Equipment data processing method and device for service-oriented collaborative integration

Abstract

The application relates to a kind of equipment data processing method and device for service-oriented collaborative integration, and the method comprises the following steps: collecting industrial sensing data from complex equipment, extracting relevant data according to application target, obtaining several dimension waveform data;Phase alignment is carried out on the data, window interception is carried out from the marker point, and abnormal waveform data segment is obtained;Entropy feature is calculated, and the dimension data of the maximum entropy value is extracted as the representative of multi-dimensional waveform data according to the dimension;The statistical characteristics of the representative of waveform data are extracted, and the data is standardized according to the characteristic value;In Cartesian coordinate system, linear transformation is carried out on the processed data representative, the offset degree of actual waveform data relative to standard waveform data at the same time is calculated, and the standardized representation of waveform data is realized.The method provided by the application can realize the standardization, availability and comparability of complex equipment data through efficient reduction processing of equipment data, and support the collaborative integration of equipment manufacturing and maintenance system.

Description

Technical Field

[0001] This invention relates to the field of equipment data processing, and more specifically to a service-oriented collaborative integration-based equipment data processing method and apparatus for reducing the amount of equipment operation data and integrating service systems. Background Technology

[0002] Complex equipment generates a large amount of production and status data throughout its lifecycle. Scientific and standardized management of this data is the foundation and prerequisite for equipment data informatization. During service collaboration and integration, differing platform architectures and inconsistent data source formats lead to poor data sharing integration and low availability between services, severely hindering interoperability and operability. Therefore, designing data integration services can improve the utilization efficiency of equipment data, unify heterogeneous data sources in complex equipment, and provide efficient support for the integration of multiple systems in complex equipment, the reduction of service communication, and the sharing of data resources.

[0003] Furthermore, these complex equipment data typically exhibit characteristics of high throughput, high dimensionality, correlation, and temporal sequence. Existing industrial time-series data processing methods mostly focus on the study of single-dimensional periodic sequences or multi-dimensional uncorrelated sequences, failing to effectively mine and utilize the correlation information of multi-dimensional sequences. Although some machine learning-based models and algorithms can process multi-dimensional time-series data, the computation process usually lacks interpretability, and its reliability is difficult to meet the needs of equipment data detection and processing with diverse patterns and changing operating conditions, resulting in low analytical performance for complex equipment data.

[0004] In the daily operation of complex equipment, sensors and other data acquisition devices are commonly used to monitor and analyze abnormal conditions in real time, with most of the sensed information manifested as waveform data. In the real world, many data formats can be converted into a superposition of sine and cosine waveform components using methods such as Fourier transform. The waveform signal of normally operating equipment is usually a standard sine wave. By analyzing standard waveform data, the characteristics of the original data can be reconstructed more accurately, leading to more comprehensive conclusions. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a service-oriented collaborative integration method and apparatus for equipment data processing. This invention extracts multidimensional data by leveraging correlations, rationally utilizes the inherent characteristics of the data, achieves efficient reduction of multidimensional related data, standardizes complex equipment data, saves bandwidth, and improves the interaction efficiency of multiple systems, thereby supporting the collaborative integration of equipment manufacturing and maintenance systems.

[0006] The technical solution of this invention is as follows:

[0007] A service-oriented collaborative integration-based equipment data processing method, comprising:

[0008] Industrial sensor data is collected from complex equipment, and relevant data is extracted according to the target to obtain multidimensional waveform data;

[0009] Phase alignment is performed on the multidimensional waveform data, and a window is truncated starting from the marked point to obtain the abnormal waveform data segment;

[0010] Calculate the entropy characteristics of multidimensional waveform data, and extract the data of the dimension with the maximum entropy value as the representative of the multidimensional waveform data.

[0011] Statistical features represented by multidimensional waveform data are extracted, and equipment data is standardized based on the extracted feature values.

[0012] In the Cartesian coordinate system, a linear transformation is performed on the standardized multidimensional waveform data to calculate the degree of deviation of the actual waveform data relative to the standard waveform data at the same moment, thereby realizing the standardized representation of the waveform data.

[0013] Furthermore, the process of acquiring industrial sensor data from complex equipment, extracting relevant data based on the target, and obtaining multidimensional waveform data includes:

[0014] Capture and record transient waveform data of different trigger types from the intelligent acquisition device of complex equipment;

[0015] Perform quality screening on the waveform recording sample data and remove erroneous sample data with inconsistent structures from the waveform recording sample data;

[0016] The distribution of recorded waveform samples under each trigger type was statistically analyzed, and samples of trigger types with concentrated distribution were selected for further analysis.

