Underwater vehicle fault diagnosis method and device based on physical data hybrid modeling strategy, equipment and medium

Abstract

The invention relates to the technical field of fault diagnosis, and discloses an underwater vehicle fault diagnosis method based on a physical data hybrid modeling strategy, at least adopts a steerage prediction physical model and a steerage prediction neural network, and comprises the following steps: inputting a kinetic parameter into the steerage prediction physical model for steerage prediction to obtain a theoretical driving prediction steerage; inputting the kinetic parameters into a steerage prediction neural network for steerage prediction to obtain a data-driven prediction steerage; performing fault diagnosis according to the theoretical driving prediction steerage, the theoretical prediction confidence, the data driving prediction steerage, the data prediction confidence and the kinetic parameters to obtain a fault diagnosis result; and performing fault emergency decision according to the fault diagnosis result and the operation state parameters of the underwater vehicle, and outputting a fault-tolerant control strategy. The method has the beneficial effects that fault diagnosis is carried out through multi-strategy steerage prediction and according to a prediction result, so that the method can adapt to different working conditions, and the accuracy and reliability of a fault diagnosis result are remarkably improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of fault diagnosis, in particular to an underwater vehicle fault diagnosis method and device based on a physical data hybrid modeling strategy, equipment and a medium. BACKGROUND

[0002] An autonomous underwater vehicle (AUV) is a robot device capable of performing corresponding tasks underwater and plays a vital role in many fields such as marine scientific research, resource exploration and military reconnaissance. In actual scenarios, the complex and changeable characteristics of the underwater environment bring great challenges to the underwater vehicle, and the navigation mode also affects the task execution process of the underwater vehicle. Under dynamic working condition changes, mechanical failures may occur in the task execution process of the underwater vehicle, which not only affects the execution of the task, but also may cause the underwater vehicle to lose control and cause serious losses. Therefore, fault diagnosis is crucial for the normal navigation and work of the underwater vehicle, and the accuracy of the fault diagnosis method in the related art still needs to be improved. SUMMARY

[0003] The application provides an underwater vehicle fault diagnosis method and device based on a physical data hybrid modeling strategy, equipment and a medium, which effectively improves the adaptability of the prediction results to different working conditions and significantly improves the accuracy and reliability of the fault diagnosis results by performing multi-strategy rudder effect prediction on the underwater vehicle and performing fault diagnosis and emergency decision-making according to the prediction results corresponding to different prediction strategies.

[0004] In order to achieve the above purpose, the main technical scheme adopted by the application includes: In a first aspect, the application provides an underwater vehicle fault diagnosis method based on a physical data hybrid modeling strategy, which adopts at least a rudder effect prediction physical model and a rudder effect prediction neural network; the method comprises: inputting the dynamic parameters of the underwater vehicle into the rudder effect prediction physical model to perform rudder effect prediction on the underwater vehicle and obtain the theoretical driving prediction rudder effect of the underwater vehicle; inputting the dynamic parameters into the rudder effect prediction neural network to perform rudder effect prediction on the underwater vehicle and obtain the data-driven prediction rudder effect of the underwater vehicle; wherein the theoretical driving prediction rudder effect corresponds to a theoretical prediction confidence, and the data-driven prediction rudder effect corresponds to a data prediction confidence; performing fault diagnosis on the underwater vehicle according to the theoretical driving prediction rudder effect, the theoretical prediction confidence, the data-driven prediction rudder effect, the data prediction confidence and the dynamic parameters to obtain the fault diagnosis result of the underwater vehicle; A fault emergency decision is made based on the fault diagnosis result and the operating status parameters of the underwater vehicle, and a fault-tolerant control strategy for handling the corresponding fault is output.

[0005] The present invention proposes a method for underwater vehicle fault diagnosis based on a physical data hybrid modeling strategy. Based on the dynamic parameters of the underwater vehicle, a multi-strategy rudder efficiency prediction method is used to predict the underwater vehicle's rudder efficiency using a rudder efficiency prediction physical model and a rudder efficiency prediction neural network. The method obtains a theoretically driven predicted rudder efficiency corresponding to the rudder efficiency prediction physical model and a data-driven predicted rudder efficiency corresponding to the rudder efficiency prediction neural network. On this basis, the underwater vehicle is diagnosed for faults based on the predicted rudder efficiency corresponding to different prediction strategies and their corresponding prediction confidences. Based on the fault diagnosis results, a fault-tolerant control strategy for handling the corresponding fault is output. Compared with the related art, the present invention utilizes a physical data hybrid modeling strategy including a rudder efficiency prediction physical model and a rudder efficiency prediction neural network to predict the underwater vehicle's rudder efficiency using a multi-strategy rudder efficiency prediction method. The method also performs comprehensive fault diagnosis based on the predicted rudder efficiency and its corresponding confidence. This method enables the fault diagnosis results to adapt to the actual operating conditions of the underwater vehicle, thereby comprehensively considering the impact of the actual operating conditions on the underwater vehicle's behavior and effectively improving the accuracy and reliability of the fault diagnosis results.

[0006] Optionally, the dynamic parameters include the heading speed and rudder angle of the underwater vehicle; and inputting the dynamic parameters of the underwater vehicle into the rudder efficiency prediction physical model to perform rudder efficiency prediction on the underwater vehicle to obtain the theoretical driving predicted rudder efficiency of the underwater vehicle includes: Calculating the torque exerted by each rudder on the underwater vehicle based on the rudder angles of each rudder on the underwater vehicle and the corresponding hydrodynamic coefficients to obtain a static rudder effect of the underwater vehicle; The static steering effect is nonlinearly corrected according to the heading speed to obtain the theoretically driven predicted steering effect.

