Hydraulic mechanism oil pump motor detection method and system based on exhaust device state

Abstract

The invention discloses a hydraulic mechanism oil pump motor detection method and system based on the state of an exhaust device, and relates to the technical field of hydraulic mechanism oil pump motor health detection.The method comprises the steps that clustering analysis is conducted on a full-life-cycle current curve of the exhaust device based on time duration alignment, and a current reference curve is obtained; calculating the current difference between the fault time point and the maintenance completion time point to construct a current difference curve; calculating the curve difference between the current reference curve and the current difference curve according to the fault type, and constructing a reference fluctuation relation corresponding to the fault type according to the environmental data; a state curve corresponding to the current exhaust device is obtained according to the reference fluctuation relation and the current reference curve, and a hydraulic mechanism oil pump motor state detection result is output according to the state curve and the real-time hydraulic mechanism oil pump motor current data. The method has the beneficial effects that the accuracy of the state of the oil pump motor of the hydraulic mechanism is improved, and fault feature learning of the oil pump motor is not needed.

Description

Technical Field

[0001] This application relates to the field of hydraulic mechanism oil pump motor health detection technology, and in particular to a method and system for detecting hydraulic mechanism oil pump motors based on the status of the exhaust device. Background Technology

[0002] In power systems, hydraulic pump motors are core power components of critical equipment such as circuit breakers. Their operating status directly determines the operational reliability of power equipment and the stability of power supply. Accurate status monitoring and fault early warning of these motors are crucial for ensuring the safe and efficient operation of power systems.

[0003] Some related technologies employ offline sampling inspection or periodic shutdown inspection, which not only fails to achieve real-time monitoring of motor operating status but also interrupts normal equipment operation processes, affecting the continuous power supply of the power system. Other related technologies use a single threshold comparison method to determine the motor status, relying solely on setting a fixed current threshold to define the presence of a fault. This makes it difficult to distinguish between current anomalies caused by motor-related faults and current anomalies caused by exhaust device malfunctions, leading to inaccurate fault tracing and failing to provide clear troubleshooting directions for operation and maintenance.

[0004] Furthermore, due to the high stability of the oil pump motor, there is a limited amount of fault sample data, making it impossible to learn the fault characteristics of the oil pump motor through deep learning, which further increases the difficulty of detecting the condition of the oil pump motor. Summary of the Invention

[0005] This application addresses the problem in existing technologies where the status detection of hydraulic mechanism oil pump motors is easily interfered with by exhaust devices. It provides a method and system for detecting hydraulic mechanism oil pump motors based on the status of exhaust devices. By using historical fault records of exhaust devices, which are more prone to failure, the common current changes corresponding to different fault types of exhaust devices are obtained. By checking whether the common current changes caused by exhaust device faults appear in the real-time hydraulic mechanism oil pump motor current data, it is determined whether the current change is caused by an abnormal status of the exhaust device, thereby improving the accuracy of the hydraulic mechanism oil pump motor status detection without the need for learning the fault characteristics of the oil pump motor.

[0006] To achieve the above technical objectives, this application provides a technical solution: a hydraulic mechanism oil pump motor detection method based on the status of an exhaust device, comprising the following steps: dividing the hydraulic mechanism oil pump motor current data according to the failure time point of the exhaust device, and obtaining several full life cycle current curves of the exhaust device; performing cluster analysis on the full life cycle current curves based on time length alignment to obtain a current reference curve corresponding to the failure type; calculating the current difference between the failure time point and the maintenance completion time point, and constructing a current difference curve; calculating the curve difference between the current reference curve and the current difference curve according to the failure type, and constructing a reference fluctuation relationship corresponding to the failure type based on environmental data; obtaining real-time hydraulic mechanism oil pump motor current data, obtaining the status curve corresponding to the current exhaust device according to the reference fluctuation relationship and the current reference curve, and outputting the hydraulic mechanism oil pump motor status detection result according to the status curve and the real-time hydraulic mechanism oil pump motor current data.

[0007] Furthermore, the step of dividing the hydraulic mechanism oil pump motor current data according to the failure time of the exhaust device and obtaining several full life cycle current curves of the exhaust device includes: taking the time from the completion time of the previous maintenance of the exhaust device to the time of the last failure of the exhaust device as a full life cycle, dividing the hydraulic mechanism oil pump motor current data, and obtaining the full life cycle current curve.

