New energy station maintenance system and method based on intelligent sensor

Abstract

The invention relates to the technical field of new energy station maintenance, in particular to a new energy station maintenance system and method based on an intelligent sensor. A data processing module; the fault identification module is used for outputting sensor abnormal data, acquiring sensor state information, adjusting each sensor identification result according to the sensor state information, constructing a fault position identification model, and outputting fault equipment positioning according to the sensor abnormal data; and the fault maintenance module is used for constructing an equipment maintenance model, acquiring equipment maintenance information, updating a preset cumulative distance database according to the error output times of each piece of equipment maintenance information, and outputting equipment maintenance alarm information according to the updating times of the preset cumulative distance database. And the control module is also used for adjusting the equipment maintenance alarm information output sequence of each equipment according to the equipment maintenance alarm information output times. According to the invention, the troubleshooting time is reduced, and the maintenance efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy station maintenance, and in particular to a new energy station maintenance and repair system and method based on intelligent sensors. Background Art

[0002] New energy stations are usually built in remote areas, with equipment scattered and far away, which brings difficulties to operation and maintenance. Since the equipment is scattered and far away, the fault response speed is slow, which will extend the equipment downtime and affect the power generation efficiency and economic benefits of the new energy stations. New energy stations require professional operation and maintenance teams to carry out daily inspections, troubleshooting and repairs. These tasks require highly qualified technical personnel and need to be stationed in remote areas for a long time, resulting in high labor costs. Long-term calculations in the background of cloud computers and master control computers may cause the entire system to crash, affecting the normal operation and monitoring of the power station.

[0003] Chinese Patent Publication No.: CN108769156A discloses a maintenance and management system for small and medium-sized solar power stations, which belongs to the technical field of solar power stations. A maintenance and management system for small and medium-sized solar power stations includes temperature sensors, light sensors, solar power station metering terminals, signal collection devices, cloud computers, general control computer backgrounds, maintenance or management personnel mobile phone text messages and maintenance or management personnel mobile phone APPs installed in small and medium-sized solar power stations. Data of small and medium-sized solar power stations can be collected, and cloud monitoring and analysis of various data of the solar power station are normal. When the data is abnormal, the fault type will be automatically generated and the maintenance personnel or relevant management personnel will be notified in time. It can be seen that the maintenance and management system for small and medium-sized solar power stations has the following problems: when using cloud monitoring and analysis, the troubleshooting time is long, the maintenance efficiency is low, and there is a lack of prediction function for potential faults. Summary of the invention

[0004] To this end, the present invention provides a new energy station maintenance and repair system and method based on intelligent sensors to overcome the problems of long troubleshooting time, low maintenance efficiency and lack of potential fault prediction function during cloud monitoring and analysis in the prior art.

[0005] To achieve the above objectives, on the one hand, the present invention provides a new energy station maintenance and repair system based on intelligent sensors, comprising: Data collection module, which collects normal equipment operation data and real-time equipment operation data; Data processing module, pre-processing normal equipment operation data and real-time equipment operation data; A fault identification module is used to obtain the accumulated distance, output sensor abnormality data according to the accumulated distance, obtain sensor status information, adjust the identification results of each sensor according to the sensor status information, obtain preset accumulated distance automatic adjustment information, and build a fault location identification model to output the faulty device location according to the sensor abnormality data; The fault maintenance module is used to build an equipment maintenance model, obtain equipment maintenance information, update the preset cumulative distance database according to the number of erroneous outputs of each equipment maintenance information, output equipment maintenance alarm information according to the number of updates of the preset cumulative distance database, and adjust the output order of equipment maintenance alarm information of each equipment according to the number of outputs of equipment maintenance alarm information.

[0006] Furthermore, the fault identification module includes: A fault time identification unit is used to perform DTW comparison between real-time time series device data and preset time series device data, obtain cumulative distance, judge the identification results of each sensor according to the cumulative distance, and output sensor abnormal data according to the judgment result, and is also used to build a sensor fault identification model, obtain sensor status information, adjust the identification results of each sensor according to the sensor status information, and is also used to build a preset cumulative distance automatic adjustment model to obtain preset cumulative distance automatic adjustment information; The fault location identification unit is used to build a fault location identification model and output the faulty device location according to the sensor abnormal data.

[0007] Furthermore, when judging the recognition results of each sensor, the fault time recognition unit performs time series processing on the normal equipment operation data, divides the normal equipment operation data into time intervals according to preset unit time intervals, and constructs preset time series equipment data from the normal equipment operation data after the time interval division in time sequence, performs time series processing on the real-time equipment operation data, obtains the real-time time series equipment data, performs DTW comparison on the real-time time series equipment data and the preset time series equipment data, obtains the cumulative distance L, compares the cumulative distance L with the preset cumulative distance L0, judges the recognition results of each sensor according to the comparison results, and outputs the sensor abnormality data according to the judgment results, wherein: When L≤L0, the fault time identification unit determines that the sensor identification result is normal and does not output sensor abnormality data; When L>L0, the fault time identification unit determines that the sensor identification result is abnormal, and outputs sensor abnormality data to the administrator.

[0008] Furthermore, the fault time identification unit constructs a sensor fault identification model, inputs the real-time time series device data into the sensor fault identification model, obtains the output result of the sensor fault identification model, and uses it as the sensor state information, wherein the sensor state information includes the normal operation of the sensor and the abnormal operation of the sensor, and adjusts the identification results of each sensor according to the sensor state information, wherein: When the sensor status information indicates that the sensor is operating normally, the recognition result of the sensor is not adjusted; When the sensor status information indicates that the sensor is working abnormally, the sensor identification result is adjusted to be normal, and sensor maintenance alarm information is output to the administrator.