[0007] Optionally, the dynamic parameters include a navigation speed and a navigation angular velocity of the underwater vehicle; and inputting the dynamic parameters into the steering efficiency prediction neural network to perform steering efficiency prediction on the underwater vehicle to obtain the data-driven predicted steering efficiency of the underwater vehicle includes: performing acceleration calculations at set time intervals based on the navigation speed to obtain navigation acceleration; performing acceleration calculations at set time intervals based on the navigation angular velocity to obtain navigation angular acceleration; Inputting the navigation acceleration and the navigation angular acceleration into the steering effect prediction neural network, and linearly combining the navigation acceleration and the navigation angular acceleration using basis functions corresponding to different weights to obtain a parameter feature map; The steering efficiency is predicted according to the parameter feature mapping to obtain the data-driven predicted steering efficiency.

[0008] Optionally, the performing fault diagnosis on the underwater vehicle based on the theoretically driven predicted steering efficiency, the theoretically predicted confidence, the data driven predicted steering efficiency, the data predicted confidence, and the dynamic parameters to obtain a fault diagnosis result of the underwater vehicle includes: performing confidence correction on the theoretical driven predicted steering effect according to the theoretical prediction confidence to obtain an adaptive theoretical predicted steering effect; performing confidence correction on the data driven predicted steering effect according to the data prediction confidence to obtain an adaptive data predicted steering effect; The rudder effect is predicted according to the adaptive theory, the rudder effect and the dynamic parameters are predicted according to the adaptive data, and the fault diagnosis network is used to perform fault diagnosis on the underwater vehicle to obtain a fault diagnosis result of the underwater vehicle.

[0009] Optionally, the theoretical prediction confidence and the data prediction confidence are obtained by: Inputting the theoretically driven predicted steering efficiency and the data driven predicted steering efficiency into the fault diagnosis network for confidence prediction, thereby obtaining an initial theoretical confidence corresponding to the theoretically driven predicted steering efficiency and an initial data confidence corresponding to the data driven predicted steering efficiency; Adjusting the weight distribution of the theoretically driven predicted steering efficiency according to the initial theoretical confidence to obtain an iterative theoretical predicted steering efficiency; adjusting the weight distribution of the data driven predicted steering efficiency according to the initial data confidence to obtain an iterative data predicted steering efficiency; Inputting the iterative theoretical predicted rudder efficiency and the iterative data predicted rudder efficiency into the fault diagnosis network for confidence prediction to update the initial theoretical confidence and the initial data confidence; repeating the above steps of confidence prediction and weight distribution adjustment until a preset stop condition is reached; The initial theoretical confidence is used as the theoretical prediction confidence, and the initial data confidence is used as the data prediction confidence.

[0010] Optionally, the fault diagnosis network is trained in the following manner: Using an ant colony algorithm to search for initial network parameters of the fault diagnosis network to obtain optimized network parameters of the fault diagnosis network; Based on the training data set, the fault diagnosis network set according to the optimized network parameters is model trained to update the optimized network parameters to obtain target network parameters; the fault diagnosis network is parameterized according to the target network parameters to complete the training.

[0011] Optionally, the fault diagnosis result includes a fault type and a fault severity, and the operating state parameters include a remaining energy state of the underwater vehicle, a navigation mission type, and an environmental disturbance level; and performing a fault emergency decision based on the fault diagnosis result and the operating state parameters of the underwater vehicle and outputting a fault-tolerant control strategy for handling the corresponding fault includes: Performing fuzzy classification on the fault type, the fault degree, the remaining energy state, the navigation mission type, and the environmental disturbance level to obtain fuzzy variables; The fuzzy variables are subjected to fuzzy reasoning based on pre-set fuzzy rules to obtain the fault-tolerant control strategy.

[0012] In a second aspect, an embodiment of the present application provides an underwater vehicle fault diagnosis device based on a physical data hybrid modeling strategy, wherein the physical data hybrid modeling strategy uses at least a rudder efficiency prediction physical model and a rudder efficiency prediction neural network; the device includes: a multi-strategy steering efficiency prediction module, configured to input the dynamic parameters of the underwater vehicle into the steering efficiency prediction physical model to perform steering efficiency prediction on the underwater vehicle, thereby obtaining a theoretical driven predicted steering efficiency of the underwater vehicle; and input the dynamic parameters into the steering efficiency prediction neural network to perform steering efficiency prediction on the underwater vehicle, thereby obtaining a data driven predicted steering efficiency of the underwater vehicle; wherein the theoretical driven predicted steering efficiency corresponds to a theoretical prediction confidence, and the data driven predicted steering efficiency corresponds to a data prediction confidence; a vehicle fault diagnosis module, configured to perform fault diagnosis on the underwater vehicle based on the theoretical drive predicted steering efficiency, the theoretical prediction confidence, the data driven predicted steering efficiency, the data prediction confidence, and the dynamic parameters, and obtain a fault diagnosis result for the underwater vehicle; The vehicle emergency decision module is used to make a fault emergency decision based on the fault diagnosis result and the operating status parameters of the underwater vehicle, and output a fault-tolerant control strategy for handling the corresponding fault.

[0013] In a third aspect, an embodiment of the present application provides a computer device comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, computer instructions are stored in the memory, and the processor executes the method described in any one of the above embodiments by executing the computer instructions.

[0014] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to enable a computer to execute any one of the methods in the above embodiments.

[0015] In a fifth aspect, an embodiment of the present application provides a computer program product, comprising computer instructions for causing a computer to execute the method described in any of the above embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the specific embodiments or prior art in the present application, the drawings needed to be used in the description of the specific embodiments or prior art will be briefly introduced as follows. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0017] Figure 1 A step diagram of the underwater vehicle fault diagnosis method based on the physical data hybrid modeling strategy provided by the embodiment of the present application is shown in FIG. 6. Figure 2 A step diagram of obtaining the theoretically driven predicted rudder effect in the embodiment of the present application is shown in FIG. 7. Figure 3 A step diagram of obtaining the data-driven predicted rudder effect in the embodiment of the present application is shown in FIG. 8. Figure 4 A step diagram of the underwater vehicle fault diagnosis in the embodiment of the present application is shown in FIG. 9. Figure 5 A step diagram of obtaining the theoretical prediction confidence and the data prediction confidence in the embodiment of the present application is shown in FIG. 10. Figure 6 A step diagram of training the fault diagnosis network in the embodiment of the present application is shown in FIG. 11. Figure 7 A flowchart of training the fault diagnosis network in the embodiment of the present application is shown in FIG. 12. Figure 8 A step diagram of outputting the fault-tolerant control strategy in the embodiment of the present application is shown in FIG. 13. Figure 9 A whole framework diagram of the underwater vehicle fault diagnosis method based on the physical data hybrid modeling strategy provided by the embodiment of the present application is shown in FIG. 14. Figure 10 A module diagram of the underwater vehicle fault diagnosis apparatus based on the physical data hybrid modeling strategy provided by the embodiment of the present application is shown in FIG. 15. Figure 11 A structural schematic diagram of a computer device provided by the embodiment of the present application is shown in FIG. 16. DETAILED DESCRIPTION

[0018] To make the purpose, technical solutions, and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of this application.

