Multi-parameter integrated acquisition terminal

Abstract

The multi-parameter integrated acquisition terminal provided in this application dynamically configures and adaptively adjusts the data acquisition tasks for the monitored object through the signal conditioning and acquisition module in collaboration with the local intelligent processing module, thereby improving the intelligence level of data acquisition by the terminal deployed at the local edge. Furthermore, based on the characteristics of the monitored device, it constructs fused features by calculating the correlation between primary features in the synchronous real-time data stream after task adjustment, rather than simply fusing individual primary features directly. This allows the fused feature vector and summary uploaded by the local intelligent processing module to be directly used by maintenance personnel, reducing the data analysis pressure on the backend monitoring system and further improving the intelligence level of data acquisition by edge devices.

Description

Technical Field

[0001] This application relates to the field of data acquisition technology, specifically to a multi-parameter integrated acquisition terminal. Background Technology

[0002] Online monitoring and early warning of the operating status of power equipment such as distribution cabinets, transformers, and power systems have become crucial for ensuring power supply reliability. Online monitoring and early warning rely heavily on data acquisition terminals deployed at various nodes to collect multiple parameters, including voltage, current, temperature, humidity, and vibration intensity.

[0003] Currently, data acquisition terminals on the market are mainly divided into two categories. One type is dedicated data acquisition devices with single functions, such as temperature detectors and power quality analyzers. These require independent installation for different parameters, resulting in complex on-site wiring, high costs, and difficulty in maintenance. The other type is integrated terminals that integrate multiple sensors. However, most of these only physically centralize the interfaces and power supplies of multiple sensors. Each channel inside the terminal works independently, and the acquisition strategy remains unchanged. They cannot intelligently adjust the monitoring focus according to the actual operating status of the monitored object, nor can they perform multi-parameter correlation analysis and preliminary fault diagnosis at the edge. All collected data still needs to be uploaded to the back-end monitoring system, which makes it difficult to meet the monitoring system's requirements for real-time and intelligent data acquisition and analysis. Summary of the Invention

[0004] This application provides a multi-parameter integrated acquisition terminal to achieve adaptive adjustment of data acquisition strategies for different monitoring objects, complete multi-parameter data fusion and feature extraction locally, thereby improving the data analysis efficiency of the back-end monitoring system and reducing the analysis pressure.

[0005] To achieve this objective, the following technical solution is adopted in this application: A multi-parameter integrated data acquisition terminal is provided, including... The signal conditioning and acquisition module dynamically configures the signal type, range and sampling rate of each channel according to the pre-loaded or remotely distributed device configuration description file, forming an initial acquisition task set for the monitored object and executing data acquisition. The local intelligent processing module is connected to the signal conditioning and acquisition module. It is used to perform real-time analysis on the initial data stream acquired by the signal conditioning and acquisition module, and dynamically generate at least one sub-task when a trigger condition is identified. It also instructs the signal conditioning and acquisition module to interrupt or adjust the initial task and prioritize the execution of the sub-task. The local intelligent processing module is also used to extract a fusion feature vector representing the current state of the monitored device from each real-time data stream synchronously acquired by the signal conditioning and acquisition module, and to upload data to the backend monitoring system when the fusion feature vector is abnormal.

[0006] Preferably, the method for forming the initial acquisition task set through dynamic configuration includes the following steps: A1, Load the device configuration description file preset for the monitored object and dynamically reconstruct hardware resources; A2, based on the hardware resources and configuration file content dynamically reconstructed for each data acquisition interface, generates an initial acquisition task for each interface to acquire data and adds it to the initial acquisition task set.

[0007] Preferably, the method by which the local intelligent processing module performs real-time analysis of the initial data stream and identifies the triggering conditions includes the following steps: B1, identify abnormal data in each of the preliminary data streams received from each interface, the identification method is as follows: For the received temperature flow data, calculate the temperature change rate, and then determine whether the temperature change rate exceeds a preset allowable temperature change rate threshold. If so, the temperature flow data is determined to be abnormal data. For the received vibration flow data, calculate the short-time standard deviation of vibration and determine whether the short-time standard deviation of vibration exceeds the preset allowable short-time standard deviation threshold. If so, the vibration flow data is determined to be abnormal data. For the received voltage and current data, calculate the voltage range or current range, and determine whether the voltage range or the current range exceeds the corresponding preset allowable range threshold. If so, the voltage and current data are determined to be abnormal data. B2, search for the correlation between various types of abnormal data, and match the trigger conditions bound to the searched correlation from the trigger condition library.

