UUV motor reliability data acquisition and analysis method and system

Abstract

The invention relates to the technical field of image recognition, in particular to a UUV (Unmanned Underwater Vehicle) motor reliability data acquisition and analysis method, which comprises the following steps: S1, tested equipment is a propelling system and comprises a main propelling motor and a servo driver; a parameter signal, read by the servo driver, of the rotating speed of the main push motor is converged into the signal conditioning and adapting equipment through an RS485 bus; s2, signals of sensors for detecting three-axis vibration, phase voltage, phase current and temperature and installed on the main push motor are collected by a data collecting and testing device and then collected into a signal conditioning and adapting device, analog signals collected by the sensors are converted into digital signals through the data collecting and testing device, and the digital signals are transmitted to a computer through a bus; by collecting and analyzing the operation data of the motor, the performance of the motor under different conditions can be better understood, so that the design of the motor is improved, the efficiency of the motor is improved, the service life of the motor is prolonged, and the data can help to predict the fault time of the motor, so that maintenance or replacement can be performed in advance, and accidental shutdown is avoided.

Description

Technical Field

[0001] The present invention relates to the field of image recognition technology, and in particular to a method and system for collecting and analyzing reliability data of a UUV motor. Background Art

[0002] The propulsion motor, as the core of the UUV (Unmanned Underwater Vehicle) propulsion system, is a key component for the normal operation of the entire UUV. Its reliability is one of the prerequisites for ensuring the safe and reliable operation of the underwater UUV. The reliability of the motor directly affects the safety of the UUV and the success rate of the mission. If the reliability level of the motor is low, it may cause the UUV to lose control, sink or be lost, resulting in mission failure and huge economic losses. Therefore, the reliability data of the motor is crucial to ensure the stability and safety of the UUV when performing missions.

[0003] The underwater unmanned vehicle propulsion motor is a special motor, and reliability data is one of the necessary conditions for the realization of propulsion motor fault detection, fault prediction, fault modeling, health status assessment and other functions. At present, the collection and analysis of reliability data of UUV propulsion motors is basically blank.

[0004] Therefore, in order to address the problem that the collection and analysis technology of the reliability data of the above-mentioned UUV propulsion motor is immature, resulting in the problem that the safety of the UUV motor cannot be guaranteed during use, a UUV motor reliability data collection and analysis method and system can be designed. The present invention adopts a semi-physical simulation architecture, which can improve the stability and safety of the UUV propulsion motor when performing tasks. The motor's reliability data can also be used to optimize the design and performance of the UUV. Summary of the Invention

[0005] In order to overcome the problem that the reliability data collection and analysis technology of UUV propulsion motors is immature, resulting in the problem that the safety of UUV motors cannot be guaranteed during use.

[0006] The technical solution of the present invention is: a UUV motor reliability data collection and analysis method, the steps of which are as follows:

[0007] S1: The device under test is a propulsion system, including a main propulsion motor and a servo drive. The parameter signal of the main propulsion motor speed read by the servo drive is fed into the signal conditioning and adaptation equipment via the RS485 bus.

[0008] S2: The sensors installed on the main propulsion motor to detect triaxial vibration, phase voltage, phase current, and temperature are collected by the data acquisition and testing equipment and then integrated into the signal conditioning and adaptation equipment. The sensors include current sensors, temperature sensors, vibration sensors, and noise sensors. The data acquisition and testing equipment is an important device that connects the sensors and the computer. It converts the analog signals collected by the sensors into digital signals and transmits them to the computer via the bus.

[0009] S3: Hardware Installation

[0010] The computer cabinet is equipped with a cloud server, time synchronization equipment, and edge processor. The above equipment, as well as external monitoring terminals, wireless networking equipment, and signal conditioning and adaptation equipment, are all connected to the network switching equipment inside the cabinet via network cables.

[0011] S4: The time synchronization device is connected to an external Beidou antenna to receive Beidou timing signals and provide NTP services for the entire network. The cloud server is equipped with corresponding data storage devices to save the collected data. The signal conditioning and adaptation equipment conditions the incoming data and sends it to the network switching equipment through the network cable for use by the cloud platform server and monitoring terminal.