[0017] Based on the trigger type and data characteristics, waveform data of several dimensions were extracted from the above-selected sample data. The relevant m-dimensional waveform data extracted are as follows:

[0018] X = [X] 1 ,X 2 ,X 3 ,…,X m ]

[0019] Furthermore, the step of performing phase alignment on the multidimensional waveform data and truncating a window starting from the marked point to obtain abnormal waveform data segments includes:

[0020] First, select alignment markers to perform initial phase alignment on the waveform data. The initial phase is defined as follows: The positive extreme points of the first cycle are selected as marker points, and the resulting marker vector is:

[0021]

[0022] In the formula, Mark i Let represent the index vector indicating the alignment of m-dimensional waveform data in the i-th waveform recording sample, and n represent the number of waveform recording samples. According to the marker vector Mark... i The element index value is used to align the marker points of each dimension of the waveform data.

[0023] Then, the duration of abnormal features in the data is statistically analyzed to determine the truncation length. The waveform data segment is then truncated to obtain a multidimensional abnormal waveform data segment under the waveform recording sample. The abnormal data length is plotted and statistically analyzed as α, and the segment is truncated from the marked point to the subsequent α data points to obtain the multidimensional abnormal waveform data segment under the waveform recording sample. The following conditions must be met:

[0024]

[0025] in, This represents the subscript index vector when the V-dimensional waveform data in the i-th waveform recording sample is aligned.

[0026] Furthermore, the calculation of the entropy characteristics of the multidimensional waveform data, and the extraction of data from the dimension containing the maximum entropy value as the representative of the multidimensional waveform data, includes:

[0027] First, using the following formula for information entropy, calculate the information entropy of each dimension of waveform data in all samples under the recorded waveform sample, that is:

[0028]

[0029] In the formula, This represents the waveform data value of dimension V at time j. This represents the frequency of the data value at time j among all data in this dimension (i.e., the number of occurrences / the total number of data). Let V represent the set of waveform data at all times in dimension V of the i-th sample; calculate the probability of each data point, and then obtain the information entropy value of the waveform data in each dimension, i.e.:

[0030]

[0031] Finally, the waveform data of the dimension containing the maximum entropy value is selected as the representative of the multidimensional waveform data of this waveform recording sample.

[0032] h i (x)=max H V (x)

[0033] Furthermore, the statistical characteristics represented by the multidimensional waveform data are extracted, including kurtosis, divergence, mean, and standard deviation. The equipment data is then standardized based on these characteristics. The formulas for calculating the mean μ, standard deviation σ, kurtosis, and divergence are as follows:

[0034]

[0035]

[0036]

[0037] Where, x j This represents the waveform data value at time j, and n represents the number of waveform samples recorded.

[0038] Then, based on the mean μ and standard deviation σ, the equipment data are processed into dimensionless quantities using the Z-score standardization method.

[0039]

[0040] Furthermore, the process of performing a linear transformation on the standardized data representation in the Cartesian coordinate system, calculating the degree of offset between the actual waveform data and the standard waveform data at the same moment, and realizing the standardized representation of the waveform data includes:

[0041] First, construct the waveform function for normal equipment operation based on information such as sampling frequency, amplitude, and phase under ideal conditions. Calculate the ideal data value f(t0) at each time step;

[0042] Then calculate the actual waveform data at the same time. Offset relative to standard waveform data f(t0) And use positive and negative signs to indicate the height of the actual waveform relative to the ideal waveform;

[0043] Finally, the waveform data points are transformed in the Cartesian coordinate system. With the standard waveform data as a reference, the y-axis component is represented as the offset of the actual waveform relative to the ideal waveform at that moment, dis(t0). The coordinates are then represented as (t0, dis(t0)).

[0044] Furthermore, the aforementioned equipment data processing methods are encapsulated into a data interface service to provide a unified data interface for service collaboration and integration. First, the data service is built using the Spring Cloud framework. Then, the equipment data processing methods are encapsulated using Java. Finally, data service standards, specifications, and invocation instructions are provided.

[0045] A service-oriented collaborative integration-oriented equipment data processing device, comprising:

[0046] The data acquisition module is used to collect industrial sensor data from complex equipment, extract relevant data according to the target, and obtain multidimensional waveform data.