[0009] Further, when the fault time identification unit makes real-time adjustment to the preset cumulative distance L0, a preset cumulative distance automatic adjustment model is constructed, and the preset cumulative distance database is divided into 70% of the preset cumulative distance training set, 15% of the preset cumulative distance verification set and 15% of the preset cumulative distance test set, and the preset cumulative distance training set is input into the random forest model for training to obtain the trained random forest model, and the preset cumulative distance verification set is input into the trained random forest model to iteratively optimize the parameters of the trained random forest model, and the preset cumulative distance test set is input into the iteratively optimized random forest model, and the iteratively optimized recurrent network neural model is tested to obtain the automatic adjustment accuracy U, and the automatic adjustment accuracy U is compared with the preset automatic adjustment accuracy U0, and the training compliance of the random forest model is judged according to the comparison result, and the random forest model is output according to the judgment result, wherein: When U≥U0, the random forest model training is determined to have met the standard, and the random forest model is output as the preset cumulative distance automatic adjustment model; When U<U0, it is determined that the random forest model training does not meet the standard, the random forest model parameters are iteratively optimized, and the training is repeated until the random forest model training meets the standard; The real-time device data is input into the preset cumulative distance automatic adjustment model to obtain the preset cumulative distance automatic adjustment information.

[0010] Furthermore, when locating the faulty equipment, the fault location identification unit constructs a fault location identification model, obtains sensor abnormality data through the fault time identification unit, inputs the sensor abnormality data into the fault location identification model, obtains the output result of the fault location identification model, and uses it as the faulty equipment location.

[0011] Furthermore, the fault maintenance module constructs an equipment maintenance model according to the sensor abnormality data and the historical maintenance database, inputs the real-time sensor abnormality data into the equipment maintenance model, obtains the output result of the equipment maintenance model, uses it as the equipment maintenance information, compares the number of error outputs S of each equipment maintenance information with the preset number of error outputs S0, judges the real-time performance of the preset cumulative distance automatic adjustment model according to the comparison result, and updates the preset cumulative distance database according to the judgment result, wherein: When S<S0, the fault maintenance module determines that the real-time performance of the preset cumulative distance automatic adjustment model is normal, and does not update the preset cumulative distance database; When S≥S0, the fault maintenance module determines that the real-time performance of the preset cumulative distance automatic adjustment model is abnormal, collects actual fault information, updates the preset cumulative distance database according to the actual fault information, and trains the preset cumulative distance automatic adjustment model according to the updated preset cumulative distance database until the preset cumulative distance automatic adjustment model training meets the standard, and outputs the updated preset cumulative distance automatic adjustment model; The fault maintenance module compares the actual operation fault with the historical maintenance database, and judges the type of fault according to the comparison result, and adjusts the real-time judgment result of the preset cumulative distance automatic adjustment model according to the judgment result, wherein: When the actual operation fault is consistent with the historical maintenance database, the fault maintenance module determines that the fault type is a new fault, and adjusts the judgment result of the real-time performance of the preset cumulative distance automatic adjustment model to the real-time performance abnormality of the preset cumulative distance automatic adjustment model; When the actual operation fault is inconsistent with the historical maintenance database, the fault maintenance module determines that the fault occurrence type is a historical fault, and the fault maintenance module does not adjust the real-time judgment result of the preset cumulative distance automatic adjustment model.

[0012] Furthermore, the fault maintenance module compares the preset cumulative distance database update times C with the preset update times C0, judges the practicality of each device according to the comparison result, and outputs device maintenance alarm information according to the judgment result, wherein: When C<C0, the fault maintenance module determines that the practicability of the device is normal and does not output the device maintenance alarm information; When C≥C0, the fault maintenance module determines that the practicability of the device is abnormal, outputs device maintenance alarm information, and notifies the administrator to perform maintenance on the device.

[0013] Furthermore, the fault maintenance module compares the output times R of the equipment maintenance alarm information with the preset output times R0 of the equipment maintenance alarm information, judges the maintenance importance of each device according to the comparison result, and adjusts the output order of the equipment maintenance alarm information of each device according to the judgment result, wherein: When R<R0, the fault maintenance module determines that the maintenance importance of the equipment is high, and does not adjust the output sequence of the equipment maintenance alarm information of the equipment; When R≥R0, the fault maintenance module determines that the maintenance importance of the equipment is high, and adjusts the output sequence of the equipment maintenance alarm information of the equipment to priority output.

[0014] On the other hand, the present invention provides a maintenance method for a new energy station based on an intelligent sensor, comprising: Step S1, collecting normal equipment operation data and real-time equipment operation data; Step S2, preprocessing normal equipment operation data and real-time equipment operation data; Step S3, performing DTW comparison on the real-time time series device data and the preset time series device data, obtaining the cumulative distance, and outputting the sensor abnormality data according to the cumulative distance; Step S4, constructing a sensor fault identification model, obtaining sensor status information, and adjusting the identification results of each sensor according to the sensor status information; Step S5, constructing a preset cumulative distance automatic adjustment model, and obtaining preset cumulative distance automatic adjustment information; Step S6, constructing a fault location recognition model and outputting the faulty device location according to the sensor abnormal data; Step S7, constructing a device maintenance model, obtaining device maintenance information, and updating a preset cumulative distance database according to the number of erroneous outputs of each device maintenance information; Step S8, outputting equipment maintenance alarm information according to the preset cumulative distance database update times; Step S9: adjusting the output sequence of the equipment maintenance alarm information of each device according to the number of times the equipment maintenance alarm information is output.