[0019] Autonomous Underwater Vehicles (AUVs) are robotic devices capable of performing tasks underwater. They play a vital role in a variety of fields, including marine scientific research, resource exploration, and military reconnaissance. In real-world scenarios, the complex and ever-changing nature of the underwater environment presents significant challenges for AUVs, and navigation modes also impact the execution of their missions. Under dynamic operating conditions, AUVs may experience mechanical failures during mission execution, which not only impacts mission execution but can also cause loss of control and lead to serious damage. Therefore, fault diagnosis is crucial to the normal navigation and operation of AUVs.

[0020] In related technologies, methods used to diagnose underwater vehicle faults typically rely on a single model. For example, they calculate the underwater vehicle's operating status based solely on a physical model, or they use a neural network to learn the underwater vehicle's operating mode to predict potential faults. Physical models utilize basic physical principles for calculations and have strong interpretability, but they have limitations when dealing with complex and dynamically changing environmental conditions. Neural networks, on the other hand, can effectively handle fault prediction under complex conditions by learning from large amounts of data. However, the prediction results obtained based on neural networks are less interpretable. Therefore, single-model fault diagnosis methods are limited by the dynamic operating conditions of underwater vehicles and often fail to provide accurate and reliable fault prediction results.

[0021] Based on the above problems, the present application provides an underwater vehicle fault diagnosis method based on a physical data hybrid modeling strategy, which uses at least a rudder efficiency prediction physical model and a rudder efficiency prediction neural network, including: inputting dynamic parameters into the rudder efficiency prediction physical model to predict the rudder efficiency, and obtaining a theoretically driven predicted rudder efficiency; inputting dynamic parameters into the rudder efficiency prediction neural network to predict the rudder efficiency, and obtaining a data-driven predicted rudder efficiency; the theoretically driven predicted rudder efficiency corresponds to a theoretical prediction confidence, and the data-driven predicted rudder efficiency corresponds to a data prediction confidence; fault diagnosis is performed according to the theoretically driven predicted rudder efficiency, the theoretical prediction confidence, the data-driven predicted rudder efficiency, the data prediction confidence and the dynamic parameters to obtain a fault diagnosis result; fault emergency decision-making is made according to the fault diagnosis result and the operating status parameters of the underwater vehicle, and a fault-tolerant control strategy is output.

[0022] The underwater vehicle fault diagnosis method based on the physical data hybrid modeling strategy provided in the application is based on the dynamic parameters of the underwater vehicle, and the rudder effect prediction physical model and the rudder effect prediction neural network are used to perform multi-strategy rudder effect prediction on the underwater vehicle respectively, to obtain a theoretical driving prediction rudder effect corresponding to the rudder effect prediction physical model and a data-driven prediction rudder effect corresponding to the rudder effect prediction neural network. On this basis, the underwater vehicle is diagnosed according to the prediction rudder effects corresponding to different prediction strategies and the prediction confidence corresponding to the prediction rudder effects, and a fault-tolerant control strategy for processing the corresponding fault is output based on the fault diagnosis result.

[0023] Compared with the related art, the application uses the physical data hybrid modeling strategy including the rudder effect prediction physical model and the rudder effect prediction neural network to perform multi-strategy rudder effect prediction on the underwater vehicle, and performs comprehensive fault diagnosis according to the prediction rudder effects and the confidence corresponding to the prediction rudder effects, so that the fault diagnosis result can be self-adaptive to the actual working condition of the underwater vehicle, thereby comprehensively considering the influence of the actual working condition on the behavior of the underwater vehicle, and effectively improving the accuracy and reliability of the fault diagnosis result.

[0024] The underwater vehicle fault diagnosis method based on the physical data hybrid modeling strategy provided in the specification can be applied to the fault diagnosis of the underwater vehicle. The underwater vehicle can include an autonomous underwater vehicle (XAUV), a remote-controlled underwater vehicle, a hybrid underwater vehicle, or an underwater robot, and the physical data hybrid modeling strategy at least uses the rudder effect prediction physical model and the rudder effect prediction neural network. It can be understood that after adaptive modification, the method can also be used for fault monitoring of working equipment other than the underwater vehicle, or the rudder effect prediction is performed by using a strategy other than the physical model and the neural network.

[0025] According to the embodiment of the application, an underwater vehicle fault diagnosis method based on a physical data hybrid modeling strategy is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0026] In this embodiment, an underwater vehicle fault diagnosis method based on a physical data hybrid modeling strategy is provided, which can be used for fault diagnosis of the underwater vehicle described above. As shown in the Figure 1 physical data hybrid modeling strategy at least uses the rudder effect prediction physical model and the rudder effect prediction neural network; the method comprises: S100. input the dynamic parameters of the underwater vehicle into a rudder effect prediction physical model to perform rudder effect prediction on the underwater vehicle, to obtain a theoretical driving prediction rudder effect of the underwater vehicle; input the dynamic parameters into a rudder effect prediction neural network to perform rudder effect prediction on the underwater vehicle, to obtain a data-driven prediction rudder effect of the underwater vehicle; wherein the theoretical driving prediction rudder effect corresponds to a theoretical prediction confidence, and the data-driven prediction rudder effect corresponds to a data prediction confidence.