[0047] The data capture module is used to perform phase alignment on multidimensional waveform data and capture a window starting from the marked point to obtain abnormal waveform data segments.

[0048] The data representative acquisition module is used to calculate the entropy characteristics of multidimensional waveform data and extract the data of the dimension with the maximum entropy value as the representative of the multidimensional waveform data.

[0049] The feature extraction module is used to extract statistical features represented by multidimensional waveform data and to standardize the data based on the extracted feature values.

[0050] The standardization representation module is used to perform a linear transformation on the standardized data representation in the Cartesian coordinate system, calculate the degree of deviation of the actual waveform data relative to the standard waveform data at the same moment, and realize the standardized representation of waveform data.

[0051] Furthermore, the aforementioned equipment data processing device for service-oriented collaborative integration also includes:

[0052] The encapsulation module is used to encapsulate the data acquisition module, data interception module, data representation acquisition module, feature extraction module, and standardized representation module into a data interface service, providing a unified data interface for service collaboration and integration.

[0053] Compared with the prior art, the present invention has the following advantages:

[0054] 1. The equipment data processing method and apparatus for service-oriented collaborative integration provided by the present invention reduces the volume of high-throughput equipment data through a series of operations such as window truncation of abnormal data and extraction of data in the dimension where the maximum entropy value is located, thereby effectively improving data quality and reducing computational complexity.

[0055] 2. Compared with traditional technologies, the service-oriented collaborative integration equipment data processing method and device provided by this invention extracts global features of the time-series waveform data itself, such as kurtosis, divergence, mean, and standard deviation, and effectively integrates the correlation relationships of multi-dimensional data for comprehensive analysis. This approach ensures the comprehensiveness of feature extraction and has strong universality in waveform data processing.

[0056] 3. The equipment data processing method for service-oriented collaborative integration provided by this invention performs a linear transformation on waveform equipment data in the Cartesian coordinate system in the time domain space to study its fluctuation relative to the standard sine curve. This approach amplifies abnormal fluctuations in the data while also achieving data standardization, consistency, and comparability, forming a standard specification system for data representation. This is of great significance for achieving high-quality data integration in the process of service-oriented collaborative integration.

[0057] 4. The equipment data processing method and apparatus for service-oriented collaborative integration provided by the present invention designs a data integration service to unify heterogeneous data sources in complex equipment, which can support the collaborative integration of equipment manufacturing and maintenance systems and provide efficient service support for the comprehensive integration of information technology and data resource sharing of complex equipment. Attached Figure Description

[0058] Figure 1 This is a flowchart of a service-oriented collaborative integration equipment data processing method according to an embodiment of the present invention;

[0059] Figure 2 This is a flowchart of step 1 in a service-oriented collaborative integration equipment data processing method according to an embodiment of the present invention: collecting industrial sensor data from complex equipment, extracting relevant data according to the target, and obtaining several-dimensional waveform data.

[0060] Figure 3 This is a flowchart of step 2 in a service-oriented collaborative integration equipment data processing method according to an embodiment of the present invention: performing phase alignment on the data and extracting a window from the marked point to obtain an abnormal waveform data segment.

[0061] Figure 4 The abnormal waveform data segment obtained after phase alignment in an embodiment of the present invention is shown, where y1 to y6 represent the six-dimensional side data of UA, UB, UC, IA, IB, and IC, respectively.

[0062] Figure 5 This is a flowchart of step 3 in a service-oriented collaborative integration equipment data processing method according to an embodiment of the present invention: calculating the entropy characteristics of multidimensional waveform data and extracting the data of the dimension with the maximum entropy value as the representative of the multidimensional waveform data.

[0063] Figure 6 This is a flowchart of step 4 in a service-oriented collaborative integration equipment data processing method according to an embodiment of the present invention: extracting statistical features represented by waveform data, including four aspects: kurtosis, divergence, mean, and standard deviation, and standardizing the data based on these feature values.

[0064] Figure 7This is a flowchart of step 5 in a service-oriented collaborative integration equipment data processing method in an embodiment of the present invention: In the Cartesian coordinate system, a linear transformation is performed on the processed data representative, and the degree of offset of the actual waveform data relative to the standard waveform data at the same moment is calculated to realize the standardized representation of the waveform data.

[0065] Figure 8 The waveform data before and after linear transformation in this embodiment of the invention are shown, where y1 and y2 represent the data before and after the transformation, respectively.