[0008] Preferably, the method by which the local intelligent processing module generates the subtask and instructs its priority execution includes the following steps: C1, Search the rule base for the task execution action bound to the triggering condition; C2, search for available interfaces to perform the task execution actions; C3 determines whether each currently used interface executing the initial data collection task can execute all the searched task execution actions. If so, each task execution action is encapsulated into a subtask and distributed to the corresponding in-use interface, and the in-use interface is instructed to execute the subtask first. If not, the task execution actions that cannot be executed by the currently used interface are randomly assigned to the available interface that can be executed, and the interface is instructed to execute them.

[0009] Preferably, the method for the local intelligent processing module to extract the fused feature vector includes the following steps: D1, extracts primary features from the received synchronous real-time data stream, including time-domain features and / or frequency-domain features; D2, cross-channel relationships are associated with the extracted primary features to construct fused features; D3, concatenates the constructed fusion features and primary features into the fusion feature vector.

[0010] Preferably, the method by which the local intelligent processing module identifies whether the fused feature vector is a data anomaly includes the following steps: E1, obtain the fusion feature historical health vector for the same monitored device; E2, perform a data consistency comparison between the extracted fusion feature vector and the fusion feature historical health vector. If the comparison is successful, the data of the fused feature vector is determined to be normal, and the fused feature historical health vector is updated according to the fused feature vector. If the comparison fails, the abnormal fused feature vector is analyzed for faults, and then the analysis result, together with the fused feature vector and the real-time data streams of each channel received synchronously, is uploaded to the backend monitoring system.

[0011] Preferably, in step E2, the method for determining whether the consistency comparison was successful is as follows: If the difference between the element values ​​of all corresponding elements in the two vectors being compared is within the preset threshold range, then the data consistency comparison between the two vectors is considered successful; otherwise, the comparison is considered to have failed.

[0012] Preferably, the method for fault analysis of the fused feature vector of anomalies includes the following steps: F1, which performs quantization calculations on outlier data in the fused feature vector; F2 matches the abnormal data quantification results from the policy library, matches the corresponding fault correction policy, and packages the quantification results together into a fault analysis result.

[0013] This application has the following beneficial effects: 1. By dynamically configuring and adaptively adjusting the data acquisition tasks for the monitored objects through the signal conditioning and acquisition module in collaboration with the local intelligent processing module, the intelligence level of data acquisition by the terminal deployed at the local edge is improved, making data acquisition more effective and efficient.

[0014] 2. Based on the characteristics of the monitored equipment, the fused features are constructed by calculating the correlation between primary features in the synchronous real-time data stream after task adjustment, rather than simply fusing the primary features directly. This allows the fused feature vector and summary uploaded by the local intelligent processing module to be directly used by operation and maintenance personnel, reducing the data analysis pressure on the backend monitoring system and achieving the goal of intelligent data collection from edge devices. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments of this application will be briefly described below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the structure of the multi-parameter integrated acquisition terminal provided in the embodiments of this application. Detailed Implementation

[0017] The technical solution of this application will be further described below with reference to the accompanying drawings and specific embodiments.

[0018] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual images. They should not be construed as limiting the scope of this application. To better illustrate the embodiments of this application, some components in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0019] In the accompanying drawings of the embodiments of this application, the same or similar reference numerals correspond to the same or similar components. In the description of this application, it should be understood that if terms such as "upper," "lower," "left," "right," "inner," and "outer" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting this application. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0020] In the description of this application, unless otherwise expressly specified and limited, the term "connection" or similar designation indicating a connection between components should be interpreted broadly. For example, it can refer to a fixed connection, a detachable connection, or an integral part; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can refer to the internal communication between two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0021] The multi-parameter integrated data acquisition terminal provided in this application includes: The signal conditioning and acquisition module dynamically configures the signal type, range and sampling rate of each channel according to the pre-loaded or remotely distributed device configuration description file, forming an initial acquisition task set for the monitored object and executing data acquisition. The local intelligent processing module is connected to the signal conditioning and acquisition module. It is used to perform real-time analysis on the initial data stream acquired by the signal conditioning and acquisition module. When a trigger condition is identified, it dynamically generates at least one subtask and instructs the signal conditioning and acquisition module to interrupt or adjust the initial task and prioritize the execution of the subtask.