[0012] S5: Data Processing

[0013] S51: Preprocess the collected data, including data cleaning and data standardization, and remove redundant, abnormal, and highly correlated data from the data set;

[0014] S52: Feature Selection

[0015] Extract key features related to faults from preprocessed data;

[0016] S53: Model selection

[0017] Based on the extracted features, select appropriate machine learning algorithms or deep learning models for training, including fault tree and SVM classification algorithms;

[0018] S53: Perform model training

[0019] Use the preprocessed dataset to train the selected model and divide the data into training set and validation set;

[0020] S54: Perform model evaluation

[0021] Use test data to evaluate the model and verify its performance and effectiveness. Depending on the analysis objectives, the model evaluation uses multiple indicators, including accuracy, recall rate, and F1 score.

[0022] Preferably, the data storage device in S4 includes a hard disk, a solid state drive or a cloud storage method.

[0023] Preferably, the feature selection method in S52 includes principal component analysis and feature importance assessment.

[0024] Preferably, in S53, the model parameters are adjusted to optimize the prediction performance. During the training process, a cross-validation method is used to evaluate the generalization ability of the model to avoid overfitting or underfitting. The model is trained using training data to optimize the model parameters.

[0025] A UUV motor reliability data acquisition and analysis system, which includes the UUV motor reliability data acquisition and analysis method as described above, and further includes a system management module, a data management module, a prediction algorithm module, and a health management module;

[0026] The system management module includes device management, network management, database management and log management;

[0027] The data management module is divided into two sub-modules: data acquisition and data processing. The data acquisition module includes data collection and transmission; the data processing module includes data analysis and data uploading:

[0028] The prediction algorithm module is divided into a data preprocessing module and a model research module. The data preprocessing module includes data cleaning, data standardization and sequence clipping; the model research includes feature selection, model selection, model training and model evaluation.

[0029] Preferably, in the system management module:

[0030] 1) Device management: Initialize, remotely operate and start testing the devices in the platform.

[0031] 2) Network management performs network configuration and connectivity testing on terminals and system applications connected to the network.

[0032] 3) Database management: Perform background management of network storage data and configuration management of the database.

[0033] 4) Log management monitors and records all system user settings and user operation behaviors.

[0034] Preferably, in the data management module:

[0035] 1) Data acquisition module

[0036] There are two ways to obtain data:

[0037] a) Read the BIT data built into the system under test

[0038] BIT uses a series of software and hardware configurations to monitor the status of key components within underwater long-endurance unmanned equipment. Based on the monitoring data of the BIT system, researchers and maintenance personnel use their professional knowledge for analysis and can also use expert systems to quickly and accurately diagnose faults and predict the life of the entire system and components.

[0039] The BIT of the propulsion system is installed on the servo drive, and the information obtained includes the motor speed, actual torque, and absolute current parameters. The BIT of the energy system is installed on the BMS, and the information obtained includes the voltage, current, and temperature parameters of the battery pack and the internal battery pack.

[0040] b) Read the sensor data installed on the system under test

[0041] The sensor information of the propulsion system includes the phase voltage, line voltage, vibration, temperature, and torque parameters of the motor, and the sensor information of the energy system includes the voltage, current, and temperature parameters of the battery pack.

[0042] 2) Data processing module

[0043] The data processing module includes data parsing and data uploading functions.

[0044] a) Data analysis

[0045] The collected BIT data information and sensor data information are parsed according to the requirements of the protocol format to extract valid data records.

[0046] b) Data upload

[0047] The parsed data records are uploaded to the cloud platform server and stored in the database.

[0048] Preferably, the prediction algorithm module also includes model research of the energy system.

[0049] As a preferred option, the model research process of the energy system is as follows:

[0050] a. Set up fault prediction and classification data model;

[0051] b. Obtain historical energy system fault data, remove duplicate values and noise points from the historical fault data, and then use empirical values to fill in the missing items; and after determining the fault type, generate prediction sample data;

[0052] c. Set 60% of the prediction sample data as training samples; set 40% of the prediction sample data as verification data;

[0053] d. Train the fault prediction data model by predicting sample data to obtain a trained fault prediction data model;

[0054] e. Validation data is used to verify through cross-validation whether the trained fault prediction data model meets the accuracy requirements;

[0055] f. Collect energy system data and perform fault prediction and abnormal data classification on the energy system data through the predictive data model;

[0056] g. Based on the prediction results and classification results, continuously optimize the fault prediction and classification data model, and set the fault prediction data model based on the decision tree model and support vector machine (SVM) algorithm.