[0066] Figure 9 This is a flowchart of step 6 in a service-oriented collaborative integration equipment data processing method according to an embodiment of the present invention: encapsulating the above equipment data processing method into a data interface service to provide a unified data interface for service collaborative integration. Detailed Implementation

[0067] This invention provides a service-oriented collaborative integration method and apparatus for equipment data processing, which realizes standardized processing of equipment anomaly data, reduces computational complexity, and is of great significance for achieving high-quality integration of system services.

[0068] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below through specific implementations and in conjunction with the accompanying drawings. To accurately describe the content of this invention, the following terms and meanings will first be explained.

[0069] Complex equipment refers to the technical equipment required for simple and expanded reproduction in industries such as machinery and manufacturing. It is characterized by system interconnectivity, diverse complex failures, inherent deficiencies, frequent state changes, high cost, high capacity generation costs, high level of application, and high usage intensity.

[0070] Service integration: Through structured integrated deployment systems, computer network technology and software technology, distributed information resources and scattered information services are flexibly connected into an organic whole according to certain information organization principles, user needs and service objectives, and activities of resources and services from different systems are seamlessly integrated.

[0071] Data services refer to information technology-driven services that provide data collection, data transmission, data storage, data processing (including calculation, analysis, visualization, etc.), data exchange, and data destruction, encompassing all forms of data evolution.

[0072] Transient waveform data: During equipment operation, under certain triggering conditions, the meter automatically records the power quality data output by the complex equipment for a period of time before and after the trigger. Voltage, current, and changes in equipment switching states are all relevant electrical quantities. Possible faults include: system charge imbalance, voltage imbalance, and line short circuits and grounding.

[0073] Kurtosis and Skewness: Kurtosis and skewness are statistical characteristics unique to time-series waveform data. Kurtosis measures the height of the peak value of the statistical distribution at the mean, reflecting the sharpness of the waveform peaks; skewness measures the direction and degree of skewness of the statistical distribution, reflecting the degree of asymmetry in the waveform distribution.

[0074] Entropy characteristics: Entropy is an indicator used to describe the uncertainty of a system. The larger the entropy value, the more chaotic the system; the smaller the entropy value, the more stable the system. In this patent, the magnitude of the entropy value reflects the degree of deviation of the abnormal waveform from the standard waveform.

[0075] Example:

[0076] In one embodiment of the present invention, the complex equipment selected is the generator on a large wind turbine. As a component with a high failure rate, the generator has the need for condition monitoring and fault diagnosis during operation. Detecting the power quality indicators of the generator set is an important way to realize the self-regulation and monitoring of the wind farm.

[0077] like Figure 1 As shown, one embodiment of the present invention provides a service-oriented collaborative integration equipment data processing method, comprising the following steps:

[0078] Step 1: Collect industrial sensor data from complex equipment, extract relevant data according to the target, and obtain multidimensional waveform data;

[0079] Step 2: Align the data by phase and extract the abnormal waveform data segment by starting from the marked point;

[0080] Step 3: Calculate the entropy feature of the multidimensional waveform data, and extract the data of the dimension with the maximum entropy value as the representative of the multidimensional waveform data;

[0081] Step 4: Extract the statistical characteristics represented by the waveform data, including kurtosis, divergence, mean, and standard deviation, and standardize the data based on these characteristic values;

[0082] Step 5: In the Cartesian coordinate system, perform a linear transformation on the processed data representative, calculate the degree of deviation of the actual waveform data relative to the standard waveform data at the same moment, and realize the standardized representation of the waveform data;

[0083] Step 6: Encapsulate the equipment data processing method into a data interface service to provide a unified data interface for service collaboration and integration.

[0084] like Figure 2 As shown, in one embodiment, step 1 above—collecting industrial sensor data from complex equipment, extracting relevant data according to the target, and obtaining multi-dimensional waveform data—specifically includes:

[0085] Step 1.1: Capture and record transient waveform data of different trigger types from the intelligent acquisition device of the complex equipment.

[0086] This embodiment collects terminal waveform data from wind turbine generators. Continuous sampling begins when a fault occurs in the wind turbine, with a sampling time of 3 seconds and a sampling density of 1600 points / second. The data within the first 100ms of sampling records the normal operating state of the wind turbine just before the fault. This embodiment contains 503 waveform samples, recording transient data such as trigger state, voltage, current, and frequency before and after the fault. Each side has a data dimension of 4800. In this embodiment, waveform sample 09152224Record.dat is used as an example for analysis.

