calls the dynamic sampling configuration interval, obtains the sampling values of multiple measurement channels at the same time node, performs ratio combination using the amplitude difference and the trend change value, constructs a parameter sequence describing the fluctuation relationship between the channels, and obtains a composite parameter coupling sequence;

[0010] The vector construction module calls the composite parameter coupling sequence, analyzes the state value changes of each parameter node in a continuous time period, extracts the change direction, change amplitude and continuous change frequency, constructs the node state change vector, and combines the angle calculation and vector projection length comparison to obtain the multi-node state linkage trend coefficient.

[0011] As a further solution of the present invention, the heterogeneous data alignment sequence includes a time alignment channel number, a key node timestamp, and an interpolated sampling value; the dynamic sampling configuration interval includes a rate threshold segment, a resolution parameter level, and a configuration index number; the composite parameter coupling sequence is specifically an amplitude difference sequence, a trend change value sequence, and a trend amplitude ratio; the multi-node state linkage trend coefficient includes a state change direction vector, a state change amplitude vector, and a change frequency characteristic value.

[0012] As a further solution of the present invention, the time alignment module includes:

[0013] The rate identification submodule obtains the continuous sampling value sequence of each measurement channel, extracts the numerical difference between adjacent sampling points, performs difference operation on adjacent sampling values in the continuous sampling segment within each channel, calculates the change rate, locates the change rate peak node in each channel, and generates a set of key nodes for rate change;

[0014] The offset comparison submodule calls the rate change key node set, extracts the timestamp data of the peak nodes in the channel, arranges the key node times in each channel in sequence according to the time dimension, identifies the time offset between key nodes in the same time period, and generates the offset time interval;

[0015] The node interpolation submodule calls the offset time interval and uses the formula:

[0016] ;

[0017] Calculate and reconstruct sampling values, perform time alignment on the measurement data of multiple channels through data interpolation, and generate heterogeneous data alignment sequences;

[0018] in, is the reconstructed sample value of the interpolation point, is the known sample value before the interpolation point, is the known sample value after the interpolation point, is the sampling value The corresponding timestamp, is the sampling value The corresponding timestamp, is the target time position of the interpolation point on the time axis, is the index number of the previous sampling point in the time interval where the interpolation point is located. is the index number of the next sampling point in the time interval where the interpolation point is located. The index number of the current interpolation target point on the time axis.

[0019] As a further solution of the present invention, the analog-to-digital conversion module includes:

[0020] The amplitude extraction submodule calls the heterogeneous data alignment sequence to calculate the amplitude variation range of the sampling value of each channel within a preset time window in real time to obtain the continuous amplitude variation interval value;

[0021] The rate fluctuation submodule calls the continuous amplitude change interval value, calculates the sampling change rate of each channel in the time window, extracts the difference change between the rates of adjacent time periods, compares the rate fluctuation frequency and the difference interval in multiple time periods, establishes a rate change ratio group, and generates a sampling rate fluctuation characteristic factor;

[0022] The resolution configuration submodule calls the sampling rate fluctuation characteristic factor and uses the formula:

[0023] ;

[0024] Calculate the resolution parameter adjustment value of analog-to-digital conversion and generate a dynamic sampling configuration interval;

[0025] in, Adjust the resolution parameter value for analog-to-digital conversion. The first The rate value of the sampling point, is the average value of all rate values in the jump section, is the fluctuation frequency per unit time in the jump section, For the The difference in velocity direction change within a time period, is the maximum amplitude of the sampling rate change within the jump section, is the mean of the measured amplitudes before and after the segment, is the peak value of the measured amplitude in this section, is the number of sampling points participating in the rate fluctuation statistics in the jump section, The number of time periods participating in the direction difference calculation within the jump section, is the index number of the rate sampling point, The index number of the segment participating in the velocity direction difference.

[0026] As a further solution of the present invention, the data coupling module includes:

[0027] The channel data acquisition submodule calls the dynamic sampling configuration interval, extracts the synchronous sampling values of multiple measurement channels at the same time node, and generates a channel synchronous sampling matrix;

[0028] The trend difference extraction submodule calls the channel synchronous sampling matrix, calculates the sampling amplitude difference of multiple channels at the same time node, combines the sampling value change direction of each channel in a continuous time period, extracts the trend change direction difference between channels, and generates a trend amplitude difference structure group;

[0029] The coupling feature construction submodule calls the trend amplitude difference structure group, performs time dimension ratio combination processing on the amplitude difference sequence and trend difference sequence between channels, constructs a structural feature sequence for describing the change relationship between channels, and generates a composite parameter coupling sequence.

[0030] As a further solution of the present invention, the vector construction module includes:

[0031] The node extraction submodule calls the composite parameter coupling sequence to extract the state value changes of each parameter node in a continuous time period, and extracts the change direction, change amplitude, and change frequency according to the direction of the state change, the amplitude increment, and the change frequency in adjacent time periods to obtain the node state feature information;

[0032] The vector generation submodule calls the node state feature information, maps each node state feature into a three-dimensional vector axis value, constructs a node state change vector, and generates a node state vector structure set;

[0033] The linkage judgment submodule calls the node state vector structure set and performs angle calculation and vector projection difference comparison on multiple node state vector pairs using the formula:

[0034] ;

[0035] Calculate the state linkage trend coefficients of multiple node pairs and obtain the state linkage trend coefficients of multiple nodes;

[0036] in, Representation node With node The state linkage trend coefficient, Representation node In the The projection value on the dimensional vector coordinates, Representation node In the The projection value on the dimensional vector coordinates, Representation node In the The projection value on the dimensional vector coordinates, Representation node In the The projection value on the dimensional vector coordinates, represents the number of dimensions of the state vector, Representation node The state change vector, Representation node The state change vector, Indicates the first node number in the state node pair, Indicates the comparison node number in the state node pair, Represents the vector dimension index used for difference calculation, Indicates the vector dimension index used for amplitude total statistics.

