A Health Status Assessment System and Method for Unmanned Surface Vessel Power Plant

By constructing a health status assessment system for unmanned surface vessel (USV) power units and using the ResCNN-TCN model to assess the health status of USVs, the problem of poor condition monitoring in existing technologies is solved, and real-time health assessment and effective maintenance decisions for USV power units are realized.

CN117775226BActive Publication Date: 2026-11-17JIANGSU UNIV OF SCI & TECH
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202311574251.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-11-22
Publication Date
2026-11-17
Estimated Expiration
2043-11-22

AI Technical Summary

Technical Problem

Existing technologies for monitoring the condition of unmanned surface vessel (USV) power systems are ineffective, making it difficult to assess their health status in real time and make effective maintenance decisions.

Method used

An unmanned surface vessel (USV) power plant health status assessment system is adopted, which includes a data acquisition module, an onboard data preprocessing module, a shore-to-USV communication module, and a shore-based monitoring and management module. The ResCNN-TCN model is used to assess the health status of the USV and provide maintenance decisions.

Benefits of technology

It improves the accuracy of condition prediction and health assessment, enabling timely and effective maintenance decisions and ensuring the reliability of the unmanned surface vessel's power plant.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117775226B_ABST
    Figure CN117775226B_ABST
Patent Text Reader

Abstract

This invention discloses a health status assessment system for an unmanned surface vessel (USV) propulsion system, comprising: an USV propulsion system, characterized in that it further comprises: a data acquisition module, an onboard data preprocessing module, a shore-to-US communication module, and a shore-based monitoring and management module; the data acquisition module is connected to the USV propulsion system and is used to acquire various parameter data of the USV propulsion system; the onboard data preprocessing module is connected to the data acquisition module and is used to preprocess the acquired parameter data; the shore-to-US communication module is communicatively connected to both the onboard data preprocessing module and the shore-based monitoring and management module, and is used to transmit data between the onboard data preprocessing module and the shore-based monitoring and management module; the shore-based monitoring and management module assesses the health status of the USV based on the acquired parameter data of the USV propulsion system and provides corresponding maintenance decisions. This invention uses a newly constructed ResCNN-TCN model to predict various characteristic states, effectively improving the accuracy of state prediction.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of unmanned surface vessel (USV) health assessment technology, and particularly to a system and method for assessing the health status of USV power units. Background Technology

[0002] Currently, the power plant, as a crucial component of unmanned surface vessels (USVs), provides them with operational power. However, it operates in a highly corrosive marine environment for extended periods, resulting in a high probability of malfunction. Furthermore, a failure can lead to a series of negative consequences, potentially causing serious accidents. Therefore, ensuring and improving the reliability and maintainability of the power plant, and conducting health status assessments and developing appropriate maintenance decisions, are of paramount importance.

[0003] To address these issues and ensure the normal operation of unmanned surface vessel (USV) power units, it is crucial to assess the health status of the power units and provide timely and effective maintenance decisions. In recent years, USV condition monitoring systems and methods have been gradually studied, but relatively few studies have been conducted. For example, Li Zhaoyang et al. from the Comprehensive Technology and Economic Research Institute of China Shipbuilding Group Corporation have invented and disclosed a condition monitoring and maintenance support system for USVs. Through module settings, it can monitor the operating status of various equipment on USVs and facilitate fault repair in a timely and effective manner. However, it lacks a specific method for real-time health status assessment. (Li Zhaoyang, Yang Cheng, Yan Ge, Jiang Tao, Liu Hongxiao, Hao Guangyao, Yan Yusheng, Huang Hua, Yang Lian, Yang Qianqian. A Condition Monitoring and Maintenance Support System for Unmanned Surface Vessels [P]. Beijing: CN115964883A, 2023-04-14.)

[0004] These methods can effectively monitor the status of equipment and make maintenance decisions, but they also have shortcomings, such as poor monitoring of the status of unmanned surface vessel power units, difficulty in real-time assessment of health status, and difficulty in making effective maintenance decisions when the device malfunctions. Therefore, this invention provides an unmanned surface vessel power unit health status assessment system and method to solve the above-mentioned shortcomings. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a health status assessment system and method for unmanned surface vessel (USV) power units, thereby solving the technical problems of poor monitoring of USV power unit status, difficulty in real-time assessment of health status, and difficulty in making effective maintenance decisions when the unit malfunctions.