[0022] The following describes the process by which the signal conditioning and acquisition module dynamically configures itself based on the device configuration description file to form the initial acquisition task: Taking the monitoring of the moving and stationary contacts of the isolating trolley in a 10kV switchgear as an example, the goal is to monitor the overall temperature rise of the isolating trolley under normal conditions and to quickly identify local overheating and vibration caused by suspected poor contact when abnormal conditions occur.

[0023] First, load the pre-set device configuration description file for the switchgear isolation truck from the cloud or local machine to configure hardware resources. An example of the file content is as follows: Monitoring object: 10kV switchgear isolating handcart; Channel configuration: [{Physical interface: CH1, Signal type: AC current, Range: 0-100A, Sampling rate: 1kHz}; {Physical interface: CH2, Signal type: Temperature, Range: 0-150℃, Sampling rate: 1Hz}; {Physical interface: CH3, Signal type: Temperature, Range: 0-150℃, Sampling rate: 1Hz}; {Physical interface: CH4, Signal type: Vibration acceleration, Range: ±50g, Sampling rate: 10Hz}]. After receiving the device configuration description file, the signal conditioning and acquisition module performs configuration actions (hardware resource reconfiguration) according to the file: For example, for the CH1 interface, the control analog switch connects the physical interface to the current transformer conditioning circuit, and sets the gain range of the programmable gain amplifier to the 0-100A range described in the file for CH1, and the sampling rate to 1kHz.

[0024] For the CH2 and CH3 interfaces, connect these two physical interfaces to the resistance measurement bridge and the ADC conditioning circuit, and set the range to 0-150℃ and the sampling rate to 1Hz.

[0025] For the CH4 interface, connect the physical interface, for example, to the charge amplifier conditioning circuit, and configure it with a range of ±50g and a sampling rate of 10Hz.

[0026] After configuration, the terminal forms an initial acquisition task set including at least one initial acquisition task. For example, based on the example above, the initial acquisition task A is: to acquire temperature values ​​every 1 second through the CH2 and CH3 interfaces, and to acquire vibration acceleration every 100 milliseconds through the CH4 interface, for monitoring the load operating status trend; the initial acquisition task B is: to record a 1-second current waveform acquired through the CH1 interface every 10 minutes, for analyzing the load status.

[0027] After generating the initial data collection task for each interface, each dynamically configured interface performs data collection according to the task content and transmits the initial data stream to the local intelligent processing module in real time.

[0028] After receiving the initial data streams from each contact, the local intelligent processing module performs real-time analysis. The analysis methods include: For the received temperature stream data, such as the temperature stream data received from the CH2 and CH3 interfaces mentioned above, the temperature change rate, such as the temperature change rate per minute, is calculated and compared with a preset allowable temperature change rate threshold to identify abnormal data. For example, when the calculated temperature change rate exceeds the preset allowable temperature change rate threshold, the temperature stream data received at this time is marked as "abnormal data".

[0029] For the received vibration flow data, calculate the short-time standard deviation of the vibration, such as the standard deviation of the vibration acceleration collected through the CH4 interface over 5 consecutive seconds. When this standard deviation exceeds the preset allowable short-time standard deviation threshold, the received 5 seconds of vibration flow data is marked as abnormal data.

[0030] For the received voltage or current data, the voltage range or current range is calculated. For example, the difference between the maximum and minimum voltage values ​​in the voltage data received over a continuous 5-second period is taken as the voltage range. When the voltage range exceeds the allowable range threshold, the voltage flow data is determined to be abnormal data.

[0031] After identifying anomalous data in the initial data streams of various types, the local intelligent processing module determines whether the triggering conditions for dynamically generating subtasks have been met. The identification method is as follows: Search for the relationships between different types of abnormal data and match the trigger conditions bound to the searched relationships from the trigger condition library.