[0037] As a further embodiment of the present invention, the system further comprises:

[0038] The event encapsulation module calls the multi-node state linkage trend coefficient, detects abnormal measurement events by identifying the directional differences of the state vectors between multiple parameter nodes, extracts the corresponding parameter types and timestamps, reconstructs the abnormal event evolution path according to the timestamp sequence, encapsulates the measurement data using the path sequence, and generates an event process encapsulation record;

[0039] The event process encapsulation record specifically refers to the exception parameter type, exception timestamp, and event path number.

[0040] As a further solution of the present invention, the event encapsulation module includes:

[0041] The vector difference identification submodule calls the multi-node state linkage trend coefficient to obtain the state change vectors of multiple parameter nodes, detects abnormal measurement events by analyzing the direction differences of the state vectors between the nodes, and generates an abnormal event identification result;

[0042] The abnormal event path reconstruction submodule extracts the parameter type and timestamp data corresponding to the target abnormal event based on the abnormal event identification result, and reconstructs the evolution path of the abnormal event according to the timestamp sequence to generate event evolution path data;

[0043] The event encapsulation and recording submodule calls the event evolution path data, encapsulates the measurement data according to the path sequence, and generates an event process encapsulation record.

[0044] Compared with the prior art, the advantages and positive effects of the present invention are:

[0045] In the present invention, by time-aligning the measurement data in multiple channels, the synchronization consistency of heterogeneous data is improved. Combined with the dynamic analysis of amplitude fluctuation characteristics, real-time adjustment of the analog-to-digital conversion resolution is achieved, and the system's response sensitivity to non-steady-state signals is enhanced. The cross-channel amplitude and trend relationship is used to construct a composite parameter sequence, which optimizes the coupling expression capability between multiple parameters. Through state vector modeling and direction difference judgment, the recognition accuracy of multi-node linkage trends is improved. Time series is used to reconstruct the abnormal event path and encapsulate process records, and the detection data is converted from raw sampling to structured event chain data, thereby enhancing the organization and traceability of abnormal detection results. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a system flow chart of the present invention;

[0047] Figure 2 This is a flow chart of the time alignment module of the present invention;

[0048] Figure 3 This is a flow chart of the analog-to-digital conversion module of the present invention;

[0049] Figure 4 This is a flow chart of the data coupling module of the present invention;

[0050] Figure 5 It is a flow chart of the vector construction module of the present invention;

[0051] Figure 6 This is a flow chart of the event encapsulation module of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0053] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0054] Example 1: Please refer to Figure 1 , a multifunctional measurement and detection system includes:

[0055] The time alignment module obtains a sequence of continuous sampling values from each measurement channel, calculates the rate of change between adjacent sampling points, extracts the time position of the peak node, compares the time offset of the peak node in each channel, and performs time alignment on the data to generate a heterogeneous data alignment sequence.

[0056] The analog-to-digital conversion module calls the heterogeneous data alignment sequence, extracts the change amplitude of the measured data within a continuous period and calculates the rate fluctuation characteristics. It adjusts the resolution parameters of the analog-to-digital conversion in real time according to the rate change and generates a dynamic sampling configuration interval.

[0057] The data coupling module calls the dynamic sampling configuration interval, obtains the sampling values of multiple measurement channels at the same time node, performs ratio combination using the amplitude difference and trend change value, constructs a parameter sequence describing the fluctuation relationship between channels, and obtains a composite parameter coupling sequence;

[0058] The vector construction module calls the composite parameter coupling sequence, analyzes the state value changes of each parameter node in a continuous time period, extracts the change direction, change amplitude and continuous change frequency, constructs the node state change vector, and combines angle calculation and vector projection length comparison to obtain the multi-node state linkage trend coefficient;

[0059] The event encapsulation module calls the multi-node state linkage trend coefficient, detects abnormal measurement events by identifying the directional differences of the state vectors between multiple parameter nodes, extracts the corresponding parameter types and timestamps, and reconstructs the abnormal event evolution path according to the timestamp sequence. The measurement data is encapsulated using the path sequence to generate event process encapsulation records.

[0060] The heterogeneous data alignment sequence includes the time alignment channel number, key node timestamp, and interpolated sampling value. The dynamic sampling configuration interval includes the rate threshold segment, resolution parameter level, and configuration index number. The composite parameter coupling sequence is specifically the amplitude difference sequence, trend change value sequence, and trend amplitude ratio. The multi-node state linkage trend coefficient includes the state change direction vector, state change amplitude vector, and change frequency characteristic value. The event process encapsulation record specifically refers to the abnormal parameter type, abnormal timestamp, and event path number.

[0061] See also Figure 2 , the time alignment module includes:

[0062] The rate identification submodule obtains the continuous sampling value sequence of each measurement channel, extracts the numerical difference between adjacent sampling points, performs difference operation on adjacent sampling values in the continuous sampling segment within each channel, calculates the change rate, locates the change rate peak node in each channel, and generates a set of key nodes for rate change;

[0063] Obtain a sequence of continuous sampling values for each measurement channel, extract the numerical difference and time difference of any pair of adjacent sampling points in each channel in the order of sampling time, divide the difference by the time interval to obtain the rate of change per unit time. For channel A, if the amplitudes of a pair of adjacent sampling points are 2.5 and 3.0 respectively, and the timestamps are 100ms and 110ms respectively, then the rate of change is (3.0-2.5) / (110-100)=0.05 / ms. For channel B, if the sampling points are 2.7 and 3.1, and the corresponding timestamps are the same, then the rate of change is 0.04 / ms. The system repeats the above operation for the complete sequence of each channel to form the rate of change per unit time. The rate sequence is obtained by sliding the window mechanism, setting the window width to 3 sampling points. The local extreme value comparison is performed on the rate values in any continuous window in the rate sequence, and the time index number corresponding to the maximum rate in the window is identified. The index number is mapped to the original data sequence, and its corresponding timestamp is obtained as the rate peak node of the window. Assuming that the rate in channel A is 0.02, 0.06, and 0.05 / ms at the 4th, 7th, and 11th sampling points respectively, the local maximum value is identified when the window slides to the 7th point, and the time point is 170ms. Finally, the positioning is repeated for each channel, and the rate peak node times of channel A are {170, 190, 250}ms and channel B are {180, 200, 260}ms, respectively, to obtain the key node set of rate change.