[0087] Step 1.2: Remove erroneous sample data with inconsistent structures from the waveform recording samples.

[0088] First, all recorded sample data were screened for quality, and one erroneous sample was excluded, resulting in 502 samples.

[0089] Step 1.3: Statistically analyze the distribution of recorded waveform samples under each trigger type, and select the trigger type samples with the most concentrated distribution for analysis.

[0090] Then, the recorded data on the triggering conditions of the samples were read, and it was found that there were three types of triggering: low voltage triggering, current imbalance triggering, and communication triggering. Among them, the number of samples with abnormalities caused by current imbalance was the largest, at 497.

[0091] Step 1.4: Based on the trigger type, extract the three-dimensional current waveform data from the sample data selected above.

[0092] In this embodiment of the invention, the waveform recording sample data includes waveform recording state data and seven-dimensional side data: voltage (UA, UB, UC), current (IA, IB, IC), and frequency (F), with each dimension having a length of 4800. The matrix X (4800×497×7) formed by the samples is represented as follows:

[0093] X = [X] UA ,X UB ,X UC ,X IA ,X IB ,X IC ,X F ]

[0094] in,

[0095]

[0096]

[0097] X V This represents a matrix consisting of all waveform sample vectors recorded on the side V, with a size of 4800×497; This represents a column vector consisting of the single-sample side face V, with a size of 4800×1. N represents the total number of samples, and m represents the data length of each dimension. The three-dimensional current waveform data (IA, IB, IC) related to anomalies are read and retained for subsequent targeted analysis.

[0098] X′=[X IA ,X IB ,X IC ]

[0099] like Figure 3 As shown, in one embodiment, step 2 above—phase alignment of the data and window truncation starting from the marked point to obtain the abnormal waveform data segment—specifically includes:

[0100] Step 2.1: Select alignment markers to perform initial phase alignment on the waveform data.

[0101] In this embodiment of the invention, for each dimension of waveform data, its initial phase is defined as follows: The positive extreme points of the first cycle are selected as marker points, and the resulting marker vector is:

[0102]

[0103] In the formula, Mark i Let represent the index vector of the three-dimensional waveform data alignment in the i-th waveform recording sample. Then, for each dimension of data IA, IB, IC, we have: According to the tag vector The element index values ​​are used to align the marker points of each dimension of the waveform data. For example, Mark1 = [20, 26, 9].

[0104] Step 2.2: Statistically analyze the duration of abnormal features in the data to determine the truncation length, truncate the waveform data segment, and obtain the three-dimensional abnormal waveform data segment under each waveform recording sample.

[0105] like Figure 4 As shown, the outlier data in the plot is of length α, and is extracted from the marked point to the next α data points, satisfying the following conditions:

[0106]

[0107] In this embodiment of the invention, α = 1200, and for the IA, IB, and IC three-dimensional data, the intervals [20~1220], [26~1226], and [9~1209] are respectively extracted. Taking the IA dimension as an example, the extracted waveform data is as follows:

[0108]

[0109] like Figure 5 As shown, in one embodiment, step 3 above: calculating the entropy feature of the multidimensional waveform data, and extracting the dimension data containing the maximum entropy value as the representative of the multidimensional waveform data, specifically includes:

[0110] Step 3.1: Calculate the entropy characteristic value of the three-dimensional abnormal waveform data segment under each waveform recording sample.

[0111] Taking IA-dimensional data as an example, The frequency P(d) of waveform data points appearing in integer data i Given the set {0.00, 0.03, 0.1, ..., 0.00}, the information entropy of the IA-dimensional data is obtained as follows: Right now:

[0112]

[0113] Similarly, the information entropy of other dimensions of data in this sample are obtained as follows: H V (x) = {0.452, 0.802, 0.511}.

[0114] Step 3.2: Select the dimension containing the maximum entropy value as the representative waveform data of the recorded sample.

[0115] h V (x)=max H V (x)

[0116] Similarly, h V (x) = max{0.452,0.802,0.511} = 0.802, therefore IB is taken as the representative of the three-dimensional waveform data of this example sample.

[0117] like Figure 6 As shown, in one embodiment, step 4 above involves extracting the statistical characteristics represented by the waveform data, including kurtosis, divergence, mean, and standard deviation. The data is then standardized based on these characteristic values, specifically including:

[0118] Step 4.1: Calculate the four statistical characteristics represented by the waveform data: kurtosis, divergence, mean, and standard deviation.