[0015] Compared with the prior art, the beneficial effect of the present invention is that the system is applied to the maintenance control terminal of the new energy station, and through real-time monitoring and comparison of data, potential faults can be discovered in time, early warning can be provided for equipment maintenance, troubleshooting time can be reduced, and maintenance efficiency can be improved. The system ensures the real-time nature of data through the data acquisition module, so that the administrator can understand the current status and performance of the equipment in time, and provide an accurate basis for decision-making. The system pre-processes the real-time equipment data through the data processing module for subsequent data comparison. The system identifies potential faults through the fault identification module, provides early warning for equipment maintenance, avoids production interruptions and safety accidents caused by sudden faults, and accurately outputs the location of the faulty equipment according to the abnormal sensor data. It helps administrators to quickly locate fault points, reduce troubleshooting time, improve maintenance efficiency, reduce the cost of repairs and replacement parts, reduce equipment maintenance costs, promptly discover potential problems in the production process, and provide data support for optimizing production processes. The system improves the speed of acquiring equipment maintenance information through the fault maintenance module, ensures the timeliness and effectiveness of maintenance work, prevents equipment failures, reduces downtime, and improves production efficiency. By dynamically updating the database, it meets the maintenance needs of different equipment and different working conditions, improves the flexibility and effectiveness of maintenance strategies, and prioritizes the maintenance needs of each equipment, which helps technicians to reasonably arrange maintenance tasks, give priority to key equipment and emergency failures, and ensure the stable operation of the production line.

[0016] In particular, the data acquisition module ensures the real-time nature of the data by capturing the device data with the latest timestamp, so that the administrator can understand the current status and performance of the device in a timely manner and provide an accurate basis for decision-making.

[0017] In particular, the data processing module performs preprocessing on the real-time device data to ensure the data quality of the real-time device data, and performs data standardization on the cleaned data to facilitate subsequent data comparison.

[0018] In particular, the fault identification module can quickly obtain the cumulative distance by comparing the real-time time series equipment data with the preset time series equipment data through DTW, and then judge the identification results of each sensor, thereby improving the efficiency of fault diagnosis. The cumulative distance is used as the basis for judgment to accurately identify abnormal sensor data. By building a sensor fault identification model, the sensor status information is obtained, and the identification results are fine-tuned to improve the accuracy of diagnosis. Through real-time monitoring and comparison of data, potential faults can be discovered in time, and early warnings can be provided for equipment maintenance to avoid production interruptions and safety accidents caused by sudden faults. By building a fault location identification model, the location of the faulty equipment can be accurately output based on the abnormal sensor data, which helps administrators quickly locate the fault point, reduce troubleshooting time, improve maintenance efficiency, reduce the cost of repair and replacement of parts, reduce the maintenance cost of equipment, and timely discover potential problems in the production process, providing data support for optimizing the production process.

[0019] In particular, the fault maintenance module improves the speed of acquiring equipment maintenance information by inputting real-time sensor abnormality data into the equipment maintenance model, ensures the timeliness and effectiveness of maintenance work, prevents equipment failures, reduces downtime, and improves production efficiency. By dynamically updating the database, it meets the maintenance needs of different equipment and different working conditions, improves the flexibility and effectiveness of maintenance strategies, helps to reduce the company's maintenance costs, improves economic benefits, helps to ensure the stable operation of equipment and extend its service life, and prioritizes the maintenance needs of each equipment according to the number of output times of equipment maintenance alarm information, which helps technicians to reasonably arrange maintenance tasks, give priority to key equipment and emergency failures, and ensure the stable operation of the production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a schematic diagram of the structure of the new energy station maintenance and repair system based on intelligent sensors in this embodiment; Figure 2 This is a schematic diagram of the structure of the fault identification module of this embodiment; Figure 3 The figure is a flowchart of the maintenance method of a new energy station based on smart sensors according to the present embodiment. DETAILED DESCRIPTION

[0021] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0022] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0023] It should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0024] See also Figure 1 As shown, it is a structural diagram of the new energy station maintenance and repair system based on intelligent sensors in this embodiment, including: Data collection module, which collects normal equipment operation data and real-time equipment operation data; Data processing module, pre-processing normal equipment operation data and real-time equipment operation data; A fault identification module is used to obtain the accumulated distance, output sensor abnormality data according to the accumulated distance, obtain sensor status information, adjust the identification results of each sensor according to the sensor status information, obtain preset accumulated distance automatic adjustment information, and build a fault location identification model to output the faulty device location according to the sensor abnormality data; The fault maintenance module is used to build an equipment maintenance model, obtain equipment maintenance information, update the preset cumulative distance database according to the number of erroneous outputs of each equipment maintenance information, output equipment maintenance alarm information according to the number of updates of the preset cumulative distance database, and adjust the output order of equipment maintenance alarm information of each equipment according to the number of outputs of equipment maintenance alarm information.

[0025] Specifically, the system is applied to the maintenance control terminal of new energy stations. Through real-time monitoring and comparison of data, potential faults can be discovered in time, early warning can be provided for equipment maintenance, troubleshooting time can be reduced, and maintenance efficiency can be improved. The system ensures the real-time nature of data through the data acquisition module, so that administrators can understand the current status and performance of the equipment in time, providing an accurate basis for decision-making. The system pre-processes real-time equipment data through the data processing module for subsequent data comparison. The system identifies potential faults through the fault identification module, provides early warning for equipment maintenance, avoids production interruptions and safety accidents caused by sudden faults, and accurately outputs the location of faulty equipment based on abnormal sensor data, which helps administrators to quickly Quickly locate fault points, reduce troubleshooting time, improve maintenance efficiency, reduce maintenance and parts replacement costs, reduce equipment maintenance costs, timely discover potential problems in the production process, and provide data support for optimizing production processes. The system improves the speed of acquiring equipment maintenance information through the fault maintenance module, ensures the timeliness and effectiveness of maintenance work, prevents equipment failures, reduces downtime, and improves production efficiency. By dynamically updating the database, it meets the maintenance needs of different equipment and different working conditions, improves the flexibility and effectiveness of maintenance strategies, and prioritizes the maintenance needs of each equipment, which helps technicians to reasonably arrange maintenance tasks, give priority to key equipment and emergency failures, and ensure the stable operation of the production line.