[0064] The offset comparison submodule calls the rate change key node set, extracts the timestamp data of the peak node in the channel, arranges the key node time in each channel according to the time dimension, identifies the time offset between key nodes in the same time period, and generates the offset time interval;

[0065] Call the rate change key node set, extract the timestamp data of the identified key nodes in each channel, sort them in ascending order according to the time axis to build a time series index, set a target time range such as [160ms, 200ms], and filter the key node time of each channel within this time period. If there are two nodes of 170ms and 190ms in channel A, and two nodes of 180ms and 200ms in channel B, then calculate the time difference between the corresponding positions of each pair of key time nodes. For example, the difference between 170ms and 180ms is 10ms, and the difference between 190ms and 200ms is also 10ms. Merge them into the offset length vector [10, 10]ms, the system sets the offset time judgment threshold to 5ms, which is the acceptable time error benchmark value for cross-channel nodes. Since the difference between the above two node pairs is greater than the threshold, the system judges that there is a synchronization offset between the key nodes of channel A and channel B during this time period. The judgment operation is to compare the numerical value of the offset value with the benchmark value. If the offset value is greater than or equal to the benchmark value, it is considered that the offset exists, otherwise it is considered that the synchronization exists. Finally, the offsets between multiple channels and their time period identifiers are combined and packaged to generate an offset time interval.

[0066] The node interpolation submodule calls the offset time interval using the formula:

[0067] ;

[0068] Calculate and reconstruct sampling values, perform time alignment on the measurement data of multiple channels through data interpolation, and generate heterogeneous data alignment sequences;

[0069] in, is the reconstructed sample value of the interpolation point, is the known sample value before the interpolation point, is the known sample value after the interpolation point, is the sampling value The corresponding timestamp, is the sampling value The corresponding timestamp, is the target time position of the interpolation point on the time axis, is the index number of the previous sampling point in the time interval where the interpolation point is located. is the index number of the next sampling point in the time interval where the interpolation point is located. The index number of the current interpolation target point on the time axis;

[0070] When calling the offset time interval and performing time alignment on the measurement data of multiple channels, first extract the known sampling values on both sides of the offset segment based on the key node time information of each channel in the offset time interval, where the sampling value at the moment before the sampling point is set as , the sampling value at the next moment is set to , and extract the corresponding timestamps of the two and , followed by the target time and The time difference between the interpolation points is used to determine the time offset of the interpolation point, and the amplitude difference between the known values is calculated to obtain the interpolation amplitude. , specifically using the formula:

[0071] ;

[0072] In the specific example, the two sample values of channel A are 3.0 ( ) and 3.5 ( ), the corresponding timestamp is 110ms ( ) and 120ms ( ), if the target interpolation point time is 115ms ( ), then the interpolation value of this point is:

[0073] ;

[0074] The interpolation value of channel A at 115ms is 3.25. Channel B is processed in the same way. 、 、 、 , then its interpolation value at 115ms is:

[0075] ;

[0076] Using all interpolation points between channels as reference nodes, the interpolation values of each channel are reconstructed in time order to form a complete time-aligned data stream sequence, ensuring that each channel has a corresponding amplitude sampling point at any time. The misaligned data is corrected through interpolation to generate a heterogeneous data alignment sequence. Represents the reconstructed sample value of the point to be interpolated, and Respectively represent the amplitude of the previous sampling point and the next sampling point in the interpolation interval, 、 are the timestamps of the two sampling points, Indicates the time position of the target interpolation point on the time axis, is the index number of the previous sampling point in the current interpolation interval, It is the index number of the current interpolation target point in the time series. The actual time displacement of the interpolation point is used as the multiplication factor, and the adjacent amplitude differences are combined to realize dynamic amplitude generation. The reconstructed sampling value represents the situation where there is a time offset between different measurement channels. The sampling value of a certain target time point is estimated through interpolation calculation to eliminate the time synchronization problem between channels and realize time alignment, effectively establishing a continuous mapping relationship between time nodes and signal amplitudes.

[0077] See also Figure 3 , the analog-to-digital conversion module includes:

[0078] The amplitude extraction submodule calls the heterogeneous data alignment sequence to calculate the amplitude variation range of the sampling value of each channel within the preset time window in real time, and obtains the continuous amplitude variation interval value;

[0079] Call the heterogeneous data alignment sequence, first obtain the continuous sampling value sequence after time interpolation processing in each measurement channel, and split it according to the set time window. Taking 5ms as a group of sliding windows as an example, the system extracts the difference between the maximum amplitude and the minimum amplitude in each group of windows as the amplitude change interval value of the time window. If the sampling value of channel A in a certain time window is {2.5, 2.8, 3.0, 3.1, 3.3}, then the maximum value in the window is 3.3, the minimum value is 2.5, and the change range is 3.3-2.5=0.8. Channel B in the same time window is {2.5, 2.8, 3.0, 3.1, 3.3}. The sampling values in the time period are {2.6, 2.9, 3.0, 3.2, 3.4}, so the amplitude change interval value is 3.4-2.6=0.8. The system slides the window in sequence to extract the amplitude interval of each time period to form the amplitude change stream sequence under the channel time series. The amplitude change interval judgment threshold is set to 0.5 unit amplitude. If the amplitude interval values of two consecutive windows exceed the threshold, it is determined that the segment is in the jump behavior stage. The window start and end time and the corresponding channel number are recorded to form a continuous amplitude change record sequence, and the continuous amplitude change interval value is generated based on this.