[0027] S200. perform fault diagnosis on the underwater vehicle according to the theoretical driving prediction rudder effect, the theoretical prediction confidence, the data-driven prediction rudder effect, the data prediction confidence, and the dynamic parameters, to obtain a fault diagnosis result of the underwater vehicle.

[0028] S300. perform fault emergency decision-making according to the fault diagnosis result and the operating state parameters of the underwater vehicle, and output a fault-tolerant control strategy for handling the corresponding fault.

[0029] The rudder effect prediction physical model can be used to deduce and simulate the actual physical behavior of the underwater vehicle in the working process according to the dynamic equation of the underwater vehicle and related physical principles, thereby obtaining the theoretical driving prediction rudder effect of the underwater vehicle. The physical principles used by the rudder effect prediction physical model include but are not limited to rudder surface mechanics model, hydrodynamics model, and physical transmission equation, etc. It can be understood that the rudder effect prediction physical model is based on known physical laws for simulation, and can accurately characterize the working process of the underwater vehicle in a stable environment, and has high explainability.

[0030] The rudder effect prediction neural network can learn the complex nonlinear relationship between the working condition data of the underwater vehicle in the working process by learning a large amount of working condition data, to predict the rudder effect change trend of the underwater vehicle, and obtain the data-driven prediction rudder effect of the underwater vehicle. Illustratively, the type of neural network used by the rudder effect prediction neural network can be a radial basis function neural network, a multilayer perceptron, a support vector machine, or a K-nearest neighbor regression, etc. It can be understood that the rudder effect prediction neural network can automatically model based on the working condition data of the underwater vehicle, and through an adaptive mechanism to adjust in real time according to the change of the working condition, to realize flexible prediction of the rudder effect of the underwater vehicle.

[0031] Specifically, the dynamic parameters generated by the underwater vehicle during operation are collected by sensors installed in the underwater vehicle. After obtaining the dynamic parameters, a physical data hybrid modeling strategy is used to establish multiple prediction models for multi-strategy steering efficiency prediction of the underwater vehicle. It should be noted that the physical data hybrid modeling strategy includes at least a steering efficiency prediction physical model and a steering efficiency prediction neural network. The steering efficiency prediction physical model predicts the steering efficiency of the underwater vehicle based on physical principles, and the steering efficiency prediction neural network predicts the steering efficiency of the underwater vehicle based on data relationships through a neural network, thereby obtaining predicted steering efficiencies from different angles, effectively improving the comprehensiveness of the prediction results. In some embodiments, the number of steering efficiency prediction physical models or steering efficiency prediction neural networks can be one or more.

[0032] Furthermore, the physical model for predicting steering efficiency is simulated based on known physical laws, and the resulting theoretically driven predicted steering efficiency is more accurate in a stable environment. The steering efficiency prediction neural network predicts steering efficiency by learning the complex relationships between data. The resulting data-driven predicted steering efficiency can be adjusted in real time according to changes in working conditions, making it more suitable for complex and changing environments. Since different prediction strategies are based on different principles and use different working conditions, after obtaining the multi-strategy steering efficiency prediction results for underwater vehicles, the prediction confidence of different prediction results under the current working conditions of the underwater vehicle is analyzed to obtain the theoretical prediction confidence corresponding to the theoretically driven predicted steering efficiency and the data prediction confidence corresponding to the data-driven predicted steering efficiency. These prediction confidences are used to fuse the multi-strategy steering efficiency prediction results, so that the steering efficiency prediction results based on different prediction strategies can complement each other's strengths, significantly improve the accuracy of the steering efficiency prediction results, and thus improve the accuracy and reliability of the fault diagnosis results.

[0033] Furthermore, after obtaining the theoretically driven predicted steering efficiency, the theoretically predicted confidence, the data-driven predicted steering efficiency, and the data-driven predicted confidence, the theoretically driven predicted steering efficiency and the data-driven predicted steering efficiency are fused based on the theoretical prediction confidence and the data prediction confidence. This allows for fault diagnosis of the underwater vehicle based on the fusion results and the underwater vehicle's dynamic parameters, yielding a fault diagnosis result for the underwater vehicle. Exemplarily, the fault diagnosis result may include fault types, including but not limited to stuck rudder faults, multi-faceted damage faults, and center-biased faults, with each fault type corresponding to a specific fault severity. Based on this, a fault emergency decision is made based on the obtained fault diagnosis results and the underwater vehicle's operating status parameters, and a fault-tolerant control strategy is output to address the corresponding fault type, ensuring the normal operation and navigation of the underwater vehicle. It is understood that the underwater vehicle's operating status parameters may be directly obtained from sensors within the underwater vehicle to represent the underwater vehicle's current state.

[0034] The underwater vehicle fault diagnosis method based on the physical data hybrid modeling strategy provided in this embodiment performs multi-strategy rudder efficiency prediction on the underwater vehicle based on the dynamic parameters of the underwater vehicle through the rudder efficiency prediction physical model and the rudder efficiency prediction neural network, and obtains the theoretically driven predicted rudder efficiency corresponding to the rudder efficiency prediction physical model and the data-driven predicted rudder efficiency corresponding to the rudder efficiency prediction neural network; on this basis, the underwater vehicle is diagnosed according to the predicted rudder efficiency corresponding to different prediction strategies and their corresponding prediction confidence levels, and a fault-tolerant control strategy for handling the corresponding fault is output based on the fault diagnosis results.

[0035] Compared with related technologies, the present application utilizes a physical data hybrid modeling strategy including a rudder efficiency prediction physical model and a rudder efficiency prediction neural network to perform multi-strategy rudder efficiency prediction on underwater vehicles, and performs comprehensive fault diagnosis based on each predicted rudder efficiency and its corresponding confidence level, so that the fault diagnosis results can adapt to the actual working conditions of the underwater vehicle, thereby comprehensively considering the impact of the actual working conditions on the behavior of the underwater vehicle, and effectively improving the accuracy and reliability of the fault diagnosis results.