[0032] For example, suppose that during operation, contact phase C (connected to the CH3 interface) in the switchgear begins to overheat abnormally due to increased contact resistance. The local intelligent processing module monitors that the temperature change rate of the temperature flow data collected by CH3 has exceeded the allowable temperature change rate threshold of 1℃ / min for three consecutive cycles. Simultaneously, it analyzes that there is a sudden increase in the short-term standard deviation of vibration in the vibration flow data transmitted by the CH4 interface, lasting over 100ms. According to the preset rules: when the CH3 interface is connected to the resistance measurement bridge and ADC conditioning circuit in the switchgear, and the CH4 interface is connected to the charge amplifier conditioning circuit in the switchgear, if the abnormal behavior of the CH3 temperature flow data is: the temperature change rate exceeds the allowable temperature change rate threshold of 1℃ / min for three consecutive cycles, and almost simultaneously (e.g., within 2 seconds), and the abnormal behavior of the CH4 vibration flow data is: a sudden increase in the short-term standard deviation of vibration lasting 80ms or more, then this sudden increase in the short-term standard deviation of vibration is determined to be caused by slight deformation or abnormal discharge due to local overheating of devices connected to the CH3 interface, such as contact phase C. At this point, it was identified that the sudden increase in the short-term standard deviation of vibration at the CH4 interface was related to the abnormal temperature flow data at the CH3 interface. This relationship is the correlation between the abnormal temperature flow data and the abnormal vibration flow data.

[0033] Subsequently, based on the pre-defined associations and binding relationships between trigger conditions, the identified association-bound trigger conditions are matched from the trigger condition library. For example, given the temporal correlation between the abnormally rapid rise in temperature and sudden vibration mentioned above, if this association is pre-bound to the trigger condition "suspected early signs of poor contact," then based on this binding relationship, the trigger condition "suspected early signs of poor contact" is matched.

[0034] The method by which the local intelligent processing module dynamically generates subtasks based on the identified triggering conditions includes the following steps: C1 searches the rule base for the task execution action that triggers the condition binding; For example, for the trigger condition of "suspected early signs of poor contact in the contactor," the task execution actions bound to the rule base may include "high-frequency temperature capture of abnormal points" and "high-frequency spectrum analysis of vibration." The local intelligent processing module will then generate the first sub-task of high-frequency temperature capture for the task execution action of "high-frequency temperature capture of abnormal points," for example, based on preset sub-task generation rules for high-frequency temperature capture. The generated first sub-task might be: target sampling object: C-phase contact of a 110kV switchgear; sampling rate increased by 10 times, such as increasing the original 1Hz sampling rate to 10Hz, and sampling duration adjusted to 60 seconds.

[0035] C2 searches for available interfaces capable of performing the task. For example, if both CH3 and CH5 interfaces in the signal conditioning and acquisition module can perform the high-frequency temperature capture task of the C-phase contact of the 110kV switchgear, then CH3 and CH5 are searched as available interfaces.

[0036] C3 determines whether each currently used interface executing the initial data collection task can perform all the searched task execution actions. If so, each task execution action is encapsulated into a corresponding subtask and distributed to the corresponding in-use interface, and the in-use interface is instructed to execute the received subtask first; for example, the CH3 interface is instructed to execute the first subtask in the example above. If not, the task execution actions that cannot be executed by the currently used interface are randomly assigned to an available interface that can be executed, and the interface is instructed to execute them.

[0037] In this embodiment, the local intelligent processing module is also used to extract a fusion feature vector representing the current state of the monitored device from each real-time data stream synchronously acquired by the signal conditioning and acquisition module.

[0038] The following example, taking a 10kV vacuum circuit breaker as the monitored device, illustrates the process of extracting and fusing feature vectors by the local intelligent processing module to assess the health of the mechanical characteristics of the tripping operation of the monitored device.