[0026] Specifically, when the data acquisition module collects normal equipment operation data and real-time equipment operation data, the normal equipment operation data refers to the specific values ​​of the operation of each device monitored and recorded by sensors when the internal equipment of the new energy station is in a normal working state, such as current, voltage and vibration. The real-time equipment data refers to the specific values ​​of the real-time operation of the internal equipment of the new energy station. This embodiment does not limit the method for obtaining the equipment operation data. Those skilled in the art can freely set it according to actual conditions, and only need to meet the real-time acquisition requirements of the equipment data. For example, it can be set to obtain the temperature data of the photovoltaic panel through a temperature sensor, obtain the solar radiation intensity data through an irradiance sensor, obtain the current data of the photovoltaic panel through a current sensor, obtain the voltage data of the photovoltaic panel through a voltage sensor, and obtain the inclination angle data of the photovoltaic panel through an inclination sensor. The equipment operation data includes the normal equipment operation data and the real-time equipment operation data. This embodiment does not limit the storage method of the equipment operation data. Those skilled in the art can freely set it according to actual conditions, and only need to meet the storage requirements of the equipment operation data. For example, it can be set to store the equipment operation data in a sensor database, and mark each sensor with an ID, and the ID mark includes the sensor type and sensor location.

[0027] Specifically, the data acquisition module ensures the real-time nature of the data by capturing the device data with the latest timestamp, so that the administrator can understand the current status and performance of the device in a timely manner, providing an accurate basis for decision-making.

[0028] Specifically, when preprocessing the normal device operation data and the real-time device operation data, the data processing module processes abnormal values ​​and noise data in the normal device operation data and the real-time device operation data, and performs data standardization processing.

[0029] Specifically, this embodiment does not limit the processing method of outliers in real-time device data. Those skilled in the art can freely set it according to actual conditions, and only need to meet the processing requirements of outliers in real-time device data. For example, a box plot method can be set to identify outliers and delete outliers. The box plot method refers to describing the distribution characteristics of data by drawing the maximum value, minimum value, median, first quartile (Q1) and third quartile (Q3) of the data, and defining data points lower than Q1-1.5×(Q3-Q1) or higher than Q3+1.5×(Q3-Q1) as outliers. This embodiment smoothes the noise data in the real-time device data through Kalman filtering. The Kalman filtering refers to a recursive algorithm for estimating the state of a dynamic system from noisy observation data. The data standardization processing refers to the process of uniformly normalizing data of different formats, different sources and different structures.

[0030] Specifically, the data processing module pre-processes the real-time device data to ensure the data quality of the real-time device data, and performs data standardization on the cleaned data to facilitate subsequent data comparison.

[0031] See also Figure 2 As shown, it is a schematic diagram of the structure of the fault identification module of this embodiment, including: A fault time identification unit is used to perform DTW comparison between real-time time series device data and preset time series device data, obtain cumulative distance, judge the identification results of each sensor according to the cumulative distance, and output sensor abnormal data according to the judgment result, and is also used to build a sensor fault identification model, obtain sensor status information, adjust the identification results of each sensor according to the sensor status information, and is also used to build a preset cumulative distance automatic adjustment model to obtain preset cumulative distance automatic adjustment information; The fault location identification unit is used to build a fault location identification model and output the faulty device location according to the sensor abnormal data.

[0032] Specifically, when judging the recognition results of each sensor, the fault time recognition unit performs time series processing on the normal equipment operation data, divides the normal equipment operation data into time intervals according to preset unit time intervals, and constructs preset time series equipment data from the normal equipment operation data after the time interval division in time sequence, performs time series processing on the real-time equipment operation data, obtains the real-time time series equipment data, performs DTW comparison on the real-time time series equipment data and the preset time series equipment data, obtains the cumulative distance L, compares the cumulative distance L with the preset cumulative distance L0, judges the recognition results of each sensor according to the comparison results, and outputs the sensor abnormality data according to the judgment results, wherein: When L≤L0, the fault time identification unit determines that the sensor identification result is normal and does not output sensor abnormality data; When L>L0, the fault time identification unit determines that the sensor identification result is abnormal, and outputs sensor abnormality data to the administrator.

[0033] Specifically, the time interval division refers to the time division of the equipment operation data according to the preset unit time interval, and the division is performed once every preset unit time interval, and the divided data set is used as the equipment operation data after the time interval division. This embodiment does not specifically limit the value of the preset unit time interval. Technical personnel in this field can freely set it according to actual conditions, and only need to meet the needs of time series processing of the equipment operation data. For example, the preset unit time interval can be set to 5 minutes. The time series equipment data refers to the data of the equipment status of the new energy station changing over time. The cumulative distance L refers to the similarity between two time series calculated by the DTW algorithm. The preset cumulative distance L0 refers to the preset value used to judge the identification results of each sensor before performing the DTW comparison, such as setting L0=80. The sensor abnormal data refers to the data in which there are significant differences between the sensor data and the normal equipment operation data, such as the time of abnormal occurrence, the type of abnormal value and the sensor type.