[0080] The rate fluctuation submodule calls the continuous amplitude change interval value, calculates the sampling change rate of each channel in the time window, extracts the difference change between the rates of adjacent time periods, compares the rate fluctuation frequency and the difference interval in multiple time periods, establishes a rate change ratio group, and generates the sampling rate fluctuation characteristic factor;

[0081] Call the continuous amplitude change interval value, set the change rate within the time period for each channel, the system performs the difference operation on the amplitude changes of two adjacent sampling moments based on the sampling point, and divides it by the corresponding time interval to form a change rate sequence per unit time. If channel A has an amplitude sequence of {2.5, 2.8, 3.1, 3.0, 3.3} within a certain period of time, and the corresponding timestamps are 100ms, 105ms, 110ms, 115ms, 120ms respectively, then the change rate is: (2.8-2.5) / 5=0.06, (3.1-2.8) / 5=0.06, (3.0-3.1) / 5=-0.02, (3.3-3.0) / 5=0.06, forming a rate sequence {0.06, 0.06, -0.02,0.06}, then the system calculates the fluctuation change of the rate in different time periods, that is, subtracting the current period rate from the previous period rate and taking the absolute value, such as |0.06-0.06|=0.00, |0.06-(-0.02)|=0.08, |-0.02-0.06|=0.08, generating a difference fluctuation sequence, setting the fluctuation difference judgment threshold to 0.05. If the difference exceeds this value, it is counted as a fluctuation event. The fluctuation frequency within the unit time window is 4, and the maximum rate change amplitude under the current jump behavior is recorded as 0.11-0.07=0.04. Finally, based on the fluctuation behavior, the frequency value, direction change difference and total rate change of each channel jump segment are constructed, and combined to form the sampling rate fluctuation characteristic factor.

[0082] The resolution configuration submodule calls the sampling rate fluctuation characteristic factor, and uses the formula according to the rate change amplitude, the fluctuation frequency per unit time, and the difference in the sampling value fluctuation direction:

[0083] ;

[0084] Calculate the resolution parameter adjustment value of analog-to-digital conversion and generate a dynamic sampling configuration interval;

[0085] in, Adjust the resolution parameter value for analog-to-digital conversion. The first The rate value of the sampling point, is the average value of all rate values in the jump section, is the fluctuation frequency per unit time in the jump section, For the The difference in velocity direction change within a time period, is the maximum amplitude of the sampling rate change within the jump section, is the mean of the measured amplitudes before and after the segment, is the peak value of the measured amplitude in this section, is the number of sampling points participating in the rate fluctuation statistics in the jump section, The number of time periods participating in the direction difference calculation within the jump section, is the index number of the rate sampling point, The index number of the segment participating in the velocity direction difference;

[0086] Calling the sampling rate fluctuation characteristic factor, the rate sampling value of the selected channel in the jump segment is {0.08, 0.10, 0.07, 0.11, 0.09}, and its average value is:

[0087] ;

[0088] The sum of the absolute differences between the rate value and the average rate value is further calculated:

[0089] ;

[0090] Fluctuation frequency Set it to 4 and take its square root:

[0091] ;

[0092] The direction change difference is set to {0.02, 0.03, 0.01}, and its absolute value sum is:

[0093] ;

[0094] Set the maximum rate change range , average amplitude , peak amplitude , then the difference is:

[0095] ;

[0096] Substituting all the above values into the formula:

[0097] ;

[0098] Finally, the resolution adjustment parameter value of analog-to-digital conversion is calculated , and generate a dynamic sampling configuration interval accordingly.

[0099] in, Indicates the The value of the rate sampling point, is the average rate within the transition segment, is the fluctuation frequency in the jump section, For the The difference in velocity direction change within a time period, is the maximum amplitude difference of the rate within the jump segment, is the average amplitude, is the peak amplitude of the current segment, It is the dynamic quantity ultimately used to adjust the accuracy of analog-to-digital conversion.

[0100] Table 1 Resolution parameter calculation input table

[0101] ;

[0102] Table 1 lists the transition rate values and direction change differences involved in the calculation. The rate value is calculated by dividing the amplitude of adjacent time segments by the time interval, while the direction change difference is calculated by taking the scalar difference of the trend direction. These values are used to determine the actual values of the parameters in the formula. Based on the data in the table and the calculation, the resolution adjustment value is 0.000727. This result indicates that the analog-to-digital conversion module at this stage needs to adopt a more precise conversion step size to adapt to the sampling rate fluctuation characteristics.

[0103] See also Figure 4 , the data coupling module includes:

[0104] The channel data acquisition submodule calls the dynamic sampling configuration interval to extract the synchronous sampling values of multiple measurement channels at the same time node and generate a channel synchronous sampling matrix;

[0105] After invoking the dynamic sampling configuration interval, the system enters the multi-channel sampling data synchronization phase, ensuring that all channels have complete measurement records at the same time. Taking channels A and B as an example, the sampling values of channel A at time points 100 milliseconds, 105 milliseconds, 110 milliseconds, 115 milliseconds, and 120 milliseconds are 2.5, 2.8, 3.0, 3.1, and 3.3, respectively. The sampling values of channel B at the same time points are 2.6, 2.9, 3.0, 3.2, and 3.4. The system iterates through each channel in turn, extracting the amplitude value at the corresponding time from the interpolated and time-aligned sampling data. The measurement values of all channels are then spliced vertically along the time dimension to generate a synchronized sampling matrix. Each matrix cell corresponds to the measurement value of a channel at a specific time point. Each row represents a unified time point, while each column corresponds to a different channel. During the matrix generation process, the system performs a one-to-one comparison of the timestamp fields of all channels, eliminating time deviations caused by sampling delays or signal loss, and retaining time points where all channels have complete sample values. The construction of the synchronous sampling matrix relies on the dynamic sampling configuration interval generated by the preceding module. This information contains information such as each channel's current sampling state, transition window, and valid time period. The system uses this sampling window as a benchmark for selecting synchronization points. In the above example, the system constructs a set of channel-synchronized records based on data from five time points between 100 milliseconds and 120 milliseconds for subsequent amplitude and trend difference analysis.

[0106] The trend difference extraction submodule calls the channel synchronous sampling matrix to calculate the sampling amplitude difference of multiple channels at the same time node. Combined with the sampling value change direction of each channel in a continuous time period, it extracts the trend change direction difference between channels and generates a trend amplitude difference structure group.