[0039] Upon receiving a "break" command (either in the initial acquisition task or a sub-task after adjustment) or detecting the start of the break current, the terminal automatically triggers a high-speed synchronous acquisition task. The signal conditioning and acquisition module records synchronous waveforms of coil current (range, for example, 0-10A, sampling rate, 50kHz) and vibration signal (range, for example, ±100g, sampling rate, 50kHz) for a duration longer than 10 milliseconds, using the same time reference.

[0040] Then, the local intelligent processing module extracts primary features from the received synchronous real-time data stream, including time-domain features and / or frequency-domain features; Taking the received coil current waveform and vibration acceleration waveform as examples, the method for extracting primary features by the local intelligent processing module is explained: For example, by analyzing the coil current waveform, the electrical timing characteristics of the coil current can be extracted, including the action start time characteristic: the moment T0 when the current rises above a threshold (e.g., 10% of the peak value) is recorded as the action start time; the current peak value characteristic: the maximum current value Ip in the waveform; and the main current pulse width characteristic: the duration Tw from the current rising to the threshold (e.g., 10% of the peak value). The output electrical timing characteristics can be expressed as Fe=[T0, Ip, Tw].

[0041] By analyzing the vibration acceleration waveform, mechanical event characteristics of the 10kV vacuum circuit breaker are extracted, including the timing characteristics of the main vibration event: the moment Tv with the strongest vibration energy is found through envelope detection or energy detection; the impact intensity characteristics: the peak value Ap of the vibration signal in the vicinity of the main vibration event, such as within an adjacent 2ms interval; and the impact spectrum characteristics: a fast Fourier transform is performed on the period before and after the main vibration event (e.g., within 1ms) to extract the dominant frequency Fr as the impact spectrum characteristics. The output mechanical event characteristics are expressed as: Fm=[Tv, Ap, Fr].

[0042] Then, the local intelligent processing module performs cross-channel relationship association on the extracted primary features to construct fused features.

[0043] In this application, the local intelligent processing module does not simply concatenate Fe and Fm as exemplified above to form a fusion feature. Instead, it calculates the correlation between primary features based on the characteristics of the monitored device, then constructs the fusion feature, and finally forms the fusion feature vector.

[0044] Continuing with the Fe and Fm examples above, the local intelligent processing module discovers that Fe contains a primary feature T0 and Fm contains a primary feature Tv. Based on the preset characteristics of the 10kV vacuum circuit breaker, for example, the core of the 10kV vacuum circuit breaker has a time delay from startup to mechanical impact, and the time delay Δt is the difference between Tv and T0. Based on this preset characteristic, the module identifies a relationship between the two primary features T0 and Tv. Then, based on this recorded characteristic, Δt is calculated as the fusion feature of these two primary features.

[0045] Similarly, the fusion feature k = Ap / Ip is constructed.

[0046] Finally, the constructed fused features and primary features are concatenated into a fused feature vector.

[0047] For example, the primary features Ip and Tw in Fe, the primary feature Fr in Fm, and the constructed fusion features Δt and k are concatenated into a fusion feature vector Ff = [Ip, Tw, Δt, k, Fr].

[0048] In this application, the local intelligent processing module also identifies whether the fused feature vector is abnormal. Only when an anomaly is detected in the fused feature vector will the real-time data stream of each channel be uploaded to the backend monitoring system. The method by which this module identifies whether there is data anomaly in the fused feature vector includes the following steps: E1, obtain the historical health vector of fused features for the same monitored device; For example, the fusion feature history health vector stored for this 10kV vacuum circuit breaker is: [22.8A (health threshold of Ip), 17.8ms (health threshold of Tw), 24.1ms (health threshold of Δt), 0.88 (health threshold of k), 1200Hz (health threshold of Fr)].

[0049] E2 compares the extracted fusion feature vector with the obtained historical health vector of the same fusion feature vector for data consistency. If the comparison is successful, the data of the fused feature vector is determined to be normal, and the fused feature vector is used as a new sample to update the historical health vector of the fused feature vector. If the comparison fails, the abnormal fusion feature vector is analyzed for faults, and then the analysis, together with the fusion feature vector and the real-time data streams from each channel received synchronously, is uploaded to the backend detection system.