[0006] This invention provides a health status assessment system for an unmanned surface vessel (USV) power unit, comprising: an USV power unit, a data acquisition module, an onboard data preprocessing module, a shore-to-ship communication module, and a shore-based monitoring and management module;

[0007] The data acquisition module is connected to the unmanned surface vessel's (USV) power unit and is used to collect various parameter data of the USV's power unit. The onboard data preprocessing module is connected to the data acquisition module and is used to preprocess the collected parameter data. The shore-based communication module is connected to both the onboard data preprocessing module and the shore-based monitoring and management module and is used to transmit data between the onboard data preprocessing module and the shore-based monitoring and management module. Based on the acquired parameter data of the USV's power unit, the shore-based monitoring and management module assesses the health status of the USV using the constructed ResCNN-TCN model and provides corresponding maintenance decisions.

[0008] Furthermore, the unmanned surface vessel's power unit includes: a diesel engine, a propeller, a waterjet propulsion system, a transmission system, and a battery energy storage element.

[0009] Furthermore, the data acquisition module includes a vibration acquisition unit, a current acquisition unit, a voltage acquisition unit, a speed acquisition unit, an oil level acquisition unit, a temperature acquisition unit, a humidity acquisition unit, an electrical energy acquisition unit, and other acquisition units;

[0010] The vibration acquisition unit is used to acquire vibration signals of the power unit; the current acquisition unit is used to acquire current signals during the operation of the power unit; the voltage acquisition unit is used to acquire voltage signals during the operation of the power unit; the speed acquisition unit is used to acquire the speed of the power unit; the oil level acquisition unit is used to acquire the oil quantity of the power unit; the temperature acquisition unit is used to acquire the temperature of the power unit itself and the ambient temperature during operation; the humidity acquisition unit is used to acquire the humidity of the environment surrounding the power unit; the electrical energy acquisition unit is used to acquire the charge of the battery energy storage element; and other acquisition units are used to acquire other parameters of the power unit.

[0011] Furthermore, the onboard data preprocessing module includes a data cleaning unit, a data normalization unit, and a data segmentation unit;

[0012] The data cleaning unit identifies and removes outlier data, and supplements and estimates missing data.

[0013] The data normalization unit adopts the maximum-minimum normalization method to unify the data to the range [0,1].

[0014] The data segmentation unit divides the normalized data, retaining the temporal correlation between adjacent sequences.

[0015] Furthermore, the shore-based monitoring and management module includes: a health status assessment unit and a maintenance decision-making unit;

[0016] The health status assessment unit performs a health status assessment based on the data information uploaded by the onboard data preprocessing module.

[0017] The maintenance decision-making unit makes decisions and maintenance recommendations based on the health status.

[0018] The present invention also provides a method for assessing the health status of an unmanned surface vessel (USV) propulsion system, applicable to the aforementioned USV propulsion system health status assessment system, comprising the following steps:

[0019] Step 1: Collect and acquire time series data of various characteristic parameters of the power unit;

[0020] Step 2: Preprocess the collected time series data;

[0021] Step 3: Mark the health status of the preprocessed data;

[0022] Step 4: Construct a ResCNN-TCN model, and use the ResCNN-TCN model to predict the state of each feature to obtain the state prediction value;

[0023] The ResCNN-TCN model features are sequentially connected as follows: an extended convolutional layer, a first max-pooling layer, a convolutional layer, a second max-pooling layer, a residual unit, a Flatten layer, and a first extended causal convolutional layer.

[0024] After the first dilated causal convolutional layer, the network splits into two paths. One path includes a first dropout layer, a second dilated causal convolutional layer, and a second dropout layer connected in sequence. The other path includes a 1*1 convolutional identity mapping layer. The second dropout layer and the 1*1 convolutional identity mapping layer are summed and then connected to a fully connected layer and an output layer in sequence.

[0025] Step 5: Train the ResCNN-TCN model using the training and test sets;

[0026] Step 6: Using the trained ResCNN-TCN model, obtain the state prediction values ​​of each feature parameter in real time, and calculate the estimated values ​​of health indicators based on the state prediction values;

[0027] Step 7: Calculate the overall health of the power unit based on the estimated health indicators of each feature, and provide health status warnings and corresponding maintenance decisions based on the overall health.