[0008] Furthermore, the step of performing cluster analysis on the full life cycle current curves based on time length alignment to obtain the current reference curve corresponding to the fault type includes: using the median time length of the full life cycle current curves of the same fault type as the time length alignment benchmark for that fault type; performing equal-length interpolation of the full life cycle current curves based on the time length alignment benchmark to obtain the corrected full life cycle current curves; calculating the curve similarity between the corrected full life cycle current curves using a dynamic time warping algorithm, and obtaining fault clusters based on the curve similarity and the clustering algorithm; and extracting common features from the fault clusters to construct the current reference curves.

[0009] Furthermore, the calculation of the current difference between the fault time point and the maintenance completion time point, and the construction of the current difference curve, includes: calculating the current difference based on the fault current data corresponding to the same fault time point and the maintenance completion current data, and constructing a current difference curve based on the current difference; wherein, the fault current data includes all current data on the day of the fault time point, and the maintenance completion current data includes all current data on the day of the maintenance completion time point.

[0010] Furthermore, the step of calculating the curve difference between the current reference curve and the current difference curve according to the fault type and constructing the reference fluctuation relationship corresponding to the fault type based on environmental data includes: retrieving the current reference curve according to the fault type corresponding to the current difference curve, calculating the curve difference between the current difference curve and the current reference curve, and recording it as the environmental fluctuation value at the corresponding fault time point; aggregating the environmental fluctuation values ​​of the same fault type and the corresponding environmental data, and constructing the reference fluctuation relationship based on the correlation between the environmental fluctuation values ​​and the environmental data.

[0011] Furthermore, the step of acquiring real-time hydraulic mechanism oil pump motor current data, obtaining the state curve corresponding to the current exhaust device based on the reference fluctuation relationship and the current reference curve, and outputting the hydraulic mechanism oil pump motor state detection result based on the state curve and the real-time hydraulic mechanism oil pump motor current data includes: acquiring real-time environmental data; obtaining the current environmental fluctuation value based on the real-time environmental data and the reference fluctuation relationship; obtaining the state curve corresponding to the current exhaust device based on the current environmental fluctuation value and the current reference curve; filtering out the interference current of the exhaust device in the real-time hydraulic mechanism oil pump motor current data based on the state curve to obtain the oil pump motor state current; and determining the hydraulic mechanism oil pump motor state based on the oil pump motor state current and a preset current threshold.

[0012] Furthermore, the step of acquiring real-time hydraulic mechanism oil pump motor current data, acquiring the state curve corresponding to the current exhaust device based on the reference fluctuation relationship and the current reference curve, and outputting the hydraulic mechanism oil pump motor state detection result based on the state curve and the real-time hydraulic mechanism oil pump motor current data includes: acquiring real-time hydraulic mechanism oil pump motor current data; if the real-time hydraulic mechanism oil pump motor current data is less than a preset idling current threshold, then outputting an idling alarm; if the real-time hydraulic mechanism oil pump motor current data is greater than or equal to the preset idling current threshold, then acquiring the state curve corresponding to the current exhaust device based on the reference fluctuation relationship and the current reference curve, and outputting the hydraulic mechanism oil pump motor state detection result based on the state curve and the real-time hydraulic mechanism oil pump motor current data.

[0013] Furthermore, the step of performing cluster analysis on the full lifecycle current curves based on time length alignment to obtain the current reference curve corresponding to the fault type includes: performing equal-length interpolation of the full lifecycle current curves based on the time length alignment benchmark to obtain the corrected full lifecycle current curves; calculating the sampling time similarity between the corrected full lifecycle current curves using a dynamic time warping algorithm; clustering the sampling time current data based on the sampling time similarity; using the cluster with the highest density as the fault cluster, reconstructing the current curves in the fault cluster as the fault curves based on consecutive sampling times; and constructing the current reference curve using the average current of the fault curves.

[0014] Furthermore, the step of filtering out the interference current of the exhaust device in the real-time hydraulic mechanism oil pump motor current data based on the state curve to obtain the oil pump motor state current includes: calculating the fault similarity of each state curve and the real-time hydraulic mechanism oil pump motor current data; using the state curve with a fault similarity greater than a preset fault similarity threshold to perform the filtering out of the interference current of the exhaust device in the real-time hydraulic mechanism oil pump motor current data to obtain the oil pump motor state current.