[0050] For example, the extracted fusion feature vector Ff = [Ip, Tw, Δt, k, Fr] = [23.5A, 18.2ms, 25.6ms, 0.85, 1250Hz]. The obtained historical health vector of the fusion feature is [22.8A (health threshold of Ip), 17.8ms (health threshold of Tw), 24.1ms (health threshold of Δt), 0.88 (health threshold of k), 1200Hz (health threshold of Fr)]. According to the preset consistency comparison rules, for example, the difference between the two values ​​is calculated for the same parameter. If the difference is lower than the preset difference threshold, the parameter value in the fusion feature vector is determined to be normal; otherwise, it is determined to be abnormal. For example, 23.5A - 22.8A = 0.7A, which is less than the preset difference threshold of 0.8A, so the Ip parameter value in the fusion feature vector is determined to be normal. For example, if 25.6ms - 24.1ms = 1.5ms is greater than the preset abnormal threshold of 1ms for Δt, then the parameter value of Δt in the fused feature vector is determined to be abnormal.

[0051] When all parameter values ​​in the fused feature vector Ff are determined to be normal, the consistency comparison between the two vectors is considered successful. At this point, it is not necessary to upload real-time streaming data to the backend monitoring system, but the fused feature vector's historical health vector needs to be updated using this fused feature vector as a new sample. The update method is as follows: The earliest parameter value is filtered out, and the newly added parameter value of the same parameter in the currently extracted fusion feature vector is added. Then, the average value of each parameter value in the time series is calculated and updated in the historical health vector of the fusion feature. For example, 22.8A in the historical health vector of the fusion feature mentioned above is the average value of Ip values ​​within the historical 10 seconds from the current time point. Assuming the sampling frequency is once per second, it includes 10 Ip values. At this time, the Ip value extracted from the fusion feature vector now replaces the Ip value collected in the first second of the history, and then the average value is recalculated as the update value of the Ip value in the historical health vector of the fusion feature.

[0052] In this application, the method for fault analysis based on the fused feature vector of anomalies is illustrated below: For example, if anomalies in the Δt and k values ​​are detected through the consistency comparison described above, the growth rate or decline rate of the corresponding parameter values ​​in the historical health vector of the fused feature is calculated. For instance, if Δt increases by 6% and k decreases by 3%, then based on these quantified anomaly features, a pre-configured fault correction strategy is matched to ultimately form the fault analysis result. For example, if the strategy library records that when the Δt of a 10kV vacuum circuit breaker increases abnormally and the k value decreases abnormally, it is recommended to check the lubrication status of the mechanical transmission components. The strategy is then matched from the strategy library using the aforementioned anomaly features and packaged together with the anomaly features to form the fault analysis result.

[0053] In summary, the multi-parameter integrated acquisition terminal provided in this application dynamically configures and adaptively adjusts the data acquisition tasks for the monitored object through the signal conditioning and acquisition module in collaboration with the local intelligent processing module, thereby improving the intelligence level of data acquisition by the terminal deployed at the local edge. Furthermore, based on the characteristics of the monitored device, it constructs fused features by calculating the correlation between primary features in the synchronous real-time data stream after task adjustment, rather than simply fusing individual primary features directly. This allows the fused feature vector and summary (fault analysis results) uploaded by the local intelligent processing module to be directly used by maintenance personnel, reducing the data analysis pressure on the backend monitoring system and further improving the intelligence level of data acquisition by edge devices.

[0054] It should be stated that the above-described specific embodiments are merely preferred embodiments and technical principles applied in this application. Those skilled in the art should understand that various modifications, equivalent substitutions, and variations can be made to this application. However, such variations, as long as they do not depart from the spirit of this application, should be within the scope of protection of this application. Furthermore, some terminology used in this application's specification and claims is not limiting but merely for ease of description.

Claims

1. A multi-parameter integrated acquisition terminal, characterized in that, Comprising. The signal conditioning and acquisition module dynamically configures the signal type, range and sampling rate of each channel according to the preloaded or remotely issued device configuration description file, forms an initial acquisition task set of the monitored object and performs data acquisition; The local intelligent processing module is connected with the signal conditioning and acquisition module, used for performing real-time analysis on the preliminary data stream collected by the signal conditioning and acquisition module, dynamically generating at least one subtask when a trigger condition is identified, and instructing the signal conditioning and acquisition module to interrupt or adjust the initial task and preferentially execute the subtask; The local intelligent processing module is also used for extracting a fusion feature vector representing the current state of the monitored device from each real-time data stream synchronously collected by the signal conditioning and acquisition module, and uploading data to the backend monitoring system when the fusion feature vector is abnormal.