[0036] Reference Figure 2 As shown in FIG. 1 , as an embodiment of the present application, the dynamic parameters include the heading speed and rudder angle of the underwater vehicle; the dynamic parameters of the underwater vehicle are input into the rudder efficiency prediction physical model to predict the rudder efficiency of the underwater vehicle, thereby obtaining the theoretical driving predicted rudder efficiency of the underwater vehicle, including: S110. Calculate the torque exerted by each rudder on the underwater vehicle based on the rudder angles of the rudders and their corresponding hydrodynamic coefficients to obtain a static rudder effect of the underwater vehicle.

[0037] S120. Perform nonlinear correction on the static steering effect according to the heading speed to obtain the theoretically driven predicted steering effect.

[0038] The heading speed of an underwater vehicle can be the forward speed of the underwater vehicle relative to the underwater environment. The rudder angle of an underwater vehicle can be the relative angle of any rudder surface on the underwater vehicle, that is, the angle between the rudder surface and the longitudinal axis of the underwater vehicle. An underwater vehicle can be equipped with different types of rudders, including but not limited to stern rudders and bow rudders. Stern rudders can be X-shaped stern rudders, and one or more rudders of each type can be installed. Different types of rudders play different roles during the navigation of the underwater vehicle and generate torque on the underwater vehicle in different ways. Therefore, each type of rudder has its own corresponding hydrodynamic coefficient. The heading speed and rudder angle determine the motion trajectory and attitude of the underwater vehicle, and together affect the controllability and stability of the underwater vehicle.

[0039] Specifically, any rudder on an underwater vehicle interacts with the water flow to apply torque to the vehicle. The torque applied by each rudder on the underwater vehicle is calculated based on the rudder angle and its corresponding hydrodynamic coefficient to obtain the static rudder efficiency of the underwater vehicle.

[0040] Furthermore, when an underwater vehicle is traveling at high speed, the nonlinear effect of water flow on the vehicle becomes more significant, affecting the relationship between rudder efficiency and heading speed. Therefore, after obtaining the static rudder efficiency of the underwater vehicle, it is necessary to perform a nonlinear correction based on the heading speed of the underwater vehicle to predict the rudder efficiency at the current heading speed and obtain the theoretically driven predicted rudder efficiency.

[0041] For example, the theoretical predicted driving rudder efficiency of an underwater vehicle includes the theoretical predicted pitching rudder efficiency and the theoretical predicted turning rudder efficiency. Accordingly, the hydrodynamic coefficient includes the pitching dynamic coefficient and the turning dynamic coefficient. When the underwater vehicle is equipped with four sets of X-shaped stern rudders and bow rudders, the theoretical predicted pitching rudder efficiency and the theoretical predicted turning rudder efficiency of the underwater vehicle can be expressed by the following formula: in, It represents the theoretical prediction of pitch rudder effect; It represents the theoretical prediction of steering efficiency; 、 、 、 and are the pitching dynamic coefficients of each rudder on the underwater vehicle; 、 、 and are the steering dynamic coefficients of each rudder on the underwater vehicle; is the heading speed of the underwater vehicle; 、 、 and are the rudder angles of the stern rudder, is the rudder angle of the bow rudder.

[0042] Reference Figure 3 As shown in FIG. 1 , as an embodiment of the present application, the dynamic parameters include the navigation speed and navigation angular velocity of the underwater vehicle; the dynamic parameters are input into the steering efficiency prediction neural network to predict the steering efficiency of the underwater vehicle, and the data-driven predicted steering efficiency of the underwater vehicle is obtained, including: S130. Perform acceleration calculation at set time intervals based on the navigation speed to obtain the navigation acceleration; perform acceleration calculation at set time intervals based on the navigation angular velocity to obtain the navigation angular acceleration.

[0043] S140. Input the navigation acceleration and the navigation angular acceleration into the rudder effect prediction neural network, and use the basis functions corresponding to different weights to linearly combine the navigation acceleration and the navigation angular acceleration to obtain a parameter feature map.

[0044] S150. Predict the steering efficiency based on the parameter feature mapping to obtain a data-driven predicted steering efficiency.

[0045] The navigation speed of the underwater vehicle includes three degrees of freedom, and the speeds in different directions are independent of each other. The navigation angular velocity of the underwater vehicle includes three degrees of freedom, and the rotation speeds in different directions are independent of each other.

[0046] Specifically, the six-degree-of-freedom acceleration of the underwater vehicle is calculated as the input of the steering efficiency prediction neural network. The six-degree-of-freedom acceleration of the underwater vehicle is expressed by the following formula: in, 、 and are the navigation accelerations of the underwater vehicle in different directions; 、 and are the navigation speeds of the underwater vehicle in different directions; 、 and are the navigation angular accelerations of the underwater vehicle in different directions; 、 and are the angular velocities of the underwater vehicle in different directions; The current moment of the rudder effectiveness prediction; To set the time interval, usually 0.1s is used.

[0047] Furthermore, the steering effect prediction neural network can be a pre-built and trained neural network. For example, the steering effect prediction neural network can be a radial basis function (RBF) neural network, which constructs a three-layer neural network consisting of an input layer, a hidden layer, and an output layer, and uses a radial basis function adaptive algorithm in the steering effect prediction neural network to perform real-time adjustments to gradually approach the nonlinear function as the prediction result. The steering effect prediction neural network receives the six-degree-of-freedom acceleration of the underwater vehicle as input, and uses basis functions corresponding to different weights in the hidden layer to linearly combine the navigation acceleration and navigation angular acceleration to obtain a parameter feature map, which is expressed by the following formula: in, represents parameter feature map; It is the input of the rudder effect prediction neural network; is the weight matrix; is the Gaussian basis function; is the approximate error.

[0048] Furthermore, the output layer receives the parameter feature map and adjusts the weight of each Gaussian basis function in the parameter feature map to predict the steering efficiency of the underwater vehicle based on the parameter feature map, thereby obtaining the data-driven predicted steering efficiency. The data-driven predicted steering efficiency includes the data-predicted pitch steering efficiency and the data-predicted turning steering efficiency, where the data-predicted pitch steering efficiency is expressed as , the data predicted steering effect is expressed as .