[0015] Another solution provided in this application is a hydraulic mechanism oil pump motor detection system based on the status of the exhaust device, used to implement the method described above, including: a data processing unit, used to divide the hydraulic mechanism oil pump motor current data according to the failure time point of the exhaust device, and obtain several full life cycle current curves of the exhaust device; a benchmark analysis unit, used to perform cluster analysis on the full life cycle current curves based on time length alignment, and obtain the current benchmark curve corresponding to the failure type; a difference analysis unit, used to calculate the current difference between the failure time point and the maintenance completion time point, and construct the current difference curve; a fluctuation analysis unit, used to calculate the curve difference between the current benchmark curve and the current difference curve according to the failure type, and construct the benchmark fluctuation relationship corresponding to the failure type based on environmental data; and a detection unit, used to acquire real-time hydraulic mechanism oil pump motor current data, acquire the status curve corresponding to the current exhaust device according to the benchmark fluctuation relationship and the current benchmark curve, and output the hydraulic mechanism oil pump motor status detection result according to the status curve and the real-time hydraulic mechanism oil pump motor current data.

[0016] The beneficial effects of this application are as follows: By dividing the oil pump motor current data of the entire life cycle of the exhaust device, cluster analysis is performed based on the fault time point to extract the current curve corresponding to the fault type. Taking advantage of the reproducible nature of the interference of exhaust device faults on oil pump motor current, common current features of different fault types are stripped away through cluster analysis. At the same time, the fluctuation of the degree of influence of faults on current under the current environment is constructed based on the current difference before and after maintenance. The interference of exhaust device status on motor current data is obtained by combining common current features and the fluctuation of the degree of influence brought by the environment. By filtering out the interference of exhaust device status in the current data, the detection accuracy of oil pump motor status is improved. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the hydraulic mechanism oil pump motor detection method based on the state of the exhaust device according to this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely one preferred embodiment of this application and are only used to explain this application. They do not limit the scope of protection of this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] like Figure 1 As shown in Embodiment 1 of this application, the hydraulic mechanism oil pump motor detection method based on the state of the exhaust device includes the following steps: Based on the failure time of the exhaust device, the hydraulic mechanism oil pump motor current data are divided to obtain the full life cycle current curves of several exhaust devices. Cluster analysis of the full life cycle current curves based on time length alignment is performed to obtain the current reference curve corresponding to the fault type; Calculate the current difference between the fault time point and the maintenance completion time point, and construct a current difference curve; The difference between the current reference curve and the current difference curve is calculated according to the fault type, and the reference fluctuation relationship corresponding to the fault type is constructed based on environmental data. The system acquires real-time hydraulic mechanism oil pump motor current data, obtains the current state curve corresponding to the current exhaust device based on the reference fluctuation relationship and current reference curve, and outputs the hydraulic mechanism oil pump motor state detection result based on the state curve and real-time hydraulic mechanism oil pump motor current data.

[0020] In this embodiment, by dividing the oil pump motor current data throughout the entire life cycle of the exhaust device, cluster analysis is performed based on the fault time point to extract the current curve corresponding to the fault type. Taking advantage of the reproducible nature of the interference of exhaust device faults on oil pump motor current, common current features of different fault types are stripped away through cluster analysis. At the same time, the fluctuation of the degree of influence of faults on current under the current environment is constructed based on the current difference before and after maintenance. The interference of exhaust device status on motor current data is obtained by combining common current features and the degree of influence fluctuation brought by the environment. By filtering out the interference of exhaust device status in the current data, the detection accuracy of oil pump motor status is improved.

[0021] Specifically, based on the failure time of the exhaust device, the hydraulic mechanism oil pump motor current data is divided, and the full life cycle current curves of several exhaust devices are obtained, including: The hydraulic mechanism oil pump motor current data is divided into two parts based on the time from the completion of the previous maintenance of the exhaust device to the time of the next failure of the exhaust device, and the current curve of the entire life cycle is obtained.