[0119] Calculate the kurtosis and skewness of the IB-dimensional data for this sample:

[0120]

[0121]

[0122] Step 4.2: Based on the above statistical characteristic values, the Z-score standardization method is used to process the equipment data into dimensionless quantities.

[0123]

[0124] like Figure 7 As shown, in one embodiment, step 5 above—performing a linear transformation on the processed data representation in the Cartesian coordinate system, calculating the offset of the actual waveform data relative to the standard waveform data at the same moment, and realizing the standardized representation of the waveform data—specifically includes:

[0125] Step 5.1: Construct the waveform function of the equipment during normal operation based on the sampling frequency, amplitude, phase and other information under ideal conditions, and calculate the ideal waveform data at each moment.

[0126] In this embodiment of the invention, A = 52, f = 50 Hz, All values ​​are taken under ideal conditions, and the constructed ideal waveform is: Where ω=2πf, substitute the value of time t0 to obtain the ideal data value f(t0) at that time.

[0127] Step 5.2: Calculate the degree of deviation of the actual waveform data relative to the standard waveform data at the same moment.

[0128] When a point A(5, 51.17) is randomly selected on the IA-dimensional data, at time t0 = 5, the actual waveform data value is... If the ideal data value is f(5) = 49.23, then the offset of the actual waveform relative to the ideal waveform can be expressed as:

[0129]

[0130] In this embodiment of the invention, positive and negative signs are used to represent the height of the actual waveform relative to the ideal waveform, and it is specified that when... When, the sign of dis(t0) is positive, when When t0, the sign of dis(t0) is negative. Therefore, dis(1) = 1.94.

[0131] Step 5.3: Redesign the meaning of the ordinate and perform coordinate transformation on the waveform data points in the Cartesian coordinate system.

[0132] In the Cartesian coordinate system, actual waveform data can be represented by coordinates. This means that the x-axis component represents time, and the y-axis component represents the actual waveform data value at that time. In this embodiment of the invention, the waveform data points in the Cartesian coordinate system are transformed. Using standard waveform data as a reference, the y-axis component is represented as the offset dis(t0) of the actual waveform at that time relative to the ideal waveform. The coordinates are then represented as (t0, dis(t0)), as shown below. Figure 8 As shown, point A has new coordinates (5, 1.94).

[0133] Step 5.4: Repeat steps 5.1 to 5.3 to achieve linear transformation of the equipment data, thereby obtaining a unified representation of the equipment waveform data in the Cartesian coordinate system.

[0134] like Figure 9 As shown, in one embodiment, step 6 above: encapsulating the above equipment data processing method into a data interface service to provide service developers with a standard, regulated, and stable data call interface, specifically including:

[0135] Step 6.1: Represent the data processing method as a service in the form of a programming language.

[0136] The data service is built using the Spring Cloud framework, and the above equipment data processing methods are encapsulated in Java.

[0137] Step 6.2: Provide data service standards, specifications, and instructions for use.

[0138] The data format is standardized: the input data format is the raw waveform recording data, and the output data format is the processed time-series data pair (t0, dis(t0)). Other services can obtain the data using the get operation. Statistical comparison of the same waveform recording sample shows that the ratio of the number of bytes after reduction processing to the number of bytes in the original data is approximately 1:30, yet it covers the key feature information of the waveform of that sample.

[0139] Another embodiment of the present invention provides an equipment data processing device for service-oriented collaborative integration, comprising:

[0140] The data acquisition module uses a dedicated electric meter sensor to collect industrial sensor data from complex equipment, extracts relevant data according to the target, and obtains multi-dimensional waveform data.

[0141] The data truncation module is used to perform phase alignment on multidimensional waveform data, analyze the data to select an appropriate truncation length, and perform window truncation starting from the marked point to obtain abnormal waveform data segments.

[0142] The data representative acquisition module is used to calculate the entropy characteristics of multidimensional waveform data and extract the data of the dimension with the maximum entropy value as the representative of the multidimensional waveform data.

[0143] The feature extraction module is used to extract four statistical features represented by multidimensional waveform data: kurtosis, divergence, mean, and standard deviation, and to standardize the data based on the extracted feature values.