[0049] Reference Figure 4 As shown, as an embodiment of the present application, fault diagnosis of an underwater vehicle is performed based on the theoretically driven predicted steering efficiency, the theoretically predicted confidence, the data driven predicted steering efficiency, the data predicted confidence, and the dynamic parameters, and the fault diagnosis result of the underwater vehicle is obtained, including: S210. Perform confidence correction on the theoretically driven predicted steering efficiency according to the theoretical prediction confidence to obtain an adaptive theoretical predicted steering efficiency; perform confidence correction on the data driven predicted steering efficiency according to the data prediction confidence to obtain an adaptive data predicted steering efficiency.

[0050] S220. Predict the rudder effect based on the adaptive theory, predict the rudder effect and dynamic parameters based on the adaptive data, and use the fault diagnosis network to diagnose the fault of the underwater vehicle to obtain the fault diagnosis result of the underwater vehicle.

[0051] Specifically, the principles behind each prediction strategy in the physical data hybrid modeling strategy are different, and different prediction strategies are suitable for different operating conditions of underwater vehicles. Among them, the rudder efficiency prediction physical model is simulated based on known physical laws, and the resulting theoretically driven prediction of rudder efficiency is more accurate in a stable environment. The rudder efficiency prediction neural network predicts rudder efficiency by learning the complex relationships between data. The resulting data-driven prediction of rudder efficiency can be adjusted in real time according to changes in operating conditions, making it more suitable for complex and changing environments. Therefore, when performing fault diagnosis on an underwater vehicle based on the multi-strategy rudder efficiency prediction results, it is necessary to fuse the multi-strategy rudder efficiency prediction results so that the fault diagnosis results can adapt to the actual operating conditions of the underwater vehicle, thereby comprehensively considering the impact of the actual operating conditions on the behavior of the underwater vehicle and effectively improving the accuracy and reliability of the fault diagnosis results.

[0052] Further, the theoretical prediction confidence and the data prediction confidence are used to correct the confidence of the respective corresponding rudder effect prediction results, to fuse the multi-strategy rudder effect prediction results, so that the rudder effect prediction results based on different prediction strategies can complement each other, significantly improve the accuracy of the rudder effect prediction results, and further improve the accuracy and reliability of the fault diagnosis results. It can be understood that the adaptive theoretical rudder effect includes adaptive theoretical pitch rudder effect and adaptive theoretical turning rudder effect, the adaptive data rudder effect includes adaptive data pitch rudder effect and adaptive data turning rudder effect, and the form of confidence correction is represented by the following formula: wherein, represents the adaptive data pitch rudder effect, represents the adaptive data turning rudder effect; represents the adaptive theoretical pitch rudder effect, represents the adaptive theoretical turning rudder effect; is the data prediction confidence, is the theoretical prediction confidence.

[0053] Further, after confidence correction, the adaptive theoretical pitch rudder effect, the adaptive theoretical turning rudder effect, the adaptive data pitch rudder effect and the adaptive data turning rudder effect, and the dynamics parameters of the underwater vehicle are input into the fault diagnosis network to diagnose the fault of the underwater vehicle, and the fault diagnosis result of the underwater vehicle is obtained.

[0054] It should be noted that the fault diagnosis network can be used for feature learning and data classification of input data to accurately identify various fault types of the rudder system of the underwater vehicle. Illustratively, the type of fault diagnosis network can be a back propagation neural network, and the overall structure is a three-layer feedforward network, including an input layer, a hidden layer and an output layer. The input data received by the input layer includes the adaptive data pitch rudder effect, the adaptive theoretical turning rudder effect, the adaptive data pitch rudder effect, the adaptive data turning rudder effect and various dynamics parameters of the underwater vehicle, the number of neurons in the input layer is the same as the number of input data, and they correspond to each other.

[0055] The output layer is used to output the fault diagnosis results. One-hot encoding can be used to represent three types of faults, and the specific fault severity under each fault type can be further distinguished by the fault category sub-label. In some embodiments, the fault types may include rudder stuck fault, multi-faceted damage fault, and center fault, among which the rudder stuck fault corresponds to 7 specific fault severity levels, including corresponding rudder stuck angles of -30°, -20°, -10°, 0°, 10°, 20°, and 30°. Multi-faceted damage fault corresponds to 3 specific fault severity levels, including corresponding damage ratios of 30%, 50%, and 80%. Center fault corresponds to 2 specific fault severity levels, including corresponding offsets of -10 and 10. The number of neurons in the output layer is determined according to the fault type and the number of specific fault severity levels corresponding to it. For example, when the total number of specific fault severity levels is 12, the number of neurons in the output layer can be 60.

[0056] The hidden layer is used for feature learning, and the number of neurons can be determined according to the empirical formula, which is expressed by the following formula: in, Indicates the number of neurons in the hidden layer; is the number of neurons in the input layer; is the number of neurons in the output layer; is a regulation factor, and its value range is between [1, 10]. In some embodiments, when the number of neurons in the input layer is 15 and the number of neurons in the output layer is 60, comprehensive experiments can determine that the optimal value of the regulation factor is 35.

[0057] Reference Figure 5 As shown, as an embodiment of the present application, the theoretical prediction confidence and the data prediction confidence are obtained in the following manner: S160. Input the theoretically driven predicted steering efficiency and the data driven predicted steering efficiency into the fault diagnosis network for confidence prediction, and obtain the initial theoretical confidence corresponding to the theoretically driven predicted steering efficiency and the initial data confidence corresponding to the data driven predicted steering efficiency.

[0058] S170. Adjust the weight distribution of the theoretically driven predicted steering efficiency according to the initial theoretical confidence to obtain the iterative theoretical predicted steering efficiency; adjust the weight distribution of the data driven predicted steering efficiency according to the initial data confidence to obtain the iterative data predicted steering efficiency.