[0022] It is understood that the "full lifecycle" in this embodiment refers to the entire operating cycle of a single fault, with the initial installation time of the exhaust device to the time of the first fault as the first full lifecycle. Here, utilizing the conventional maintenance approach of maintenance personnel, when a fault occurs, maintenance personnel, while maintaining the exhaust device, will necessarily simultaneously check whether other components are likely to fail. If so, they will simultaneously maintain and record the components that are about to fail. Therefore, even if they belong to different fault types, there must be a period of normal operation between the two fault times of the exhaust device. Using the time from the completion of the previous maintenance of the exhaust device to the time of the next fault as a full lifecycle, it must include the process from normal operation to the occurrence of the fault, thus allowing the acquisition of the current change characteristics from normal to fault corresponding to the fault type.

[0023] Meanwhile, in this embodiment, the maintenance period is not included in the entire life cycle to avoid current fluctuations during the maintenance process and test data being used for fault feature identification, thus ensuring the accuracy of fault feature separation.

[0024] Cluster analysis of the full lifecycle current curves based on time length alignment yields current baseline curves corresponding to fault types, including: The median time length of the full life cycle current curve for the same fault type is used as the time length alignment benchmark for that fault type. Perform equal-length interpolation of the full life cycle current curve using a time-length alignment reference to obtain the corrected full life cycle current curve; The dynamic time warping algorithm is used to calculate the curve similarity between the corrected full life cycle current curves, and the fault clusters are obtained based on the curve similarity and clustering algorithm. Common features are extracted from fault clusters to construct current reference curves.

[0025] In this embodiment, the fault type corresponding to each fault time point is obtained based on the maintenance records. Time length alignment is performed based on the median time length of the full life cycle current curves of the same fault type, so that full life cycle current curves of different lengths can be clustered on the same benchmark. This avoids interference from excessively long or short period curves on the alignment scale, ensuring that the time length benchmark is consistent with the typical evolution cycle of the fault type. At the same time, dynamic time warping is used to calculate curve similarity, eliminating the interference of local feature offsets and ensuring the accuracy of the current benchmark curve construction.

[0026] Specifically, common feature extraction is performed on fault clusters to construct current reference curves, including: Calculate the average value of the curves in the fault cluster, and construct a current reference curve based on the average value of the curves.

[0027] Understandably, any existing similarity-based clustering algorithm can be used for clustering.

[0028] Furthermore, the current difference between the fault time point and the maintenance completion time point is calculated, and a current difference curve is constructed, including: The current difference is calculated based on the fault current data corresponding to the same fault time point and the maintenance completion current data, and a current difference curve is constructed based on the current difference.

[0029] At this point, the fault current data includes all current data for the day of the fault, and the maintenance completion current data includes all current data for the day of the maintenance completion. A current difference curve for the two days is constructed based on the current differences at each sampling time between the fault current data and the maintenance completion current data. It can be understood that if the fault time and the maintenance completion time are on the same day, the current data for the 24 hours prior to the fault time (inclusive) is used as the fault current data, and the current data for the 24 hours following the maintenance completion time (inclusive) is used as the maintenance completion current data. The fault current data and the maintenance completion current data reflect the current deviation between the fault state and the healthy state corresponding to the fault time.

[0030] The difference between the current reference curve and the current difference curve is calculated according to the fault type. The reference fluctuation relationship corresponding to the fault type is constructed based on environmental data, including: Based on the fault type corresponding to the current difference curve, retrieve the current reference curve, calculate the curve difference between the current difference curve and the current reference curve, and record it as the environmental fluctuation value at the corresponding fault time point. Aggregate environmental fluctuation values ​​and corresponding environmental data for the same fault type, and construct a benchmark fluctuation relationship based on the correlation between environmental fluctuation values ​​and environmental data.

[0031] By comparing the common fault current characteristics of the same fault type with the individual fault current characteristics at a single fault time point, the fluctuation of fault current characteristics affected by environmental data at different fault time points can be obtained. At the same time, both common and individual fault current characteristics are retained to ensure the accuracy of the final fault current separation and improve the accuracy of health detection of hydraulic mechanism oil pump motor.

[0032] Specifically, environmental data includes at least temperature, humidity, and load data. The influence of environmental data on environmental fluctuation values ​​is obtained based on the similarity between environmental data and these fluctuations. This establishes a baseline fluctuation relationship, quantifying the environmental impact on current changes caused by exhaust device malfunctions. This facilitates dynamic adjustment of the current baseline curve to adapt to actual environmental conditions.