[0144] The standardization representation module is used to perform a linear transformation on the standardized data representation in the Cartesian coordinate system, calculate the degree of deviation of the actual waveform data relative to the standard waveform data at the same moment, and realize the standardized representation of waveform data.

[0145] Optionally, the above-mentioned equipment data processing device for service collaborative integration further includes: an encapsulation module, used to encapsulate the data acquisition module, data interception module, data representation acquisition module, feature extraction module, and standardized representation module into a data interface service, providing a unified data interface for service collaborative integration.

[0146] For the specific implementation process of each module, please refer to the description of the method of the present invention above.

[0147] Another embodiment of the present invention provides an electronic device (computer, server, smartphone, etc.) including a memory and a processor, the memory storing a computer program configured to be executed by the processor, the computer program including instructions for performing the steps of the method of the present invention.

[0148] Another embodiment of the present invention provides a computer-readable storage medium (such as ROM / RAM, disk, optical disk) storing a computer program that, when executed by a computer, implements the various steps of the method of the present invention.

[0149] The equipment data processing method and apparatus for service-oriented collaborative integration provided by this invention reduces the volume of high-throughput equipment data through a series of operations such as window truncation of abnormal data and extraction of data in the dimension where the maximum entropy value is located, effectively improving data quality and reducing computational complexity.

[0150] Compared to traditional technologies, the service-oriented collaborative integration equipment data processing method and device provided by this invention extracts global features of time-series waveform data, such as kurtosis, divergence, mean, and standard deviation, and effectively integrates the correlations of multi-dimensional data for comprehensive analysis. This approach ensures the comprehensiveness of feature extraction while possessing strong universality in waveform data processing.

[0151] The equipment data processing method for service-oriented collaborative integration provided by this invention performs a linear transformation on waveform equipment data in the Cartesian coordinate system in the time domain to study its fluctuation relative to the standard sine curve. This approach amplifies abnormal fluctuations in the data while also achieving data standardization, consistency, and comparability, forming a standardized system for data representation. This is of great significance for achieving high-quality data integration in the process of service-oriented collaborative integration.

[0152] The equipment data processing method and apparatus for service-oriented collaborative integration provided by this invention are designed with data integration services to unify heterogeneous data sources in complex equipment. This enables the collaborative integration of equipment manufacturing and maintenance systems and provides efficient service support for the comprehensive integration of information technology and data resource sharing of complex equipment.

[0153] The above embodiments are provided merely for the purpose of describing the present invention and are not intended to limit the scope of the invention. The scope of the invention is defined by the appended claims. Various equivalent substitutions and modifications made without departing from the spirit and principles of the invention should be covered within the scope of the invention.

Claims

1. A service-oriented collaborative integration-based equipment data processing method, characterized in that, Includes the following steps: Industrial sensor data is collected from complex equipment, and relevant data is extracted according to the target to obtain multidimensional waveform data; Phase alignment is performed on the multidimensional waveform data, and a window is truncated starting from the marked point to obtain the abnormal waveform data segment; Calculate the entropy characteristics of multidimensional waveform data, and extract the data of the dimension with the maximum entropy value as the representative of the multidimensional waveform data. Statistical features represented by multidimensional waveform data are extracted, and the data is standardized based on the extracted feature values. A linear transformation is performed on the standardized multidimensional waveform data in the Cartesian coordinate system to calculate the degree of deviation of the actual waveform data relative to the standard waveform data at the same moment, thereby realizing the standardized representation of the waveform data. The process of aligning the phase of the multidimensional waveform data and extracting a window segment starting from the marked point to obtain the abnormal waveform data segment includes: First, select alignment markers to perform initial phase alignment on the waveform data, specifying the initial phase as... The positive extreme points of the first cycle are selected as marker points, and the resulting marker vector is: In the formula, Mark i This represents the index vector of the m-dimensional waveform data alignment in the i-th waveform recording sample, where n represents the number of waveform recording samples; according to the marker vector Mark i The element index value in the data is used to align the marker points of each dimension of the waveform data; Then, the duration of abnormal features in the data is statistically analyzed to determine the truncation length. The waveform data segment is then truncated to obtain a multidimensional abnormal waveform data segment under the waveform recording sample. The abnormal data length is plotted and statistically analyzed. The truncation is performed from the marked point to the subsequent α data points to obtain a multidimensional abnormal waveform segment under the waveform recording sample, which satisfies the following conditions: in, This represents the subscript index vector when the V-dimensional waveform data in the i-th waveform recording sample is aligned; The calculation of the entropy feature of the multidimensional waveform data, which involves extracting data from the dimension containing the maximum entropy value as the representative of the multidimensional waveform data, includes: First, using the following formula for information entropy, calculate the information entropy of each dimension of waveform data in all samples under the recorded waveform sample: In the formula, This represents the waveform data value of dimension V at time j. This represents the frequency of the data value at time j among all data in this dimension. Let V represent the set of waveform data at all times in dimension V of the i-th sample; calculate the probability of each data point, and then obtain the information entropy value of the waveform data in each dimension: Finally, the waveform data of the dimension containing the maximum entropy value is selected as the representative of the multidimensional waveform data of this waveform recording sample: h i (x)=maxH V (x) 2. The equipment data processing method for service-oriented collaborative integration according to claim 1, characterized in that, The process of collecting industrial sensor data from complex equipment, extracting relevant data based on the target, and obtaining multidimensional waveform data includes: Capture and record transient waveform data of different trigger types from the intelligent acquisition device of complex equipment; Remove erroneous sample data with inconsistent structures from the waveform recording samples; The distribution of recorded waveform samples under each trigger type is statistically analyzed, and samples of trigger types with concentrated distribution are selected for further analysis. Based on the trigger type and data characteristics, waveform data of several dimensions were extracted from the sample data selected above.