[0107] The channel synchronous sampling matrix constructed in the previous step is called, and the system begins to perform trend difference extraction operations. First, the amplitude difference of the sampling values of different channels at each time node is calculated. Taking channel A and channel B as an example, at 100 milliseconds, the sampling values of the two channels are 2.5 and 2.6 respectively. The system subtracts the smaller value from the larger value and takes the absolute value, and the amplitude difference at this time point is 0.1. 105 milliseconds, 110 milliseconds, 115 milliseconds, and 120 milliseconds are processed in turn to obtain a set of amplitude difference sequences. Subsequently, the system enters the trend identification stage, and the difference between the sampling values of two consecutive sampling points in the time dimension of each channel is judged. If the latter sampling value is greater than the previous sampling value, it is recorded as an upward trend, represented by 1; if the latter is less than the former, it is recorded as a downward trend, represented by -1; if they are equal, it is recorded as stable, represented by 0. After processing the sample values of channel A ({2.5, 2.8, 3.0, 3.1, 3.3}), the trend sequence becomes {1, 1, 1, 1}. The sample values of channel B ({2.6, 2.9, 3.0, 3.2, 3.4}) also yield a trend sequence of {1, 1, 1, 1}. The system then compares the trend codes of the two channels over the same time period. A difference of 0 indicates a consistent trend, a difference of 2 indicates opposite trends (one rising and one falling), and a difference of 1 indicates stability in one channel while the other fluctuates. The system then extracts the trend difference for each time period and structurally binds it to the amplitude difference, creating a corresponding structure of trend and amplitude differences. This ultimately forms a trend-amplitude difference structure group.

[0108] The coupling feature construction submodule calls the trend amplitude difference structure group, performs time dimension ratio combination processing on the amplitude difference sequence and trend difference sequence between channels, constructs a structural feature sequence for describing the change relationship between channels, and generates a composite parameter coupling sequence;

[0109] After obtaining the trend amplitude difference structure group, the system performs ratio combination processing on the structural unit of each time node. Specifically, the system extracts the trend difference and the amplitude difference, and performs the ratio combination operation in sequence. In order to avoid the situation where the denominator is zero, the system adds a constant term 1 when processing the trend difference, and defines the ratio as the amplitude difference divided by the trend difference plus 1. For example, if the amplitude difference at a certain time point is 0.1 and the trend difference is 0, the calculated value is 0.1 divided by 1, that is, the result is 0.1; if the amplitude difference is 0.2 and the trend difference is 1, the result is 0.2 divided by 2, which is 0.1; if the trend difference is 2 and the amplitude difference is 0.1, the result is 0.1 divided by 3, which is about 0.033. The system records the result of the ratio operation as a new structural feature point, repeats the above calculation steps for each time node, and generates a sequence of structural feature values. The structural eigenvalue sequence represents the degree of coordination between the sampling value changes across different channels at each sampling moment. High values indicate large amplitude differences but consistent trends, while low values indicate small amplitude differences when there is directional conflict. The time series of structural eigenvalues will be used as input to subsequent models or as an auxiliary indicator for state decision-making. Ultimately, the system stores these combined values at all time nodes as structured data entries, organized chronologically into a structural sequence, and generates a composite parameter coupling sequence that describes the relative coupling of the changing quantities during the multi-channel signal linkage process.

[0110] See also Figure 5 , vector building blocks include:

[0111] The node extraction submodule calls the composite parameter coupling sequence to extract the state value changes of each parameter node in a continuous time period. According to the direction of state change, amplitude increment, and change frequency in adjacent time periods, the change direction, change amplitude, and change frequency are extracted to obtain node state feature information;

[0112] During the execution of the node extraction submodule, the composite parameter coupling sequence is first called to extract the state value changes of each measurement node within a continuous time period. At this time, the state value changes of each node are considered as a time series. For example, suppose the sampling data of node A within a continuous time period is: . This data sequence indicates how the measurement value of the node changes at each time point. Next, the state value change of this sequence is analyzed. First, the direction of state change, amplitude increment and change frequency in each time period are determined. Specifically, the direction of state change is determined by the positive or negative difference between the current time point and the previous time point. If the difference is positive, the direction is considered to be "+1"; if the difference is negative, the direction is considered to be "-1". For example, in the data sequence of node A, the measurement value from the 1st to the 2nd time point increases from 12 to 14, the difference is +2, so the direction is +1; the measurement value from the 2nd to the 3rd time point increases from 14 to 17, the difference is +3, so the direction is still +1. And so on, until all time points have been calculated, the final direction sequence is In addition, the change in amplitude is defined as the absolute difference between the current sample value and the previous sample value. For example, the change in amplitude from the 1st to the 2nd time point is , the amplitude change from the 2nd to the 3rd time point is , and so on, the amplitude change sequence is . Next, calculate the frequency of change. The frequency of change refers to the number of times the state direction changes within a given time window. Taking the direction change sequence of node A as an example, the state direction changes from +1 to -1 and from -1 to +1 twice, respectively, with a frequency value of 2. By extracting the state direction, amplitude and change frequency of node A in this time period, the state feature information of the node is obtained. This feature information contains three dimensions: direction, amplitude and frequency, and is structured and recorded in the form of triples to form a node state feature information with a consistent time interval. For example, the direction, amplitude and frequency data of node A are respectively 、 and frequency 2. This node state feature information serves as input for further processing and analysis in subsequent modules. Through this structured extraction, the system maintains time synchronization across multiple nodes, ensuring the temporal consistency of node data and providing the necessary data foundation for the subsequent vector generation submodule.