[0033] In this embodiment, a current reference curve for the same fault type is pre-constructed based on the common fault current characteristics of the same fault type, reflecting the common impact of the fault type. Then, based on the environmental data at a single fault time point and the difference between the actual fault current and the normal current, the specific impact of the environmental data on a single fault is obtained. Without the need to customize the current reference curve according to different operating conditions, the operating condition adaptation of the current reference curve can still be achieved. This adapts to situations where there may be multiple hydraulic mechanisms in the same power grid scenario. It does not require excessive software computing power. In actual calculations, it is only necessary to incorporate the current environmental data to update the pre-stored current reference curve to achieve targeted separation of fault current data.

[0034] Acquire real-time hydraulic mechanism oil pump motor current data, obtain the current state curve corresponding to the current exhaust device based on the reference fluctuation relationship and current reference curve, and output the hydraulic mechanism oil pump motor state detection results based on the state curve and real-time hydraulic mechanism oil pump motor current data, including: Acquire real-time environmental data, and obtain the current environmental fluctuation value based on the real-time environmental data and the baseline fluctuation relationship; The current state curve of the exhaust device is obtained based on the current environmental fluctuation value and the current reference curve. Based on the state curve, the interference current of the exhaust device in the real-time hydraulic mechanism oil pump motor current data is screened out to obtain the state current of the oil pump motor. The status of the hydraulic pump motor is determined based on the state current of the pump motor and the preset current threshold.

[0035] In this embodiment, considering that the number of fault data samples of the oil pump motor is small, but the oil pump motor status judgment based on the oil pump motor current data will be affected by the status of the exhaust device, the fault current feature of the exhaust device is extracted based on the fault data of the exhaust device, and then the interference current feature of the exhaust device is screened out based on the fault current feature, so as to avoid misjudgment of the oil pump motor status caused by the exhaust device abnormality and improve the accuracy of oil pump motor status judgment.

[0036] In this embodiment, the preset fault similarity threshold can be set based on expert experience. In some cases, the preset fault similarity threshold can also be set based on the maximum similarity between historical normal operation data and the current reference curve. Historical normal operation data can be obtained from test data that the maintenance personnel determine belongs to a normal state during the maintenance process.

[0037] Determining the hydraulic mechanism's oil pump motor status based on the oil pump motor's current and a preset current threshold includes: If the fluctuation value of the oil pump motor status current exceeds the preset current threshold, an alarm for abnormal status of the hydraulic mechanism oil pump motor will be output.

[0038] In this embodiment, the preset current threshold is 10% of the maximum historical state current of the oil pump motor. It is understood that the probability of an anomaly occurring during the initial use of the oil pump motor is almost nonexistent. Therefore, using 10% of the maximum historical state current as the preset current threshold, an anomaly is considered to have occurred when the fluctuation value of the oil pump motor state current exceeds the preset current threshold. In applications requiring more precise targeting, state currents identified as abnormal from the historical oil pump motor state currents are excluded, and the maximum value of the excluded historical oil pump motor state current is used as the preset current threshold to avoid interference caused by abnormal currents.

[0039] In other cases, real-time hydraulic mechanism oil pump motor current data is acquired, and the current status curve corresponding to the current exhaust device is obtained based on the reference fluctuation relationship and the current reference curve. Based on the status curve and the real-time hydraulic mechanism oil pump motor current data, the hydraulic mechanism oil pump motor status detection results are output, including: Obtain real-time hydraulic mechanism oil pump motor current data; If the real-time hydraulic mechanism oil pump motor current data is less than the preset idling current threshold, an idling alarm will be output. If the real-time hydraulic mechanism oil pump motor current data is greater than or equal to the preset idling current threshold, the reference fluctuation relationship and current reference curve are used to obtain the current state curve corresponding to the exhaust device. Based on the state curve and the real-time hydraulic mechanism oil pump motor current data, the hydraulic mechanism oil pump motor state detection result is output.

[0040] In this situation, it's crucial to pre-determine if the oil pump motor is idling. Since the current drops sharply when the oil pump motor is idling, it's necessary to stop the motor immediately, even if there are no inherent abnormalities. Only when the oil pump motor is not idling should a status curve be output, and a health check of the oil pump motor should be performed based on whether the fluctuations in the status curve are abnormal.