[0057] The beneficial effects of the present invention are as follows: By collecting and analyzing motor operation data, we can better understand the performance of the motor under different conditions, thereby improving the motor design and increasing its efficiency and lifespan. This data can also help predict the failure time of the motor so that maintenance or replacement can be carried out in advance, thus avoiding unexpected downtime. The reliability data of UUV motors is of great significance for ensuring the safe operation of UUVs, improving mission success rates and optimizing system design. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 Shown is a schematic diagram of the platform hardware composition of the UUV motor reliability data acquisition and analysis method of the present invention;

[0059] Figure 2 Shown is a schematic diagram of the data management module of the UUV motor reliability data acquisition and analysis system of the present invention;

[0060] Figure 3 Shown is a schematic diagram of the prediction algorithm module composition of the UUV motor reliability data acquisition and analysis system of the present invention;

[0061] Figure 4 What is presented is a flow chart of the energy system model research of the UUV motor reliability data acquisition and analysis system of the present invention. DETAILED DESCRIPTION

[0062] The present invention will be further described below with reference to the accompanying drawings and examples.

[0063] See also Figure 1-Figure 4 The present invention provides an embodiment: a method for collecting and analyzing reliability data of a UUV motor, the steps of which are as follows:

[0064] S1: The device under test is a propulsion system, including a main propulsion motor and a servo drive. The parameter signal of the main propulsion motor speed read by the servo drive is fed into the signal conditioning and adaptation equipment via the RS485 bus.

[0065] S2: The sensors installed on the main propulsion motor to detect triaxial vibration, phase voltage, phase current, and temperature are collected by the data acquisition and testing equipment and then integrated into the signal conditioning and adaptation equipment. The sensors include current sensors, temperature sensors, vibration sensors, and noise sensors. The data acquisition and testing equipment is an important device that connects the sensors and the computer. It converts the analog signals collected by the sensors into digital signals and transmits them to the computer via the bus.

[0066] S3: Hardware Installation

[0067] The computer cabinet is equipped with a cloud server, time synchronization equipment, and edge processor. The above equipment, as well as external monitoring terminals, wireless networking equipment, and signal conditioning and adaptation equipment, are all connected to the network switching equipment inside the cabinet via network cables.

[0068] S4: The time synchronization device is connected to the Beidou antenna to receive Beidou timing signals and provide NTP services for the entire network. The cloud server is equipped with corresponding data storage devices to save the collected data. The signal conditioning and adaptation equipment conditions the incoming data and sends it to the network switching equipment through the network cable for access by the cloud platform server and monitoring terminal. The data storage device includes a hard disk, solid-state drive or cloud storage.

[0069] S5: Data Processing

[0070] S51: Preprocess the collected data, including data cleaning and data standardization, and remove redundant, abnormal, and highly correlated data from the data set;

[0071] S52: Feature Selection

[0072] Extract key features related to faults from preprocessed data. Feature selection methods include principal component analysis and feature importance assessment.

[0073] S53: Model selection

[0074] Based on the extracted features, select appropriate machine learning algorithms or deep learning models for training. Algorithms include fault tree and SVM classification algorithms. Optimize prediction performance by adjusting model parameters. During training, use cross-validation methods to evaluate the generalization ability of the model to avoid overfitting or underfitting. Use training data to train the model and optimize model parameters.

[0075] S53: Perform model training

[0076] Use the preprocessed dataset to train the selected model and divide the data into training set and validation set;

[0077] S54: Perform model evaluation

[0078] Use test data to evaluate the model and verify its performance and effectiveness. Depending on the analysis objectives, the model evaluation uses multiple indicators, including accuracy, recall rate, and F1 score.

[0079] A UUV motor reliability data acquisition and analysis system, which includes the UUV motor reliability data acquisition and analysis method as described above, and further includes a system management module, a data management module, a prediction algorithm module, and a health management module;

[0080] The system management module includes device management, network management, database management, and log management, among which:

[0081] 1) Device management: Initialize, remotely operate and start testing the devices in the platform.

[0082] 2) Network management performs network configuration and connectivity testing on terminals and system applications connected to the network.

[0083] 3) Database management: Perform background management of network storage data and configuration management of the database.