[0059] S180. Input the iterative theoretical prediction of the rudder efficiency and the iterative data prediction of the rudder efficiency into the fault diagnosis network for confidence prediction to update the initial theoretical confidence and the initial data confidence; repeat the above steps of confidence prediction and weight distribution adjustment until the preset stop condition is reached.

[0060] S190. Use the initial theoretical confidence as the theoretical prediction confidence, and use the initial data confidence as the data prediction confidence.

[0061] Specifically, while outputting the fault diagnosis results, the fault diagnosis network also outputs prediction confidence levels corresponding to different steering efficiency prediction results to indicate the degree of adaptability of different prediction strategies to the underwater vehicle's current operating conditions. Therefore, after obtaining the multi-strategy steering efficiency prediction results for the underwater vehicle, the fault diagnosis network is used to perform confidence predictions on the different prediction results based on the underwater vehicle's current operating conditions, resulting in initial theoretical confidence levels corresponding to the theoretically driven predicted steering efficiency and initial data confidence levels corresponding to the data-driven predicted steering efficiency. It is understood that these initial theoretical confidence levels and initial data confidence levels are obtained through preliminary learning by the fault diagnosis network based on the theoretically driven predicted steering efficiency and the data-driven predicted steering efficiency. There is still an error between these initial theoretical confidence levels and the actual confidence levels of each predicted steering efficiency, and they cannot fully represent the degree of adaptability of the different prediction strategies to the current operating conditions.

[0062] Furthermore, the weight distribution of the theoretically driven predicted steering efficiency is adjusted according to the initial theoretical confidence to obtain the iterative theoretical predicted steering efficiency. The weight distribution of the data driven predicted steering efficiency is adjusted according to the initial data confidence to obtain the iterative data predicted steering efficiency. The iterative theoretical predicted steering efficiency and the iterative data predicted steering efficiency are input into the fault diagnosis network for confidence prediction to update the initial theoretical confidence and the initial data confidence. It should be noted that in the above process, the fault diagnosis network further learns different steering efficiency prediction results based on the initial theoretical confidence and the initial data confidence, so that the adaptability of the steering efficiency prediction results to the working conditions can be adjusted according to the initial confidence of the previous iterative round, thereby improving the accuracy of the adjusted initial confidence.

[0063] Furthermore, the above steps of confidence prediction and weight distribution adjustment are repeated until a preset stopping condition is reached, and the initial theoretical confidence of the current iteration round is used as the theoretical prediction confidence, and the initial data confidence of the current iteration round is used as the data prediction confidence. It is understood that the preset stopping condition includes but is not limited to an iteration number condition, a confidence convergence condition, and a confidence threshold condition, wherein the iteration number condition can be that the number of repeated iterations of the above steps reaches the predicted iteration number, the confidence convergence condition can be that the rate of change of the initial confidence within a certain number of rounds does not exceed a preset rate of change threshold, and the confidence threshold condition can be that the value of the initial confidence is less than a preset value threshold.

[0064] Reference Figure 6 As shown, as an embodiment of the present application, the fault diagnosis network is trained in the following manner: S230. Use the ant colony algorithm to search for the initial network parameters of the fault diagnosis network to obtain the optimized network parameters of the fault diagnosis network.

[0065] S240. Based on the training data set, the fault diagnosis network set according to the optimized network parameters is trained to update the optimized network parameters and obtain target network parameters; the fault diagnosis network is parameterized according to the target network parameters to complete the training.

[0066] Specifically, the Ant Colony Optimization (ACO) is a heuristic random search algorithm that solves combinatorial optimization problems by simulating the path-finding process of ants in nature. In this embodiment, the ACO is used to search for the initial network parameters of the fault diagnosis network to optimize the initial network parameters, which include the weights and thresholds of the fault diagnosis network. Figure 7 As shown in the figure, during the search process, the ant colony and the pheromone distribution in the search space are first initialized to construct the corresponding search space. Next, the ants are allowed to move in the current iteration, and the pheromone distribution in the search space is updated based on their movement. Next, the colony's state transition probability, pheromone concentration, and fitness for the current iteration are calculated based on the ants' movement. If the fitness meets the preset fitness condition, the search results are output as the optimized network parameters of the fault diagnosis network. Otherwise, the process continues to the next iteration.

[0067] In some embodiments, in order to further improve the search efficiency of the ant colony algorithm, the path selection probability of the ants can be adjusted to affect the efficiency and accuracy of the algorithm. The path selection probability can be adjusted dynamically and adaptively, and its form is expressed as follows: in, and represents the parameters that control the weights of pheromones and heuristic factors; and is the parameter of the previous iteration round; is the current iteration step size, is the maximum iteration step size. By adjusting the path selection probability through the above formula, ants can explore the solution space more flexibly during the search process, effectively improving the search efficiency of network parameters.

[0068] Furthermore, based on the training data set, the fault diagnosis network set according to the optimized network parameters is trained, the errors in the model training process and their corresponding weights are used to update the threshold of the fault diagnosis network, and the model training is completed when the training stop conditions are met to obtain a trained fault diagnosis network.

[0069] It should be noted that the ant colony algorithm is used to search the initial network parameters of the fault diagnosis network, which improves the convergence speed of the initial network parameters, so that the optimized network parameters can be determined more quickly. The optimized network parameters can effectively improve the generalization ability and robustness of the fault diagnosis network in the fault diagnosis process, thereby improving the accuracy and reliability of the fault diagnosis results.

[0070] Reference Figure 8 As shown in FIG. 1 , as an embodiment of the present application, the fault diagnosis result includes the fault type and fault severity, and the operating status parameters include the remaining energy state of the underwater vehicle, the navigation mission type, and the environmental disturbance level. A fault emergency decision is made based on the fault diagnosis result and the operating status parameters of the underwater vehicle, and a fault-tolerant control strategy for handling the corresponding fault is output, including: S310. Perform fuzzy classification on the fault type, fault degree, remaining energy state, navigation mission type and environmental disturbance level to obtain fuzzy variables.

[0071] S320. Perform fuzzy reasoning on fuzzy variables based on pre-set fuzzy rules to obtain a fault-tolerant control strategy.