3. The equipment data processing method for service-oriented collaborative integration according to claim 1, characterized in that, The process involves extracting statistical features from the multidimensional waveform data, including kurtosis, divergence, mean, and standard deviation. The data is then standardized based on these features, specifically including: Calculate the four statistical characteristics represented by the waveform data: kurtosis, divergence, mean, and standard deviation. Based on the above statistical characteristics, the Z-score standardization method is used to process the equipment data into dimensionless quantities.

4. The equipment data processing method for service-oriented collaborative integration according to claim 1, characterized in that, The process of performing a linear transformation on the standardized multidimensional waveform data in the Cartesian coordinate system, calculating the offset of the actual waveform data relative to the standard waveform data at the same moment, and realizing the standardized representation of the waveform data includes: First, construct the waveform function for normal operation of the equipment based on the sampling frequency, amplitude, and phase under ideal conditions. Calculate the ideal data value f(t0) at each time step; Then calculate the actual waveform data at the same moment. Offset relative to standard waveform data f(t0) And use positive and negative signs to indicate the height of the actual waveform relative to the ideal waveform; Finally, the waveform data points are transformed in the Cartesian coordinate system. With the standard waveform data as a reference, the y-axis component is represented as the offset of the actual waveform relative to the ideal waveform at that moment, dis(t0). The coordinates are then represented as (t0, dis(t0)).

5. The equipment data processing method for service-oriented collaborative integration according to claim 1, characterized in that, The equipment data processing method is encapsulated into a data interface service to provide a unified data interface for service collaboration and integration, specifically including: Represent data processing methods as services in the form of programming languages; Provide data service standards, specifications, and usage instructions.

6. A service-oriented collaborative integration equipment data processing device employing the method described in any one of claims 1 to 5, characterized in that, include: The data acquisition module is used to collect industrial sensor data from complex equipment, extract relevant data according to the target, and obtain multidimensional waveform data. The data capture module is used to perform phase alignment on multidimensional waveform data and capture a window starting from the marked point to obtain abnormal waveform data segments. The data representative acquisition module is used to calculate the entropy characteristics of multidimensional waveform data and extract the data of the dimension with the maximum entropy value as the representative of the multidimensional waveform data. The feature extraction module is used to extract statistical features represented by multidimensional waveform data and to standardize the data based on the extracted feature values. The standardization representation module is used to perform a linear transformation on the standardized multidimensional waveform data in the Cartesian coordinate system, calculate the degree of deviation of the actual waveform data relative to the standard waveform data at the same moment, and realize the standardized representation of waveform data.

7. The equipment data processing device for service-oriented collaborative integration according to claim 6, characterized in that, Also includes: The encapsulation module is used to encapsulate the data acquisition module, data interception module, data representation acquisition module, feature extraction module, and standardized representation module into a data interface service, providing a unified data interface for service collaboration and integration.

8. An electronic device, characterized in that, It includes a memory and a processor, the memory storing a computer program configured to be executed by the processor, the computer program including instructions for performing the method of any one of claims 1 to 5.

Images