[0084] 4) Log management monitors and records all system user settings and user operations;

[0085] The data management module is divided into two sub-modules: data acquisition and data processing. The data acquisition module includes data collection and transmission; the data processing module includes data analysis and data upload.

[0086] 1) Data acquisition module

[0087] There are two ways to obtain data:

[0088] a) Read the BIT data built into the system under test

[0089] BIT uses a series of software and hardware configurations to monitor the status of key components within underwater long-endurance unmanned equipment. Based on the monitoring data of the BIT system, researchers and maintenance personnel use their professional knowledge for analysis and can also use expert systems to quickly and accurately diagnose faults and predict the life of the entire system and components.

[0090] The BIT of the propulsion system is installed on the servo drive, and the information obtained includes the motor speed, actual torque, and absolute current parameters. The BIT of the energy system is installed on the BMS, and the information obtained includes the voltage, current, and temperature parameters of the battery pack and the internal battery pack.

[0091] b) Read the sensor data installed on the system under test

[0092] The sensor information of the propulsion system includes the phase voltage, line voltage, vibration, temperature, and torque parameters of the motor, and the sensor information of the energy system includes the voltage, current, and temperature parameters of the battery pack.

[0093] 2) Data processing module

[0094] The data processing module includes data parsing and data uploading functions.

[0095] a) Data analysis

[0096] The collected BIT data information and sensor data information are parsed according to the requirements of the protocol format to extract valid data records.

[0097] b) Data upload

[0098] Upload the parsed data records to the cloud platform server and store them in the database:

[0099] The prediction algorithm module is divided into a data preprocessing module and a model research module. The data preprocessing module includes data cleaning, data standardization, and sequence clipping; the model research module includes feature selection, model selection, model training, and model evaluation. The prediction algorithm module also includes energy system model research. The energy system model research process is as follows:

[0100] a. Set up fault prediction and classification data model;

[0101] b. Obtain historical energy system fault data, remove duplicate values and noise points from the historical fault data, and then use empirical values to fill in the missing items; and after determining the fault type, generate prediction sample data;

[0102] c. Set 60% of the prediction sample data as training samples; set 40% of the prediction sample data as verification data;

[0103] d. Train the fault prediction data model by predicting sample data to obtain a trained fault prediction data model;

[0104] e. Validation data is used to verify through cross-validation whether the trained fault prediction data model meets the accuracy requirements;

[0105] f. Collect energy system data and perform fault prediction and abnormal data classification on the energy system data through the predictive data model;

[0106] g. Based on the prediction results and classification results, continuously optimize the fault prediction and classification data model, and set the fault prediction data model based on the decision tree model and support vector machine (SVM) algorithm.

[0107] The embodiments of the present invention are described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Various changes can be made within the scope of knowledge of those skilled in the art without departing from the spirit of the present invention.

Claims

1. A UUV motor reliability data collection and analysis method, characterized in that: The steps are as follows: S1: The device under test is a propulsion system, including a main propulsion motor and a servo drive. The parameter signal of the main propulsion motor speed read by the servo drive is fed into the signal conditioning and adaptation equipment via the RS485 bus. S2: The sensors installed on the main propulsion motor to detect triaxial vibration, phase voltage, phase current, and temperature are collected by the data acquisition and testing equipment and then integrated into the signal conditioning and adaptation equipment. The sensors include current sensors, temperature sensors, vibration sensors, and noise sensors. The data acquisition and testing equipment is an important device that connects the sensors and the computer. It converts the analog signals collected by the sensors into digital signals and transmits them to the computer via the bus. S3: Hardware Installation The computer cabinet is equipped with a cloud server, time synchronization equipment, and edge processor. The above equipment, as well as external monitoring terminals, wireless networking equipment, and signal conditioning and adaptation equipment, are all connected to the network switching equipment inside the cabinet via network cables. S4: The time synchronization device is connected to an external Beidou antenna to receive Beidou timing signals and provide NTP services for the entire network. The cloud server is equipped with corresponding data storage devices to save the collected data. The signal conditioning and adaptation equipment conditions the incoming data and sends it to the network switching equipment through the network cable for use by the cloud platform server and monitoring terminal. S5: Data Processing S51: Preprocess the collected data, including data cleaning and data standardization, and remove redundant, abnormal, and highly correlated data from the data set; S52: Feature Selection Extract key features related to faults from preprocessed data; S53: Model selection Based on the extracted features, select appropriate machine learning algorithms or deep learning models for training, including fault tree and SVM classification algorithms; S53: Perform model training Use the preprocessed dataset to train the selected model and divide the data into training set and validation set; S54: Perform model evaluation Use test data to evaluate the model and verify its performance and effectiveness. Depending on the analysis objectives, the model evaluation uses multiple indicators, including accuracy, recall rate, and F1 score.