[0034] Specifically, the fault time identification unit constructs a sensor fault identification model, inputs real-time time series device data into the sensor fault identification model, obtains the output result of the sensor fault identification model, and uses it as sensor state information, wherein the sensor state information includes normal operation of the sensor and abnormal operation of the sensor, and adjusts the identification results of each sensor according to the sensor state information, wherein: When the sensor status information indicates that the sensor is operating normally, the recognition result of the sensor is not adjusted; When the sensor status information indicates that the sensor is working abnormally, the sensor identification result is adjusted to be normal, and sensor maintenance alarm information is output to the administrator.

[0035] It can be understood that the present embodiment does not limit the construction process of the sensor fault identification model. Those skilled in the art can freely set it according to actual conditions, and only need to meet the identification requirements of the sensor fault. For example, a data combination with time series device data as an index and sensor state information as the index corresponding content can be set as sensor fault identification model construction data, 75% of the fault location identification model construction data is divided into a model training set, and 25% of the sensor fault identification model construction data is divided into a model test set. The convolutional neural network model is trained according to the model training set, and the trained convolutional neural network model is tested according to the model test set. When the test accuracy reaches 97%, the trained convolutional neural network model is output as a fault location identification model. When the test accuracy does not reach 97%, the trained convolutional neural network model continues to be trained until the test accuracy reaches 97% and then it is output as a sensor fault identification model. The present embodiment does not limit the output method of the sensor maintenance alarm information. Those skilled in the art can freely set it according to actual needs, and only need to meet the output requirements of the sensor maintenance alarm information. For example, the sensor maintenance alarm information can be set to be sent to the administrator's mobile phone in the form of a text message.

[0036] Specifically, when the fault time identification unit makes real-time adjustment to the preset cumulative distance L0, a preset cumulative distance automatic adjustment model is constructed, and the preset cumulative distance database is divided into 70% of the preset cumulative distance training set, 15% of the preset cumulative distance verification set and 15% of the preset cumulative distance test set, and the preset cumulative distance training set is input into the random forest model for training to obtain the trained random forest model, and the preset cumulative distance verification set is input into the trained random forest model to iteratively optimize the parameters of the trained random forest model, and the preset cumulative distance test set is input into the iteratively optimized random forest model, and the iteratively optimized recurrent network neural model is tested to obtain the automatic adjustment accuracy U, and the automatic adjustment accuracy U is compared with the preset automatic adjustment accuracy U0, and the training compliance of the random forest model is judged according to the comparison result, and the random forest model is output according to the judgment result, wherein: When U≥U0, the random forest model training is determined to have met the standard, and the random forest model is output as the preset cumulative distance automatic adjustment model; When U<U0, it is determined that the random forest model training does not meet the standard, the random forest model parameters are iteratively optimized, and the training is repeated until the random forest model training meets the standard; The real-time device data is input into the preset cumulative distance automatic adjustment model to obtain the preset cumulative distance automatic adjustment information.

[0037] Specifically, the preset cumulative distance database refers to a historical database used to store, manage and analyze historical equipment operation data and corresponding preset cumulative distance data, the preset cumulative distance training set refers to a data set used to train the random forest model, the preset cumulative distance verification set refers to a data set used to optimize and evaluate the performance of the random forest model, the preset cumulative distance test set refers to a data set used to test the preset cumulative distance automatic adjustment model, the random forest model refers to a neural network model used to process sequence data, the automatic adjustment accuracy U refers to the proportion of the number of samples Y correctly output by the random forest model to the total number of samples Y0 in the preset cumulative distance test set, that is, U=Y / Y0, the preset automatic adjustment accuracy U0 refers to a preset value for judging whether the training of the random forest model meets the standard, this embodiment does not limit the value of the preset automatic adjustment accuracy U0, relevant technical personnel in this field can freely set it according to actual needs, as long as the need to judge whether the training of the random forest model meets the standard is met, such as the preset automatic adjustment accuracy U0=0.96 can be set, the preset cumulative distance automatic adjustment information refers to predicting the future operation status of the equipment according to the real-time equipment operation data, and automatically adjusting the preset cumulative distance L0 according to the prediction result.

[0038] Specifically, when locating a faulty device, the fault location identification unit constructs a fault location identification model, obtains sensor abnormality data through the fault time identification unit, inputs the sensor abnormality data into the fault location identification model, obtains the output result of the fault location identification model, and uses it as the location of the faulty device.

[0039] It can be understood that the present embodiment does not limit the construction process of the fault location identification model, and those skilled in the art can freely set it according to actual conditions, and only need to meet the identification requirements for locating the faulty equipment. For example, a data combination with sensor abnormal data as the index and sensor ID as the index corresponding content can be set as the fault location identification model construction data, and 70% of the fault location identification model construction data is divided into a model training set, and 30% of the fault location identification model construction data is divided into a model test set. The convolutional neural network model is trained according to the model training set, and the trained convolutional neural network model is tested according to the model test set. When the test accuracy reaches 98%, the trained convolutional neural network model is output as the fault location identification model. When the test accuracy does not reach 98%, the trained convolutional neural network model continues to be trained until the test accuracy reaches 98% and then it is output as the fault location identification model.