[0113] The vector generation submodule calls the node state feature information, maps each node state feature into a three-dimensional vector axis value, constructs the node state change vector, and generates a node state vector structure set;

[0114] The state linkage trends between nodes are assessed by analyzing the structure of node state vectors. First, the submodule processes multiple pairs of node state vectors. Each pair of node state vectors originates from different measurement points or sensors in the system, and their changes reflect the correlations between various parameters in the system. Based on this, the included angle between the state vectors of each pair of nodes is calculated. The size of the included angle reflects whether the change directions of the two nodes are consistent. A smaller angle indicates more similar change trends between the two nodes, while a smaller angle indicates less similarity. Next, a vector projection difference comparison is performed. For each pair of node state vectors, the system calculates the projection value of the node state vector along each dimension and then calculates the difference between the projections of the two nodes along the same dimension. This difference reveals the degree of consistency in the node state changes along that dimension. Specifically, smaller projection differences indicate more similar change characteristics between the nodes along that dimension, while smaller projection differences indicate greater differences. The state linkage trend coefficients for multiple node pairs are then calculated. The included angle and projection difference are calculated and then combined to obtain the final state linkage trend coefficient. The coefficient indicates the degree of linkage between different nodes. A higher coefficient indicates stronger linkage, while a lower coefficient indicates weaker linkage. These coefficients provide multi-dimensional data support, helping the system determine whether different nodes exhibit the same or similar trends during measurement. Through the operation of the linkage judgment submodule, the system effectively identifies correlations between multiple nodes and quantifies their linkage strength using trend coefficients. This process not only reveals the correlations between different parameters or measurement channels but also provides a scientific basis for system fault diagnosis and trend prediction. For example, if the linkage trend coefficients between nodes A and B increase significantly under certain operating conditions, this may indicate a strong linkage response between these two nodes under the current operating conditions, prompting system attention. This may indicate the presence of a strongly correlated physical phenomenon or potential fault. Ultimately, the calculated linkage trend coefficients for multiple nodes serve as the basis for system judgment and decision-making. These results not only provide essential data support for subsequent anomaly detection and prediction, but also, by comparing them with historical data, help the system assess whether the current state is within normal range and provide a reference for adjusting system operation.

[0115] The linkage judgment submodule calls the node state vector structure set and performs angle calculation and vector projection difference comparison on multiple node state vector pairs using the formula:

[0116] ;

[0117] Calculate the state linkage trend coefficients of multiple node pairs and obtain the state linkage trend coefficients of multiple nodes;

[0118] in, Representation node With node The state linkage trend coefficient, Representation node In the The projection value on the dimensional vector coordinates, Representation node In the The projection value on the dimensional vector coordinates, Representation node In the The projection value on the dimensional vector coordinates, Representation node In the The projection value on the dimensional vector coordinates, represents the number of dimensions of the state vector, Representation node The state change vector, Representation node The state change vector, Indicates the first node number in the state node pair, Indicates the comparison node number in the state node pair, Represents the vector dimension index used for difference calculation, Represents the vector dimension index used for amplitude total statistics;

[0119] To identify the state change relationships between multiple node pairs, we call upon the node state vector structure set. This involves pairing the three-dimensional state change vectors expressed by multiple nodes within each time period. For each set of node vectors, we perform angle calculations and projection difference comparisons. The angle reflects the consistency of the vector's direction, while the projection difference reflects the proximity of its amplitude and configuration. For example, consider three node vectors within a specific time period. Nodes A, B, and C each have state change components (change direction, amplitude, and frequency) along three dimensions, as shown in the table. The values have been dimensionalized.

[0120] Table 2 Vector dimension component representation table

[0121] ;

[0122] As shown in Table 2, the numerical distribution of nodes A, B, and C in the three state feature dimensions (direction, amplitude, and frequency) provides basic data for subsequent vector angle calculation and trend coefficient determination.

[0123] According to the values in Table 2, the state change vectors of nodes A and B can be calculated as follows:

[0124] ;

[0125] Calculate its inner product:

[0126] ;

[0127] The module lengths are:

[0128] ;

[0129] Then the cosine of the angle is:

[0130] ;

[0131] At the same time, the vector projection difference calculation is performed, and the projection differences of each dimension of A and B are:

[0132] ;

[0133] Then the sum of the projection differences is:

[0134] ;

[0135] The sum of the amplitudes of each dimension is:

[0136] ;

[0137] The final formula is as follows:

[0138] ;

[0139] The results show that the directions of the state change vectors of nodes A and B are similar, the difference in amplitude is small, and the linkage trends are highly similar. The state linkage trend coefficient is 0.818, which is used to construct the trend correlation matrix between nodes. All node pairs can be executed according to this method. Representation node With node The state linkage trend coefficient reflects the consistency of the two state behaviors in direction and magnitude. Representation node In the The projection value on the dimensional coordinate (direction, amplitude, frequency), Representation node The projection value on the same dimension, 、 The same principle is used for the total amplitude normalization calculation. is the number of dimensions of the state vector, usually three-dimensional, 、 is the vector form of node state change, 、 is the number index of the state node, 、 are the dimension indices involved in the projection difference calculation and amplitude summation, respectively. By introducing the angle cosine and the vector projection difference normalization coefficient, the consistency and intensity of state behavior are coupled and evaluated as a single trend coefficient, improving the comprehensive judgment of node state covariation. The multi-node state linkage trend coefficient is used to measure the coordination of state changes between different measurement nodes and indicates the strength of the coordinated trend between multiple node state changes. Higher values indicate more consistent changes and stronger linkage between these nodes. Refer to the node vector values shown in Table 2. The trend matrix between all nodes can be quickly derived for use in subsequent modules.

[0140] See also Figure 6 , the event encapsulation module includes:

[0141] The vector difference identification submodule calls the multi-node state linkage trend coefficient to obtain the state change vectors of multiple parameter nodes. By analyzing the directional differences of the state vectors between nodes, it detects abnormal measurement events and generates abnormal event identification results.