2. The UUV motor reliability data acquisition and analysis method according to claim 1 is characterized in that: The data storage devices in S4 include hard disks, solid-state drives, or cloud storage.

3. The UUV motor reliability data acquisition and analysis method according to claim 1 is characterized in that: The feature selection methods in S52 include principal component analysis and feature importance assessment.

4. The UUV motor reliability data acquisition and analysis method according to claim 1 is characterized in that: In S53, the model parameters are adjusted to optimize the prediction performance. During the training process, the cross-validation method is used to evaluate the generalization ability of the model to avoid overfitting or underfitting. The model is trained using training data to optimize the model parameters.

5. UUV motor reliability data acquisition and analysis system, characterized by The method for collecting and analyzing reliability data of a UUV motor according to claim 1 further comprises a system management module, a data management module, a prediction algorithm module, and a health management module; The system management module includes device management, network management, database management and log management; The data management module is divided into two sub-modules: data acquisition and data processing. The data acquisition module includes data collection and transmission; the data processing module includes data analysis and data uploading: The prediction algorithm module is divided into a data preprocessing module and a model research module. The data preprocessing module includes data cleaning, data standardization and sequence clipping; the model research includes feature selection, model selection, model training and model evaluation.

6. The UUV motor reliability data acquisition and analysis system according to claim 5, characterized in that: In the system management module: 1) Device management: Initialize, remotely operate and start testing the devices in the platform. 2) Network management performs network configuration and connectivity testing on terminals and system applications connected to the network. 3) Database management: Perform background management of network storage data and configuration management of the database. 4) Log management monitors and records all system user settings and user operation behaviors.

7. The UUV motor reliability data acquisition and analysis system according to claim 5, characterized in that: In the data management module: 1) Data acquisition module There are two ways to obtain data: a) Read the BIT data built into the system under test BIT uses a series of software and hardware configurations to monitor the status of key components within underwater long-endurance unmanned equipment. Based on the monitoring data of the BIT system, researchers and maintenance personnel use their professional knowledge for analysis and can also use expert systems to quickly and accurately diagnose faults and predict the life of the entire system and components. The BIT of the propulsion system is installed on the servo drive, and the information obtained includes the motor speed, actual torque, and absolute current parameters. The BIT of the energy system is installed on the BMS, and the information obtained includes the voltage, current, and temperature parameters of the battery pack and the internal battery pack. b) Read the sensor data installed on the system under test The sensor information of the propulsion system includes the phase voltage, line voltage, vibration, temperature, and torque parameters of the motor, and the sensor information of the energy system includes the voltage, current, and temperature parameters of the battery pack. 2) Data processing module The data processing module includes data parsing and data uploading functions. a) Data analysis The collected BIT data information and sensor data information are parsed according to the requirements of the protocol format to extract valid data records. b) Data upload The parsed data records are uploaded to the cloud platform server and stored in the database.

8. The UUV motor reliability data acquisition and analysis system according to claim 5, characterized in that: The prediction algorithm module also includes model research on energy systems.

9. The UUV motor reliability data acquisition and analysis system according to claim 8, characterized in that: The model research process of the energy system is as follows: a. Set up fault prediction and classification data model; b. Obtain historical energy system fault data, remove duplicate values and noise points from the historical fault data, and then use empirical values to fill in the missing items; and after determining the fault type, generate prediction sample data; c. Set 60% of the prediction sample data as training samples; Set 40% of the prediction sample data as validation data; d. Train the fault prediction data model by predicting sample data to obtain a trained fault prediction data model; e. Validation data is used to verify through cross-validation whether the trained fault prediction data model meets the accuracy requirements; f. Collect energy system data and perform fault prediction and abnormal data classification on the energy system data through the predictive data model; g. Based on the prediction results and classification results, continuously optimize the fault prediction and classification data model, and set the fault prediction data model based on the decision tree model and support vector machine (SVM) algorithm.

Images