[0041] Specifically, if the real-time hydraulic mechanism oil pump motor current data is less than the preset no-load current threshold, the current reference curve for a blockage fault in the exhaust device is retrieved for matching. If the matching is successful, the output indicates an exhaust device blockage abnormality; if the matching fails, the output indicates an oil pump motor abnormality. It is understandable that when the real-time hydraulic mechanism oil pump motor current data is less than the preset no-load current threshold, the oil pump motor should be immediately stopped.

[0042] In some cases, a correlation curve between idling pressure and current is pre-constructed. Real-time idling pressure is obtained based on real-time hydraulic mechanism pump motor current data and the correlation curve. A reaction time threshold is then determined based on the overshoot pressure and the real-time idling pressure. Within this threshold, only an idling alarm is triggered, without a shutdown. When the threshold is reached, a shutdown is immediately initiated. This avoids disrupting the power grid by triggering a shutdown during minor maintenance that doesn't require it.

[0043] Understandably, the preset idling current threshold can be set between 10% and 20% of the rated current of the oil pump motor. The overcharge pressure is less than the actual overcharge pressure.

[0044] As a second embodiment of this application, cluster analysis of the full lifecycle current curves based on time length alignment is performed to obtain the current reference curve corresponding to the fault type, including: Perform equal-length interpolation of the full life cycle current curve using a time-length alignment reference to obtain the corrected full life cycle current curve; The sampling time similarity between the corrected full-lifecycle current curves is calculated using a dynamic time warping algorithm. Cluster the current data at sampling times based on the similarity of sampling times; The cluster with the highest density is taken as the fault cluster, and the current curve in the fault cluster is reconstructed based on the continuous sampling time as the fault curve. A current reference curve is constructed using the average current value of the fault curve.

[0045] In this embodiment, the current similarity between curves at each sampling time is calculated using a dynamic time warping algorithm as the sampling time similarity. Clustering is then performed based on this similarity, and the cluster with the highest density is selected as the fault cluster. Since the cluster with the highest density under the same fault type is often the set of sampling time similarities during the fault occurrence period, using the cluster with the highest density as the fault cluster filters out data from fault-free times in the exhaust device, ensuring that the current reference curve reflects the common fault current characteristics corresponding to the fault type.

[0046] Furthermore, the current data corresponding to the similarity of each sampling time in the fault cluster are reconstructed according to the order of sampling time. At this time, the current data separated from the current curve of the same life cycle are reconstructed to ensure that the current reference curve can reflect the time sequence characteristics in the common current characteristics.

[0047] At this point, interference current from the exhaust device is filtered out from the real-time hydraulic mechanism pump motor current data based on the state curve, and the pump motor state current is obtained, including: Calculate the fault similarity of each state curve and the real-time hydraulic mechanism oil pump motor current data. Interference current of the exhaust device in the real-time hydraulic mechanism oil pump motor current data is filtered out using the state curve with a fault similarity greater than the preset fault similarity threshold, and the state current of the oil pump motor is obtained.

[0048] Since it is impossible to determine whether the problem is with the oil pump motor or the exhaust device in actual application scenarios, the similarity of the exhaust device's state curves with the real-time hydraulic mechanism's oil pump motor current data is calculated. If a certain state curve shows a similar trend to the real-time hydraulic mechanism's oil pump motor current data, that is, the fault similarity is greater than the preset fault similarity threshold, it is considered that the exhaust device is very likely to be faulty. The interference current of the exhaust device is then filtered out based on the state curves to obtain the state current of the oil pump motor.

[0049] Understandably, in some cases, if the fault similarity is greater than the preset fault similarity threshold, a fault warning for the exhaust device corresponding to the fault type will be output synchronously.

[0050] In this embodiment, the preset fault similarity threshold can be set based on expert experience. In some cases, the preset fault similarity threshold can also be set based on the maximum similarity between historical normal operation data and the current reference curve. Historical normal operation data can be obtained from test data that the maintenance personnel determine belongs to a normal state during the maintenance process.

[0051] As a third embodiment of this application, the hydraulic mechanism oil pump motor detection system based on the state of the exhaust device includes: The data processing unit is used to divide the hydraulic mechanism oil pump motor current data according to the failure time of the exhaust device and obtain the full life cycle current curve of several exhaust devices. The benchmark analysis unit is used to perform cluster analysis on the full life cycle current curves based on time length alignment to obtain the current benchmark curves corresponding to the fault type. The difference analysis unit is used to calculate the current difference between the fault time point and the maintenance completion time point, and to construct a current difference curve. The fluctuation analysis unit is used to calculate the curve difference between the current reference curve and the current difference curve according to the fault type, and to construct the reference fluctuation relationship corresponding to the fault type based on environmental data. The detection unit is used to acquire real-time hydraulic mechanism oil pump motor current data, obtain the current state curve corresponding to the current exhaust device based on the reference fluctuation relationship and current reference curve, and output the hydraulic mechanism oil pump motor state detection result based on the state curve and real-time hydraulic mechanism oil pump motor current data.