[0028] Furthermore, the first dilated causal convolutional layer and the first dropout layer undergo layer normalization and ReLU activation function processing; the second dilated causal convolutional layer and the second dropout layer undergo layer normalization and ReLU activation function processing.

[0029] Furthermore, in step 5, the model is continuously optimized using the Adam algorithm during the training process of the ResCNN-TCN model; the accuracy of the prediction results of the ResCNN-TCN model is evaluated by the root mean square error and the mean absolute error.

[0030] Furthermore, in step 6, the formula for calculating the health indicator assessment value is as follows:

[0031]

[0032] In the formula, h j y represents the health indicator assessment value of feature parameter j; j Y represents the state prediction output value of feature parameter j; j The average value of characteristic parameter j within the normal operating limits of the data; y min The minimum limit value for the characteristic parameter j; y max This represents the maximum limit value for the characteristic parameter j.

[0033] Furthermore, in step 7, the formula for calculating the overall health of the power unit is as follows:

[0034]

[0035] In the formula, m is the number of characteristic parameters; H represents the weight of each characteristic parameter on the health status of the device, and H represents the overall health of the power unit affected by multiple characteristic parameters.

[0036] The beneficial effects of this invention are:

[0037] This invention uses a newly constructed ResCNN-TCN model to predict the state of each feature. It can fully extract the degradation features of the data and the temporal features in the data, which can effectively improve the accuracy of state prediction and thus improve the accuracy of the health index evaluation values ​​of each feature. At the same time, it combines the weights of multiple feature parameters to obtain the comprehensive health of the power unit, thereby predicting the health status of the unmanned surface vessel's power unit and providing corresponding maintenance decisions for the unmanned surface vessel's power unit in a timely manner.

[0038] This invention uses a health index assessment formula to perform real-time health assessment of feature parameters, further processes the state prediction values ​​output by the model, thereby improving the accuracy of health assessment. Furthermore, by assigning weights to each feature parameter based on its influence on the health status of the power unit, the final comprehensive health score can objectively and accurately predict the health status of the power unit. Attached Figure Description

[0039] The features and advantages of the invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the invention in any way. In the drawings:

[0040] Figure 1 This is a system block diagram of a specific embodiment of the present invention;

[0041] Figure 2 This is a flowchart of a specific embodiment of the invention;

[0042] Figure 3 This is a schematic diagram of the ResCNN-TCN model structure according to a specific embodiment of the invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] The present invention will be further illustrated below with reference to specific embodiments. Those skilled in the art should understand that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Modifications to the present invention in various equivalent forms all fall within the scope defined by the appended claims.

[0045] like Figure 1 As shown in the figure, a specific embodiment of the present invention provides a health status assessment system for an unmanned surface vessel (USV) power unit, including: an USV power unit, a data acquisition module, an onboard data preprocessing module, a shore-to-surface communication module, and a shore-based monitoring and management module;

[0046] The data acquisition module is connected to the unmanned surface vessel's (USV) power unit and is used to collect various parameter data of the USV's power unit. The onboard data preprocessing module is connected to the data acquisition module and is used to preprocess the collected parameter data. The shore-based communication module is connected to both the onboard data preprocessing module and the shore-based monitoring and management module and is used to transmit data between the onboard data preprocessing module and the shore-based monitoring and management module. Based on the acquired parameter data of the USV's power unit, the shore-based monitoring and management module assesses the health status of the USV using the constructed ResCNN-TCN model and provides corresponding maintenance decisions.

[0047] The unmanned surface vessel's power plant includes: a diesel engine, a propeller, a waterjet propulsion system, a transmission system, and battery energy storage components.

[0048] The data acquisition module includes a vibration acquisition unit, a current acquisition unit, a voltage acquisition unit, a speed acquisition unit, an oil level acquisition unit, a temperature acquisition unit, a humidity acquisition unit, an electrical energy acquisition unit, and other acquisition units;

[0049] The vibration acquisition unit is used to acquire vibration signals of the power unit; the current acquisition unit is used to acquire current signals during the operation of the power unit; the voltage acquisition unit is used to acquire voltage signals during the operation of the power unit; the speed acquisition unit is used to acquire the speed of the power unit; the oil level acquisition unit is used to acquire the oil quantity of the power unit; the temperature acquisition unit is used to acquire the temperature of the power unit itself and the ambient temperature during operation; the humidity acquisition unit is used to acquire the humidity of the environment surrounding the power unit; the electrical energy acquisition unit is used to acquire the charge of the battery energy storage element; and other acquisition units are used to acquire other parameters of the power unit.