[0142] First, we need to obtain state change vectors for multiple measurement channels and use these vectors to reflect the state changes of each channel over time. A state change vector is an array containing state values at multiple time points. By analyzing these state values, we can identify the dynamic change trend of each channel. During the execution of this module, we compare the state values within each channel, specifically detecting the directional differences between the state vectors at each measurement point, to identify nodes with significant changes. For example, suppose that during a certain time period, the state vector of channel 1 is (0.2, 0.5, 0.8, 1.0, 1.2) while the state vector of channel 2 is (0.1, 0.4, 0.7, 1.1, 1.5). The difference between the two state values is significant and the direction of change is significantly different. This difference may indicate the occurrence of an abnormal event. In this case, we set a threshold to identify these nodes with significant differences. These nodes are considered trigger points for an abnormal event and are marked as candidate data for the abnormal event. Furthermore, the directional differences between these state vectors are quantified by calculating the vector angle or magnitude of directional change. When the vector difference exceeds a set threshold, the system identifies that node as a marker for an abnormal measurement event. For example, if the threshold is set to 0.3, when the state vector difference exceeds this value, it is determined that an abnormality has occurred at that moment. Therefore, by analyzing the directional differences in the state change vectors of multiple channels, it is possible to locate which measurement data within a specific time period exhibit significant abnormal fluctuations, providing critical data support for subsequent abnormal event identification. These identified abnormal nodes provide important evidence for subsequent steps, laying the foundation for further tracking and path reconstruction of abnormal events, ultimately generating abnormal event identification results.

[0143] The abnormal event path reconstruction submodule extracts the parameter type and timestamp data corresponding to the target abnormal event based on the abnormal event identification results, and reconstructs the evolution path of the abnormal event according to the timestamp sequence to generate event evolution path data;

[0144] The system extracts relevant parameter types and timestamp data from the abnormal event identifiers. By analyzing the timestamps of each abnormal event and its corresponding state change data, it identifies and reconstructs a complete abnormal event path. For example, suppose that during a measurement, the states of channels 1, 2, and 3 change dramatically at certain time points, with the timestamps of these changes occurring at 5 seconds, 10 seconds, 15 seconds, and 20 seconds, respectively. These timestamps constitute the abnormal event's evolutionary path. During path reconstruction, the system first arranges the timestamps of these events in chronological order, thereby establishing a temporally continuous, gradually evolving event sequence. During this process, the system reorganizes the state values corresponding to each timestamp, combining the state changes of multiple channels to ensure that the state data at each time point is accurately mapped to the event path. For example, if the event path data is [5 seconds, 10 seconds, 15 seconds, 20 seconds], the data at each time point (e.g., the state value of channel 1 at 5 seconds is 0.2, at 10 seconds is 0.5, and so on) is extracted one by one and combined in chronological order to ultimately reconstruct a complete abnormal event evolutionary path. This reconstruction process provides a clear timeframe and detailed data at each measurement point, allowing a clear overview of the entire process from the occurrence to the evolution of an abnormal event and generating data on the event's evolution path. This data provides strong support for subsequent event packaging and analysis, ensuring the accuracy and reliability of the entire process.

[0145] The event encapsulation and recording submodule calls the event evolution path data, encapsulates the measurement data according to the path sequence, and generates an event process encapsulation record;

[0146] By analyzing the sequence of timestamps extracted from the abnormal event path, the measurement data associated with these timestamps is retrieved one by one and packaged in chronological order. For example, assuming the event path is [5 seconds, 10 seconds, 15 seconds, 20 seconds], the corresponding measurement data at these times are: at 5 seconds, the state value of channel 1 is 0.2, the state value of channel 2 is 0.1, and the state value of channel 3 is 0.3; at 10 seconds, the state value of channel 1 is 0.5, the state value of channel 2 is 0.4, and the state value of channel 3 is 0.6; at 15 seconds, the state value of channel 1 is 0.8, the state value of channel 2 is 0.7, and the state value of channel 3 is 1.0; at 20 seconds, the state value of channel 1 is 1.0, the state value of channel 2 is 1.1, and the state value of channel 3 is 1.3. This data is extracted one by one in the order of timestamps and constructed into a complete event flow record. Ultimately, the packaged data is recorded in a unified format for easy subsequent use and analysis. For example, an event flow record might look like this: [5 seconds: [0.2, 0.1, 0.3], 10 seconds: [0.5, 0.4, 0.6], 15 seconds: [0.8, 0.7, 1.0], 20 seconds: [1.0, 1.1, 1.3]]. This encapsulated record includes not only the time the event occurred but also the status values of each associated channel. This encapsulation method makes event data more structured, facilitating storage, transmission, and subsequent analysis. Especially in complex monitoring and maintenance processes, these encapsulated records effectively support rapid data query and backtracking, providing a convenient tool for subsequent fault diagnosis and event tracking. The core of this process is to ensure the timeliness and accuracy of data, ultimately generating an encapsulated event flow record.

[0147] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A multifunctional measurement and detection system, characterized in that: The system comprises: The time alignment module obtains a sequence of continuous sampling values from each measurement channel, calculates the rate of change between adjacent sampling points, extracts the time position of the peak node, compares the time offset of the peak node in each channel, and performs time alignment on the data to generate a heterogeneous data alignment sequence. The analog-to-digital conversion module calls the heterogeneous data alignment sequence, extracts the change amplitude of the measurement data in a continuous period and calculates the rate fluctuation characteristics, adjusts the resolution parameters of the analog-to-digital conversion in real time according to the rate change, and generates a dynamic sampling configuration interval; The data coupling module calls the dynamic sampling configuration interval, obtains the sampling values of multiple measurement channels at the same time node, performs ratio combination using the amplitude difference and the trend change value, constructs a parameter sequence describing the fluctuation relationship between the channels, and obtains a composite parameter coupling sequence; The vector construction module calls the composite parameter coupling sequence, analyzes the state value changes of each parameter node in a continuous time period, extracts the change direction, change amplitude and continuous change frequency, constructs the node state change vector, and combines the angle calculation and vector projection length comparison to obtain the multi-node state linkage trend coefficient.

2. The multifunctional measurement and detection system according to claim 1, characterized in that: The heterogeneous data alignment sequence includes the time alignment channel number, the key node timestamp, and the interpolated sampling value; the dynamic sampling configuration interval includes the rate threshold segment, the resolution parameter level, and the configuration index number; the composite parameter coupling sequence is specifically an amplitude difference sequence, a trend change value sequence, and a trend amplitude ratio; the multi-node state linkage trend coefficient includes a state change direction vector, a state change amplitude vector, and a change frequency characteristic value.