[0052] In this embodiment, the data processing unit is connected to the data storage unit of the hydraulic system to obtain historical fault records of the exhaust device. The data processing unit is connected to the benchmark analysis unit and the difference analysis unit. The fluctuation analysis unit is connected to the benchmark analysis unit and the difference analysis unit. The fluctuation analysis unit and the benchmark analysis unit are connected to the detection unit.

[0053] The specific embodiments described above are preferred embodiments of the hydraulic mechanism oil pump motor detection method and system based on the state of the exhaust device in this application, and are not intended to limit the specific implementation scope of this application. The scope of this application includes but is not limited to the specific embodiments described above. All equivalent changes made in accordance with the shape and structure of this application are within the protection scope of this application.

Claims

1. A method for detecting the oil pump motor of a hydraulic mechanism based on the state of the exhaust device, characterized in that: Includes the following steps: Based on the failure time of the exhaust device, the hydraulic mechanism oil pump motor current data are divided to obtain the full life cycle current curves of several exhaust devices. Cluster analysis of the full life cycle current curves based on time length alignment is performed to obtain the current reference curve corresponding to the fault type; Calculate the current difference between the fault time point and the maintenance completion time point, and construct a current difference curve; The difference between the current reference curve and the current difference curve is calculated according to the fault type, and the reference fluctuation relationship corresponding to the fault type is constructed based on environmental data. The system acquires real-time hydraulic mechanism oil pump motor current data, obtains the current state curve corresponding to the current exhaust device based on the reference fluctuation relationship and current reference curve, and outputs the hydraulic mechanism oil pump motor state detection result based on the state curve and real-time hydraulic mechanism oil pump motor current data.

2. The hydraulic mechanism oil pump motor detection method based on the state of the exhaust device as described in claim 1, characterized in that: The process of dividing the hydraulic mechanism oil pump motor current data according to the failure time of the exhaust device and obtaining the full life cycle current curves of several exhaust devices includes: The hydraulic mechanism oil pump motor current data is divided into two parts based on the time from the completion of the previous maintenance of the exhaust device to the time of the next failure of the exhaust device, and the current curve of the entire life cycle is obtained.

3. The hydraulic mechanism oil pump motor detection method based on the state of the exhaust device as described in claim 1, characterized in that: The clustering analysis of the full lifecycle current curves based on time length alignment to obtain the current reference curve corresponding to the fault type includes: The median time length of the full life cycle current curve for the same fault type is used as the time length alignment benchmark for that fault type. Perform equal-length interpolation of the full life cycle current curve using a time-length alignment reference to obtain the corrected full life cycle current curve; The dynamic time warping algorithm is used to calculate the curve similarity between the corrected full life cycle current curves, and the fault clusters are obtained based on the curve similarity and clustering algorithm. Common features are extracted from fault clusters to construct current reference curves.

4. The hydraulic mechanism oil pump motor detection method based on the state of the exhaust device as described in claim 1, characterized in that: The calculation of the current difference between the fault time point and the maintenance completion time point, and the construction of the current difference curve, include: The current difference is calculated based on the fault current data and maintenance completion current data corresponding to the same fault time point, and a current difference curve is constructed based on the current difference; where the fault current data includes all current data on the day of the fault time point, and the maintenance completion current data includes all current data on the day of the maintenance completion time point.

5. The hydraulic mechanism oil pump motor detection method based on the state of the exhaust device as described in claim 1, characterized in that: The calculation of the curve difference between the current reference curve and the current difference curve according to the fault type, and the construction of the reference fluctuation relationship corresponding to the fault type based on environmental data, include: Based on the fault type corresponding to the current difference curve, retrieve the current reference curve, calculate the curve difference between the current difference curve and the current reference curve, and record it as the environmental fluctuation value at the corresponding fault time point. Aggregate environmental fluctuation values ​​and corresponding environmental data for the same fault type, and construct a benchmark fluctuation relationship based on the correlation between environmental fluctuation values ​​and environmental data.