[0050] The onboard data preprocessing module includes a data cleaning unit, a data normalization unit, and a data segmentation unit;

[0051] The data cleaning unit identifies and removes outlier data, and supplements and estimates missing data.

[0052] The data normalization unit adopts the maximum-minimum normalization method to unify the data to the range [0,1].

[0053] The data segmentation unit divides the normalized data.

[0054] The shore-based monitoring and management module includes: a health status assessment unit and a maintenance decision unit;

[0055] The health status assessment unit performs a health status assessment based on the data information uploaded by the onboard data preprocessing module.

[0056] The maintenance decision-making unit makes decisions and maintenance recommendations based on the health status.

[0057] like Figure 2 As shown in the figure, a specific embodiment of the present invention also provides a method for assessing the health status of an unmanned surface vessel (USV) propulsion system, applicable to the aforementioned USV propulsion system health status assessment system, comprising the following steps:

[0058] Step 1: Collect and acquire time series data of various characteristic parameters of the power unit;

[0059] Characteristic parameters to select include device vibration, operating current, operating voltage, speed, oil level, device temperature, ambient temperature, ambient humidity, and remaining power.

[0060] Step 2: Preprocess the collected time series data;

[0061] The onboard data preprocessing module cleans and normalizes the data collected by the data acquisition module, and then uses a sliding time window segmentation method to further segment the normalized data.

[0062] Data normalization is performed using max-min normalization, with the specific formula as follows:

[0063]

[0064] Where y' is the normalized data, y is the original data, and y' is the normalized data. max and y min These are the maximum and minimum values ​​of the original data, respectively.

[0065] Step 3: Mark the processed data as healthy;

[0066] The processed data is uploaded to the shore-based storage via the shore-to-ship communication module. The historically stored datasets are labeled with health information, and the labeled datasets are divided into training and test sets at a ratio of 80% and 20%, respectively. Newly stored data is not labeled, and the unlabeled datasets are used as the dataset for real-time health status assessment.

[0067] Time series data y with health markers and various characteristic parameters j (t=1),y j (t=2),...,y j The value of (t=n) is between [0,1], where j represents the feature parameter, n represents the time series length, 0 represents that the feature parameter is in a completely faulty state at time t, and 1 represents that the feature parameter is in a completely healthy state at time t. The specific health status labels are shown in Table 1.

[0068]

[0069] Table 1

[0070] Step 4: As Figure 3 As shown, a ResCNN-TCN model is constructed, and the state of each feature is predicted by the ResCNN-TCN model to obtain the state prediction value;

[0071] The ResCNN-TCN model features are sequentially connected as follows: an extended convolutional layer, a first max-pooling layer, a convolutional layer, a second max-pooling layer, a residual unit, a Flatten layer, and a first extended causal convolutional layer.

[0072] After the first dilated causal convolutional layer, the network splits into two paths. One path includes a first dropout layer, a second dilated causal convolutional layer, and a second dropout layer connected in sequence. The other path includes a 1*1 convolutional identity mapping layer. The second dropout layer and the 1*1 convolutional identity mapping layer are summed and then connected to a fully connected layer and an output layer in sequence.

[0073] The first dilated causal convolutional layer and the first dropout layer are processed by layer normalization and ReLU activation function; the second dilated causal convolutional layer and the second dropout layer are processed by layer normalization and ReLU activation function.

[0074] Because the environment in which the unmanned surface vessel's propulsion system operates is complex and variable, and the failure degradation mechanism is also quite complex, in order to extract the global features of the degradation process in a shorter time, we first expand the convolutional layer and use a large convolutional kernel to expand the receptive field and better extract short-term local features. The expanded convolutional layer is followed by a first max pooling layer to expand the dimensionality reduction of the data after convolution. Then, a convolutional layer is connected, which uses a small convolutional kernel to reduce the network parameters, which is beneficial to deepen the network and suppress overfitting. Finally, it is connected to a second max pooling layer to continue the dimensionality reduction of the data.