2. The multi-parameter integrated acquisition terminal according to claim 1, characterized in that, The method for dynamically configuring the initial acquisition task set includes the following steps: A1, loading the device configuration description file preset for the monitored object and performing hardware resource dynamic reconstruction; A2, generating an initial acquisition task for each data acquisition interface according to the hardware resource dynamically reconstructed for each data acquisition interface and the content of the configuration file, and adding the initial acquisition task to the initial acquisition task set.

3. The multi-parameter integrated acquisition terminal of claim 1, wherein, The method for the local intelligent processing module to perform real-time analysis on the preliminary data stream and identify the trigger condition includes the following steps: B1, identifying abnormal data in each preliminary data stream received from each interface, and the identification method is as follows: For the received temperature flow data, the temperature change rate is calculated, and then it is judged whether the temperature change rate exceeds the preset allowed temperature change rate threshold, if yes, the temperature flow data is determined as abnormal data; For the received vibration flow data, the vibration short-time standard deviation is calculated, and it is judged whether the vibration short-time standard deviation exceeds the preset allowed vibration short-time standard deviation threshold, if yes, the vibration flow data is determined as abnormal data; For the received voltage and current data, the voltage range or current range is calculated, and it is judged whether the voltage range and the current range exceed the corresponding preset allowed range threshold, if yes, the voltage and current data is determined as abnormal data; B2, searching for the correlation between abnormal data of each type, and matching the trigger condition bound by the searched correlation from the trigger condition library.

4. The multi-parameter integrated acquisition terminal of claim 3, wherein, The method for the local intelligent processing module to generate the subtask and instruct the preferential execution includes the following steps: C1, searching the task execution action bound by the trigger condition from the rule library; C2, searching the available interface for executing the task execution action; C3, judging whether each in-use interface currently executing the initial acquisition task can execute all the searched task execution actions, If yes, each task execution action is encapsulated as the subtask and distributed to the corresponding in-use interface, and the in-use interface is instructed to preferentially execute the subtask; If no, the task execution action that the in-use interface cannot execute is randomly allocated to the available interface that can execute, and the available interface is instructed to execute.

5. The multi-parameter integrated acquisition terminal of claim 1, wherein, The method for the local intelligent processing module to extract the fusion feature vector includes the following steps: D1, extracting primary features from the received synchronous real-time data stream, including time domain features and / or frequency domain features; D2, correlating the extracted primary features with cross-channel relationships to build fusion features; D3, concatenating the built fusion features and primary features into the fusion feature vector.

6. The multi-parameter integrated acquisition terminal of claim 1, wherein, The method for the local intelligent processing module to identify whether the fusion feature vector is data abnormal includes steps of: E1, obtaining a fusion feature historical health vector for the same monitored device; E2, performing data consistency comparison between the extracted fusion feature vector and the fusion feature historical health vector, if the comparison is successful, determining that the data of the fusion feature vector is normal, and updating the fusion feature historical health vector according to the fusion feature vector; if the comparison fails, performing fault analysis on the abnormal fusion feature vector, and then uploading the analysis result, the fusion feature vector and the real-time data stream of each channel received synchronously to the back-end monitoring system.

7. The multi-parameter integrated acquisition terminal of claim 6, wherein, In step E2, the method for determining whether the consistency comparison is successful is: if the difference between the element values of all corresponding elements in the two vectors is within the preset corresponding difference threshold range, it is determined that the data consistency comparison of the two vectors is successful, otherwise it is determined that the comparison fails.

8. The multi-parameter integrated acquisition terminal of claim 6, wherein, The method for performing fault analysis on the abnormal fusion feature vector includes steps of: F1, quantitatively calculating the abnormal data in the fusion feature vector; F2, matching the abnormal data quantization result from the strategy library, matching the corresponding fault correction strategy and packaging the quantization result into the fault analysis result.

Images