3. The multifunctional measurement and detection system according to claim 1, characterized in that: The time alignment module includes: The rate identification submodule obtains the continuous sampling value sequence of each measurement channel, extracts the numerical difference between adjacent sampling points, performs difference operation on adjacent sampling values in the continuous sampling segment within each channel, calculates the change rate, locates the change rate peak node in each channel, and generates a set of key nodes for rate change; The offset comparison submodule calls the rate change key node set, extracts the timestamp data of the peak nodes in the channel, arranges the key node times in each channel in sequence according to the time dimension, identifies the time offset between key nodes in the same time period, and generates the offset time interval; The node interpolation submodule calls the offset time interval and uses the formula: ； Calculate and reconstruct sampling values, perform time alignment on the measurement data of multiple channels through data interpolation, and generate heterogeneous data alignment sequences; in, is the reconstructed sample value of the interpolation point, is the known sample value before the interpolation point, is the known sample value after the interpolation point, is the sampling value The corresponding timestamp, is the sampling value The corresponding timestamp, is the target time position of the interpolation point on the time axis, is the index number of the previous sampling point in the time interval where the interpolation point is located. is the index number of the next sampling point in the time interval where the interpolation point is located. The index number of the current interpolation target point on the time axis.

4. The multifunctional measurement and detection system according to claim 3, characterized in that: The analog-to-digital conversion module includes: The amplitude extraction submodule calls the heterogeneous data alignment sequence to calculate the amplitude variation range of the sampling value of each channel within a preset time window in real time to obtain the continuous amplitude variation interval value; The rate fluctuation submodule calls the continuous amplitude change interval value, calculates the sampling change rate of each channel in the time window, extracts the difference change between the rates of adjacent time periods, compares the rate fluctuation frequency and the difference interval in multiple time periods, establishes a rate change ratio group, and generates a sampling rate fluctuation characteristic factor; The resolution configuration submodule calls the sampling rate fluctuation characteristic factor and uses the formula: ； Calculate the resolution parameter adjustment value of analog-to-digital conversion and generate a dynamic sampling configuration interval; in, Adjust the resolution parameter value for analog-to-digital conversion. The first The rate value of the sampling point, is the average value of all rate values in the jump section, is the fluctuation frequency per unit time in the jump section, For the The difference in velocity direction change within a time period, is the maximum amplitude of the sampling rate change within the jump section, is the mean of the measured amplitudes before and after the segment, is the peak value of the measured amplitude in this section, is the number of sampling points participating in the rate fluctuation statistics in the jump section, The number of time periods participating in the direction difference calculation within the jump section, is the index number of the rate sampling point, The index number of the segment participating in the velocity direction difference.

5. The multifunctional measurement and detection system according to claim 4, characterized in that: The data coupling module includes: The channel data acquisition submodule calls the dynamic sampling configuration interval, extracts the synchronous sampling values of multiple measurement channels at the same time node, and generates a channel synchronous sampling matrix; The trend difference extraction submodule calls the channel synchronous sampling matrix, calculates the sampling amplitude difference of multiple channels at the same time node, combines the sampling value change direction of each channel in a continuous time period, extracts the trend change direction difference between channels, and generates a trend amplitude difference structure group; The coupling feature construction submodule calls the trend amplitude difference structure group, performs time dimension ratio combination processing on the amplitude difference sequence and trend difference sequence between channels, constructs a structural feature sequence for describing the change relationship between channels, and generates a composite parameter coupling sequence.

6. The multifunctional measurement and detection system according to claim 5, characterized in that: The vector building block includes: The node extraction submodule calls the composite parameter coupling sequence to extract the state value changes of each parameter node in a continuous time period, and extracts the change direction, change amplitude, and change frequency according to the direction of the state change, the amplitude increment, and the change frequency in adjacent time periods to obtain the node state feature information; The vector generation submodule calls the node state feature information, maps each node state feature into a three-dimensional vector axis value, constructs a node state change vector, and generates a node state vector structure set; The linkage judgment submodule calls the node state vector structure set and performs angle calculation and vector projection difference comparison on multiple node state vector pairs using the formula: ； Calculate the state linkage trend coefficients of multiple node pairs and obtain the state linkage trend coefficients of multiple nodes; in, Representation node With node The state linkage trend coefficient, Representation node In the The projection value on the dimensional vector coordinates, Representation node In the The projection value on the dimensional vector coordinates, Representation node In the The projection value on the dimensional vector coordinates, Representation node In the The projection value on the dimensional vector coordinates, represents the number of dimensions of the state vector, Representation node The state change vector, Representation node The state change vector, Indicates the first node number in the state node pair, Indicates the comparison node number in the state node pair, Represents the vector dimension index used for difference calculation, Indicates the vector dimension index used for amplitude total statistics.

7. The multifunctional measurement and detection system according to claim 1, characterized in that: The system further comprises: The event encapsulation module calls the multi-node state linkage trend coefficient, detects abnormal measurement events by identifying the directional differences of the state vectors between multiple parameter nodes, extracts the corresponding parameter types and timestamps, reconstructs the abnormal event evolution path according to the timestamp sequence, encapsulates the measurement data using the path sequence, and generates an event process encapsulation record; The event process encapsulation record specifically refers to the exception parameter type, exception timestamp, and event path number.

8. The multifunctional measurement and detection system according to claim 7, characterized in that: The event encapsulation module includes: The vector difference identification submodule calls the multi-node state linkage trend coefficient to obtain the state change vectors of multiple parameter nodes, detects abnormal measurement events by analyzing the direction differences of the state vectors between the nodes, and generates an abnormal event identification result; The abnormal event path reconstruction submodule extracts the parameter type and timestamp data corresponding to the target abnormal event based on the abnormal event identification result, and reconstructs the evolution path of the abnormal event according to the timestamp sequence to generate event evolution path data; The event encapsulation and recording submodule calls the event evolution path data, encapsulates the measurement data according to the path sequence, and generates an event process encapsulation record.

Images