[0075] As the network depth increases, the training error increases, which leads to insufficient feature extraction from the unmanned surface vessel's power unit signal data. Therefore, residual units are used to continue to extract features at a deeper level. Residual units can effectively solve the network degradation phenomenon caused by the increase in network depth.

[0076] The features output by the residual unit are transformed into a one-dimensional vector through the Flatten layer. In order to extract features on the time scale more quickly, an extended causal convolutional layer is used. It is better to use the extended causal convolution technique to improve the receptive field without increasing the computational cost. The extended causal convolutional layer does not use data after time t. Then it is input into the first extended causal convolutional layer and the second extended causal convolutional layer.

[0077] A 1x1 convolutional layer is used to implement cross-layer connections to solve the gradient vanishing problem caused by convolutional degradation. This ensures that the scale of the input to the dilated causal convolutional layer is the same as the scale of the output after cross-layer connections through the 1x1 convolutional layer. Finally, through a fully connected layer and an output layer, the state prediction output values ​​of each feature parameter are output.

[0078] Step 5: Train the ResCNN-TCN model using the training and test sets; continuously optimize the model using the Adam algorithm during training; evaluate the accuracy of the ResCNN-TCN model's prediction results using root mean square error and mean absolute error.

[0079] Step 6: Using the trained ResCNN-TCN model, obtain the state prediction values ​​of each root feature parameter in real-time data, and calculate the estimated values ​​of health indicators based on the state prediction values; the formula for calculating the health indicator evaluation values ​​is as follows:

[0080]

[0081] In the formula, h j y represents the health indicator assessment value of feature parameter j; j Y represents the state prediction output value of feature parameter j; j ymin is the average value of feature parameter j within its normal operating limits in the data; ymin is the minimum limit value of feature parameter j; y max This represents the maximum limit value of the characteristic parameter j during operation.

[0082] The maximum limit, minimum limit, and average value within the normal operating limit range of the characteristic parameter are determined using the law of large numbers and the normal distribution method. The specific formula is as follows:

[0083]

[0084] y min =Y j -3σ

[0085] y max =Y j +3σ

[0086] That

[0087] In the formula, n is the time series within the normal operating limit range, and Y j Let y be the average value of the characteristic parameter j within its normal operating limits in the data. min The minimum limit value of the characteristic parameter j is y. max This represents the maximum limit value of the characteristic parameter j during operation. Standard deviation;

[0088] Step 7: Calculate the overall health status of the power unit based on the estimated health indicators of each feature, and provide health status warnings and corresponding maintenance decisions based on the overall health status;

[0089] The formula for calculating the overall health of the power unit is as follows:

[0090]

[0091] In the formula, m is the number of characteristic parameters; H represents the weight of each characteristic parameter on the health status of the device, and H represents the overall health of the power unit affected by multiple characteristic parameters.

[0092] Historical data shows that the impact of each characteristic parameter on the power unit varies considerably. The overall health score obtained by assigning weights can accurately measure the health status of the power unit.

[0093] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for assessing the health status of an unmanned surface vessel's propulsion system, characterized in that, The method for assessing the health status of the unmanned surface vessel's power unit includes the following steps: Step 1: Collect and acquire time series data of various characteristic parameters of the power unit; Step 2: Preprocess the collected time series data; Step 3: Mark the health status of the preprocessed data; Step 4: Construct a ResCNN-TCN model, and use the ResCNN-TCN model to predict the state of each feature to obtain the state prediction value; The ResCNN-TCN model features are sequentially connected as follows: an extended convolutional layer, a first max-pooling layer, a convolutional layer, a second max-pooling layer, a residual unit, a Flatten layer, and a first extended causal convolutional layer. After the first dilated causal convolutional layer, the network splits into two paths. One path includes a first dropout layer, a second dilated causal convolutional layer, and a second dropout layer connected in sequence. The other path includes a 1*1 convolutional identity mapping layer. The second dropout layer and the 1*1 convolutional identity mapping layer are summed and then connected to a fully connected layer and an output layer in sequence. The first dilated causal convolutional layer and the first dropout layer are processed by layer normalization and ReLU activation function; the second dilated causal convolutional layer and the second dropout layer are processed by layer normalization and ReLU activation function. Step 5: Train the ResCNN-TCN model using the training and test sets. Continuously optimize the model using the Adam algorithm during training. Evaluate the accuracy of the ResCNN-TCN model's predictions using root mean square error and mean absolute error. Step 6: Using the trained ResCNN-TCN model, obtain the state prediction values ​​of each root feature parameter in real-time data, and calculate the estimated values ​​of health indicators based on the state prediction values. The formula for calculating the health indicator evaluation values ​​is as follows: ; In the formula, h j y represents the health indicator assessment value of feature parameter j; j Y represents the state prediction output value of feature parameter j; j The average value of characteristic parameter j within the normal operating limits of the data; y min The minimum limit value for the characteristic parameter j; y max This represents the maximum limit value of the characteristic parameter j during operation; Step 7: Calculate the overall health of the power unit based on the estimated health indicators of each feature, and provide health status warnings and corresponding maintenance decisions based on the overall health. The formula for calculating the overall health of the power unit is as follows: ; In the formula, m is the number of characteristic parameters; p j (p j ∈[0,1], ) represents the weight of each characteristic parameter on the health status of the device, and H represents the overall health of the power unit affected by multiple characteristic parameters.