6. The hydraulic mechanism oil pump motor detection method based on the state of the exhaust device as described in claim 5, characterized in that: The process of acquiring real-time hydraulic mechanism oil pump motor current data, obtaining the current state curve corresponding to the current exhaust device based on the reference fluctuation relationship and current reference curve, and outputting the hydraulic mechanism oil pump motor state detection result based on the state curve and real-time hydraulic mechanism oil pump motor current data includes: Acquire real-time environmental data, and obtain the current environmental fluctuation value based on the real-time environmental data and the baseline fluctuation relationship; The current state curve of the exhaust device is obtained based on the current environmental fluctuation value and the current reference curve. Based on the state curve, the interference current of the exhaust device in the real-time hydraulic mechanism oil pump motor current data is screened out to obtain the state current of the oil pump motor. The status of the hydraulic pump motor is determined based on the state current of the pump motor and the preset current threshold.

7. The hydraulic mechanism oil pump motor detection method based on the state of the exhaust device as described in claim 1, characterized in that: The process of acquiring real-time hydraulic mechanism oil pump motor current data, obtaining the current state curve corresponding to the current exhaust device based on the reference fluctuation relationship and current reference curve, and outputting the hydraulic mechanism oil pump motor state detection result based on the state curve and real-time hydraulic mechanism oil pump motor current data includes: Obtain real-time hydraulic mechanism oil pump motor current data; If the real-time hydraulic mechanism oil pump motor current data is less than the preset idling current threshold, an idling alarm will be output. If the real-time hydraulic mechanism oil pump motor current data is greater than or equal to the preset idling current threshold, the reference fluctuation relationship and current reference curve are used to obtain the current state curve corresponding to the exhaust device. Based on the state curve and the real-time hydraulic mechanism oil pump motor current data, the hydraulic mechanism oil pump motor state detection result is output.

8. The hydraulic mechanism oil pump motor detection method based on the state of the exhaust device as described in claim 3, characterized in that: The clustering analysis of the full lifecycle current curves based on time length alignment to obtain the current reference curve corresponding to the fault type includes: Perform equal-length interpolation of the full life cycle current curve using a time-length alignment reference to obtain the corrected full life cycle current curve; The sampling time similarity between the corrected full-lifecycle current curves is calculated using a dynamic time warping algorithm. Cluster the current data at sampling times based on the similarity of sampling times; The cluster with the highest density is taken as the fault cluster, and the current curve in the fault cluster is reconstructed based on the continuous sampling time as the fault curve. A current reference curve is constructed using the average current value of the fault curve.

9. The hydraulic mechanism oil pump motor detection method based on the state of the exhaust device as described in claim 8, characterized in that: The step of filtering out interference current from the exhaust device in the real-time hydraulic mechanism oil pump motor current data based on the state curve to obtain the oil pump motor state current includes: Calculate the fault similarity of each state curve and the real-time hydraulic mechanism oil pump motor current data. Interference current of the exhaust device in the real-time hydraulic mechanism oil pump motor current data is filtered out using the state curve with a fault similarity greater than the preset fault similarity threshold, and the state current of the oil pump motor is obtained.

10. A hydraulic mechanism oil pump motor detection system based on the state of the exhaust device, used to implement the method as described in any one of claims 1 to 9, characterized in that: include: The data processing unit is used to divide the hydraulic mechanism oil pump motor current data according to the failure time of the exhaust device and obtain the full life cycle current curve of several exhaust devices. The benchmark analysis unit is used to perform cluster analysis on the full life cycle current curves based on time length alignment to obtain the current benchmark curves corresponding to the fault type. The difference analysis unit is used to calculate the current difference between the fault time point and the maintenance completion time point, and to construct a current difference curve. The fluctuation analysis unit is used to calculate the curve difference between the current reference curve and the current difference curve according to the fault type, and to construct the reference fluctuation relationship corresponding to the fault type based on environmental data. The detection unit is used to acquire real-time hydraulic mechanism oil pump motor current data, obtain the current state curve corresponding to the current exhaust device based on the reference fluctuation relationship and current reference curve, and output the hydraulic mechanism oil pump motor state detection result based on the state curve and real-time hydraulic mechanism oil pump motor current data.

Images