[0040] Specifically, the fault identification module can quickly obtain the cumulative distance by comparing the real-time time series equipment data with the preset time series equipment data through DTW, and then judge the identification results of each sensor to improve the efficiency of fault diagnosis. The cumulative distance is used as the basis for judgment to accurately identify abnormal sensor data. By building a sensor fault identification model, the sensor status information is obtained, and the identification results are fine-tuned to improve the accuracy of diagnosis. Through real-time monitoring and comparison of data, potential faults are discovered in time, and early warnings are provided for equipment maintenance to avoid production interruptions and safety accidents caused by sudden faults. By building a fault location identification model, the location of the faulty equipment is accurately output according to the abnormal sensor data, which helps administrators quickly locate the fault point, reduce troubleshooting time, improve maintenance efficiency, reduce the cost of repair and replacement of parts, reduce the maintenance cost of equipment, and discover potential problems in the production process in time, providing data support for optimizing the production process.

[0041] Specifically, the fault maintenance module builds an equipment maintenance model based on sensor abnormality data and a historical maintenance database, inputs real-time sensor abnormality data into the equipment maintenance model, obtains the output result of the equipment maintenance model, uses it as equipment maintenance information, compares the number of error outputs S of each equipment maintenance information with the preset number of error outputs S0, judges the real-time performance of the preset cumulative distance automatic adjustment model based on the comparison result, and updates the preset cumulative distance database based on the judgment result, wherein: When S<S0, the fault maintenance module determines that the real-time performance of the preset cumulative distance automatic adjustment model is normal, and does not update the preset cumulative distance database; When S≥S0, the fault maintenance module determines that the real-time performance of the preset cumulative distance automatic adjustment model is abnormal, collects actual fault information, updates the preset cumulative distance database according to the actual fault information, and trains the preset cumulative distance automatic adjustment model according to the updated preset cumulative distance database until the preset cumulative distance automatic adjustment model training meets the standard, and outputs the updated preset cumulative distance automatic adjustment model; The fault maintenance module compares the actual operation fault with the historical maintenance database, and judges the type of fault according to the comparison result, and adjusts the judgment result of the real-time performance of the preset cumulative distance automatic adjustment model according to the judgment result, wherein: When the actual operation fault is consistent with the historical maintenance database, the fault maintenance module determines that the fault type is a new fault, and adjusts the judgment result of the real-time performance of the preset cumulative distance automatic adjustment model to the real-time performance abnormality of the preset cumulative distance automatic adjustment model; When the actual operation fault is inconsistent with the historical maintenance database, the fault maintenance module determines that the fault occurrence type is a historical fault, and the fault maintenance module does not adjust the real-time judgment result of the preset cumulative distance automatic adjustment model.

[0042] It can be understood that the present embodiment does not limit the construction process of the equipment maintenance model, and those skilled in the art can freely set it according to the actual situation, and only need to meet the output requirements of the equipment maintenance information. For example, the data combination with the sensor abnormality data and the historical maintenance database as the index and the equipment maintenance information as the index corresponding content can be set as the equipment maintenance model construction data, and 85% of the fault location identification model construction data is divided into a model training set, and 15% of the equipment maintenance model construction data is divided into a model test set. The convolutional neural network model is trained according to the model training set, and the trained convolutional neural network model is tested according to the model test set. When the test accuracy reaches 96%, the trained convolutional neural network model is output as the fault location identification model. When the test accuracy does not reach 96%, the trained convolutional neural network model continues to be trained until the test accuracy reaches 96% and then it is output as the equipment maintenance model. The real-time sensor abnormality data refers to the sensor abnormality data obtained by real-time measurement, and the equipment maintenance information refers to the data and records of the equipment during the maintenance process, including the fault time, fault phenomenon, etc. The embodiment of the present invention is as follows: the number of times S of the erroneous output of the maintenance information of each device refers to the number of times the system incorrectly judges the status of the device when monitoring the real-time operating status of the device within the evaluation period. The present embodiment does not limit the specific value of the evaluation period. The relevant technical personnel in this field can freely set it according to actual needs, and only need to meet the evaluation requirements of the real-time operating status of each device. For example, the evaluation period can be set to one month. The preset number of erroneous outputs S0 refers to the preset value for measuring the real-time performance of the preset cumulative distance automatic adjustment model. For example, S0=4 times is set. The present embodiment does not limit the output mode of the device maintenance alarm information. The relevant technical personnel in this field can freely set it according to actual needs, and only need to meet the output requirements of the device maintenance alarm information. For example, the device maintenance alarm information can be set to be sent to the administrator's mobile phone in the form of a text message. The actual operation failure refers to the situation in which the device cannot operate normally according to the expected performance, specifications and standards in actual operation. The new failure refers to the failure phenomenon that occurs for the first time in the actual operation of the device. The historical failure refers to the failure phenomenon that has occurred in the actual operation of the device.

[0043] Specifically, the fault maintenance module compares the preset cumulative distance database update times C with the preset update times C0, judges the practicality of each device according to the comparison result, and outputs the device maintenance alarm information according to the judgment result, wherein: When C<C0, the fault maintenance module determines that the practicability of the device is normal and does not output the device maintenance alarm information; When C≥C0, the fault maintenance module determines that the practicability of the device is abnormal, outputs device maintenance alarm information, and notifies the administrator to perform maintenance on the device.

[0044] Specifically, the preset error output number S0 refers to a preset value for measuring whether the real-time performance of the preset cumulative distance automatic adjustment model is normal, such as setting S0=3 times. The preset cumulative distance database update number C refers to the number of times the preset cumulative distance database is updated within the evaluation period. The preset update number C0 refers to a preset value for measuring whether the practicality of each device is normal, such as setting C0=3 times.