2. An evaluation system for performing the unmanned surface vessel power plant health status assessment method as described in claim 1, characterized in that, include: The unmanned surface vessel (USV) power unit is characterized by further comprising: a data acquisition module, an onboard data preprocessing module, a shore-to-surface communication module, and a shore-based monitoring and management module; The data acquisition module is connected to the unmanned surface vessel's (USV) power unit and is used to collect various parameter data of the USV's power unit. The onboard data preprocessing module is connected to the data acquisition module and is used to preprocess the collected parameter data. The shore-based communication module is connected to both the onboard data preprocessing module and the shore-based monitoring and management module to transmit data between the onboard data preprocessing module and the shore-based monitoring and management module. Based on the acquired parameter data of the USV's power unit, the shore-based monitoring and management module assesses the health status of the USV using a constructed ResCNN-TCN model and provides corresponding maintenance decisions.

3. The unmanned surface vessel power plant health status assessment system as described in claim 2, characterized in that, The unmanned surface vessel's power unit includes: a diesel engine, a propeller, a waterjet propulsion system, a transmission system, and a battery energy storage element.

4. The unmanned surface vessel power plant health status assessment system as described in claim 2, characterized in that, The data acquisition module includes a vibration acquisition unit, a current acquisition unit, a voltage acquisition unit, a speed acquisition unit, an oil level acquisition unit, a temperature acquisition unit, a humidity acquisition unit, an electrical energy acquisition unit, and other acquisition units; The vibration acquisition unit is used to acquire vibration signals of the power unit; the current acquisition unit is used to acquire current signals during the operation of the power unit; the voltage acquisition unit is used to acquire voltage signals during the operation of the power unit; the speed acquisition unit is used to acquire the speed of the power unit; the oil level acquisition unit is used to acquire the oil quantity of the power unit; the temperature acquisition unit is used to acquire the temperature of the power unit itself and the ambient temperature during operation; the humidity acquisition unit is used to acquire the humidity of the environment surrounding the power unit; the electrical energy acquisition unit is used to acquire the charge of the battery energy storage element; and other acquisition units are used to acquire other parameters of the power unit.

5. The unmanned surface vessel power plant health status assessment system as described in claim 2, characterized in that, The onboard data preprocessing module includes a data cleaning unit, a data normalization unit, and a data segmentation unit. The data cleaning unit identifies and removes outlier data, and supplements and estimates missing data. The data normalization unit adopts the maximum-minimum normalization method to unify the data to the range [0,1]. The data segmentation unit divides the normalized data, retaining the temporal correlation between adjacent sequences.

6. The unmanned surface vessel power plant health status assessment system as described in claim 2, characterized in that, The shore-based monitoring and management module includes: a health status assessment unit and a maintenance decision-making unit; The health status assessment unit performs a health status assessment based on the data information uploaded by the onboard data preprocessing module. The maintenance decision-making unit makes decisions and maintenance recommendations based on the health status.

Citation Information

Patent Citations

  • State monitoring and maintenance support system of unmanned ship

    CN115964883A

  • Ship diesel engine health management and control system and management and control method thereof

    CN114048629A

  • System and method for predicting residual life of ship power equipment

    CN114881342A