[0045] Specifically, the fault maintenance module compares the equipment maintenance alarm information output times R with the preset equipment maintenance alarm information output times R0, judges the maintenance importance of each device according to the comparison result, and adjusts the output order of the equipment maintenance alarm information of each device according to the judgment result, wherein: When R<R0, the fault maintenance module determines that the maintenance importance of the equipment is high, and does not adjust the output sequence of the equipment maintenance alarm information of the equipment; When R≥R0, the fault maintenance module determines that the maintenance importance of the equipment is high, and adjusts the output sequence of the equipment maintenance alarm information of the equipment to priority output.

[0046] Specifically, the fault maintenance module inputs real-time sensor abnormality data into the equipment maintenance model to improve the speed of acquiring equipment maintenance information, ensure the timeliness and effectiveness of maintenance work, prevent equipment failures, reduce downtime, and improve production efficiency. By dynamically updating the database, it meets the maintenance needs of different equipment and different working conditions, improves the flexibility and effectiveness of maintenance strategies, helps to reduce the company's maintenance costs, improves economic benefits, helps to ensure the stable operation of equipment and extend its service life. According to the number of output times of equipment maintenance alarm information, the maintenance needs of each equipment are prioritized, which helps technicians to reasonably arrange maintenance tasks, give priority to key equipment and emergency failures, and ensure the stable operation of the production line.

[0047] See also Figure 3 As shown, it is a flow chart of the maintenance method of a new energy station based on a smart sensor in this embodiment, including: Step S1, collecting normal equipment operation data and real-time equipment operation data; Step S2, preprocessing normal equipment operation data and real-time equipment operation data; Step S3, performing DTW comparison on the real-time time series device data and the preset time series device data, obtaining the cumulative distance, and outputting the sensor abnormality data according to the cumulative distance; Step S4, constructing a sensor fault identification model, obtaining sensor status information, and adjusting the identification results of each sensor according to the sensor status information; Step S5, constructing a preset cumulative distance automatic adjustment model, and obtaining preset cumulative distance automatic adjustment information; Step S6, constructing a fault location recognition model and outputting the faulty device location according to the sensor abnormal data; Step S7, constructing a device maintenance model, obtaining device maintenance information, and updating a preset cumulative distance database according to the number of erroneous outputs of each device maintenance information; Step S8, outputting equipment maintenance alarm information according to the preset cumulative distance database update times; Step S9: adjusting the output sequence of the equipment maintenance alarm information of each device according to the number of times the equipment maintenance alarm information is output.

[0048] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A new energy station maintenance and repair system based on intelligent sensors, characterized in that: include: Data collection module, used to collect normal equipment operation data and real-time equipment operation data; A data processing module, used for preprocessing normal equipment operation data and real-time equipment operation data; A fault identification module is used to obtain the accumulated distance, output sensor abnormality data according to the accumulated distance, obtain sensor status information, adjust the identification results of each sensor according to the sensor status information, obtain preset accumulated distance automatic adjustment information, and build a fault location identification model to output the faulty device location according to the sensor abnormality data; The fault maintenance module is used to build an equipment maintenance model, obtain equipment maintenance information, update the preset cumulative distance database according to the number of erroneous outputs of each equipment maintenance information, output equipment maintenance alarm information according to the number of updates of the preset cumulative distance database, and adjust the output order of equipment maintenance alarm information of each equipment according to the number of outputs of equipment maintenance alarm information.

2. The intelligent sensor-based new energy station maintenance and repair system according to claim 1 is characterized in that: The fault identification module comprises: A fault time identification unit is used to perform DTW comparison between real-time time series device data and preset time series device data, obtain cumulative distance, judge the identification results of each sensor according to the cumulative distance, and output sensor abnormal data according to the judgment result, and is also used to build a sensor fault identification model, obtain sensor status information, adjust the identification results of each sensor according to the sensor status information, and is also used to build a preset cumulative distance automatic adjustment model to obtain preset cumulative distance automatic adjustment information; The fault location identification unit is used to build a fault location identification model and output the faulty device location according to the sensor abnormal data.

3. The new energy station maintenance and repair system based on intelligent sensors according to claim 2 is characterized in that: When judging the recognition results of each sensor, the fault time recognition unit performs time series processing on the normal equipment operation data, divides the normal equipment operation data into time intervals according to preset unit time intervals, and constructs preset time series equipment data from the normal equipment operation data after the time interval division in time sequence, performs time series processing on the real-time equipment operation data, obtains the real-time time series equipment data, performs DTW comparison on the real-time time series equipment data and the preset time series equipment data, obtains the cumulative distance L, compares the cumulative distance L with the preset cumulative distance L0, judges the recognition results of each sensor according to the comparison results, and outputs the sensor abnormality data according to the judgment results, wherein: When L≤L0, the fault time identification unit determines that the sensor identification result is normal and does not output sensor abnormality data; When L>L0, the fault time identification unit determines that the sensor identification result is abnormal, and outputs sensor abnormality data to the administrator.

4. The new energy station maintenance and repair system based on intelligent sensors according to claim 3 is characterized in that: The fault time identification unit constructs a sensor fault identification model, inputs the real-time time series device data into the sensor fault identification model, obtains the output result of the sensor fault identification model, and uses it as the sensor state information, wherein the sensor state information includes the normal operation of the sensor and the abnormal operation of the sensor, and adjusts the identification results of each sensor according to the sensor state information, wherein: When the sensor status information indicates that the sensor is operating normally, the recognition result of the sensor is not adjusted; When the sensor status information indicates that the sensor is working abnormally, the sensor identification result is adjusted to be normal, and sensor maintenance alarm information is output to the administrator.

5. The intelligent sensor-based new energy station maintenance and repair system according to claim 4 is characterized in that: When the fault time identification unit makes real-time adjustment to the preset cumulative distance L0, a preset cumulative distance automatic adjustment model is constructed, and the preset cumulative distance database is divided into 70% of the preset cumulative distance training set, 15% of the preset cumulative distance verification set and 15% of the preset cumulative distance test set, and the preset cumulative distance training set is input into the random forest model for training to obtain the trained random forest model, and the preset cumulative distance verification set is input into the trained random forest model to iteratively optimize the parameters of the trained random forest model, and the preset cumulative distance test set is input into the iteratively optimized random forest model, and the iteratively optimized recurrent network neural model is tested to obtain the automatic adjustment accuracy U, and the automatic adjustment accuracy U is compared with the preset automatic adjustment accuracy U0, and the training compliance of the random forest model is judged according to the comparison result, and the random forest model is output according to the judgment result, wherein: When U≥U0, the random forest model training is determined to have met the standard, and the random forest model is output as the preset cumulative distance automatic adjustment model; When U<U0, it is determined that the random forest model training does not meet the standard, the random forest model parameters are iteratively optimized, and the training is repeated until the random forest model training meets the standard; The real-time device data is input into the preset cumulative distance automatic adjustment model to obtain the preset cumulative distance automatic adjustment information.

6. The new energy station maintenance and repair system based on intelligent sensors according to claim 2 is characterized in that: When locating a faulty device, the fault location identification unit constructs a fault location identification model, obtains sensor abnormality data through the fault time identification unit, inputs the sensor abnormality data into the fault location identification model, obtains the output result of the fault location identification model, and uses it as the faulty device location.

7. The new energy station maintenance and repair system based on intelligent sensors according to claim 1 is characterized in that: The fault maintenance module constructs an equipment maintenance model based on sensor abnormality data and a historical maintenance database, inputs real-time sensor abnormality data into the equipment maintenance model, obtains the output result of the equipment maintenance model, uses it as equipment maintenance information, compares the number of error outputs S of each equipment maintenance information with the preset number of error outputs S0, judges the real-time performance of the preset cumulative distance automatic adjustment model based on the comparison result, and updates the preset cumulative distance database based on the judgment result, wherein: When S<S0, the fault maintenance module determines that the real-time performance of the preset cumulative distance automatic adjustment model is normal, and does not update the preset cumulative distance database; When S≥S0, the fault maintenance module determines that the real-time performance of the preset cumulative distance automatic adjustment model is abnormal, collects actual fault information, updates the preset cumulative distance database according to the actual fault information, and trains the preset cumulative distance automatic adjustment model according to the updated preset cumulative distance database until the preset cumulative distance automatic adjustment model training meets the standard, and outputs the updated preset cumulative distance automatic adjustment model; The fault maintenance module compares the actual operation fault with the historical maintenance database, and judges the type of fault according to the comparison result, and adjusts the real-time judgment result of the preset cumulative distance automatic adjustment model according to the judgment result, wherein: When the actual operation fault is consistent with the historical maintenance database, the fault maintenance module determines that the fault type is a new fault, and adjusts the judgment result of the real-time performance of the preset cumulative distance automatic adjustment model to the real-time performance abnormality of the preset cumulative distance automatic adjustment model; When the actual operation fault is inconsistent with the historical maintenance database, the fault maintenance module determines that the fault occurrence type is a historical fault, and the fault maintenance module does not adjust the real-time judgment result of the preset cumulative distance automatic adjustment model.

8. The intelligent sensor-based new energy station maintenance and repair system according to claim 7 is characterized in that: The fault maintenance module compares the preset cumulative distance database update times C with the preset update times C0, judges the practicality of each device according to the comparison result, and outputs the device maintenance alarm information according to the judgment result, wherein: When C<C0, the fault maintenance module determines that the practicability of the device is normal and does not output the device maintenance alarm information; When C≥C0, the fault maintenance module determines that the practicability of the device is abnormal, outputs device maintenance alarm information, and notifies the administrator to perform maintenance on the device.

9. The intelligent sensor-based new energy station maintenance and repair system according to claim 8 is characterized in that: The fault maintenance module compares the equipment maintenance alarm information output times R with the preset equipment maintenance alarm information output times R0, judges the maintenance importance of each device according to the comparison result, and adjusts the output order of the equipment maintenance alarm information of each device according to the judgment result, wherein: When R<R0, the fault maintenance module determines that the maintenance importance of the equipment is high, and does not adjust the output sequence of the equipment maintenance alarm information of the equipment; When R≥R0, the fault maintenance module determines that the maintenance importance of the equipment is high, and adjusts the output sequence of the equipment maintenance alarm information of the equipment to priority output.

10. The maintenance and repair method of new energy stations based on intelligent sensors according to claim 1 is characterized in that: include: Step S1, collecting normal equipment operation data and real-time equipment operation data; Step S2, preprocessing normal equipment operation data and real-time equipment operation data; Step S3, performing DTW comparison on the real-time time series device data and the preset time series device data, obtaining the cumulative distance, and outputting the sensor abnormality data according to the cumulative distance; Step S4, constructing a sensor fault identification model, obtaining sensor status information, and adjusting the identification results of each sensor according to the sensor status information; Step S5, constructing a preset cumulative distance automatic adjustment model, and obtaining preset cumulative distance automatic adjustment information; Step S6, constructing a fault location recognition model and outputting the faulty device location according to the sensor abnormal data; Step S7, constructing a device maintenance model, obtaining device maintenance information, and updating a preset cumulative distance database according to the number of erroneous outputs of each device maintenance information; Step S8, outputting equipment maintenance alarm information according to the preset cumulative distance database update times; Step S9: adjusting the output sequence of the equipment maintenance alarm information of each device according to the number of times the equipment maintenance alarm information is output.

Images