[0072] Reference Figure 9 As shown in the overall framework of this application, the decision layer receives the fault diagnosis results output by the diagnosis layer and uses the fuzzy decision unit to generate a corresponding fault-tolerant control strategy based on the operating state parameters of the underwater vehicle to deal with emergency situations during the working process. The fault diagnosis results output by the diagnosis layer can include the fault type and fault degree, where the fault confidence is expressed as , the fault type is expressed as , the fault degree is expressed as The operating state parameters may include the remaining energy state of the underwater vehicle, the navigation mission type and the environmental disturbance level, where the remaining energy state is expressed as , the navigation mission type is expressed as , the environmental disturbance level is expressed as The fault-tolerant control strategy is expressed as .

[0073] Specifically, a fuzzy classification is performed for fault confidence, fault type, fault severity, remaining energy state, navigation mission type, and environmental disturbance level. The domain of fault confidence is [0, 1]. Fault types, which include stuck rudder faults, multi-surface damage faults, center faults, and no faults, have a domain of [0, 3]. Fault severity includes high-risk and low-risk faults, and has a domain of [0, 1]. Stuck rudder faults with a rudder angle of 30° or -30° and multi-surface damage faults with a damage ratio of 80% are both high-risk faults, while other fault types are low-risk. The domain of remaining energy state is [0, 1]. Navigation mission types include search phase and operation phase, and have a domain of [0, 1]. The domain of environmental disturbance level is [0, 4].

[0074] The above fuzzy inputs are described in linguistic form, where the fault confidence is divided into {S, L}, representing low fault confidence and high fault confidence, respectively. The fault type is divided into {NB, NS, PS, PB}, representing no fault, damage fault, center fault, and rudder stuck fault, respectively. The fault severity is divided into {S, L}, representing low dangerous fault and high dangerous fault, respectively. The remaining energy state is divided into {S, L}, representing low remaining energy and high remaining energy, respectively. The navigation mission type is divided into {S, L}, representing the search phase and the operation phase, respectively. The environmental disturbance level is divided into {S, L}, representing low environmental disturbance level and high environmental disturbance level, respectively.

[0075] Similarly, the fault-tolerant control strategies corresponding to the fuzzy variables include continued navigation, optimized allocation, fault reconstruction, control remapping, and emergency floating, with a domain of [0, 4]. The output of the fuzzy decision-making unit is described in linguistic form, and the fault-tolerant control strategies are divided into {NB, NS, Z, PS, PB}, representing continued navigation, optimized allocation, control remapping, fault reconstruction, and emergency floating, respectively.

[0076] Furthermore, the fuzzy rules of the fuzzy decision unit can be shown in Table 1. After determining the fault-tolerant control strategy, the centroid method is used for defuzzification to output the control parameters to control the underwater vehicle. It should be noted that in the fuzzy weight allocator, the Mamdani inference algorithm is used as the fuzzy inference algorithm.

[0077] Table 1 Fuzzy rules table of fuzzy decision making unit Accordingly, please refer to Figure 10The present invention provides an underwater vehicle fault diagnosis device based on a physical data hybrid modeling strategy, wherein the physical data hybrid modeling strategy adopts at least a rudder efficiency prediction physical model and a rudder efficiency prediction neural network; the device includes: The multi-strategy steering efficiency prediction module 1010 is used to input the dynamic parameters of the underwater vehicle into the steering efficiency prediction physical model to predict the steering efficiency of the underwater vehicle and obtain the theoretical driven predicted steering efficiency of the underwater vehicle; input the dynamic parameters into the steering efficiency prediction neural network to predict the steering efficiency of the underwater vehicle and obtain the data driven predicted steering efficiency of the underwater vehicle; wherein, the theoretical driven predicted steering efficiency corresponds to a theoretical prediction confidence, and the data driven predicted steering efficiency corresponds to a data prediction confidence.

[0078] The vehicle fault diagnosis module 1020 is used to diagnose the fault of the underwater vehicle based on the theoretical drive predicted steering effect, the theoretical prediction confidence, the data driven predicted steering effect, the data prediction confidence and the dynamic parameters, and obtain the fault diagnosis result of the underwater vehicle.

[0079] The vehicle emergency decision module 1030 is used to make a fault emergency decision based on the fault diagnosis result and the operating status parameters of the underwater vehicle, and output a fault-tolerant control strategy for handling the corresponding fault.

[0080] In some optional implementations, the multi-strategy steering efficiency prediction module 1010 includes: The static rudder effect calculation unit is used to calculate the torque of each rudder on the underwater vehicle according to the rudder angle of each rudder on the underwater vehicle and its corresponding hydrodynamic coefficient, so as to obtain the static rudder effect of the underwater vehicle.

[0081] The heading speed correction unit is used to perform nonlinear correction on the static steering effect according to the heading speed to obtain the theoretical drive predicted steering effect.

[0082] In some optional implementations, the multi-strategy steering efficiency prediction module 1010 includes: The acceleration unit is used to calculate the acceleration at set time intervals according to the navigation speed to obtain the navigation acceleration; and to calculate the acceleration at set time intervals according to the navigation angular velocity to obtain the navigation angular acceleration.

[0083] The parameter mapping unit is used to input the navigation acceleration and the navigation angular acceleration into the rudder effect prediction neural network, and use the basis functions corresponding to different weights to linearly combine the navigation acceleration and the navigation angular acceleration to obtain the parameter feature mapping.

[0084] The steering efficiency prediction unit is used to predict the steering efficiency based on parameter feature mapping to obtain data-driven predicted steering efficiency.

[0085] In some optional implementations, the aircraft fault diagnosis module 1020 includes: a confidence correction unit configured to perform confidence correction on the theoretical driving predicted rudder effect according to a theoretical prediction confidence to obtain an adaptive theoretical predicted rudder effect, and perform confidence correction on the data driving predicted rudder effect according to a data prediction confidence to obtain an adaptive data predicted rudder effect.

[0086] a fault diagnosis unit configured to perform fault diagnosis on the underwater vehicle by using a fault diagnosis network according to the adaptive theoretical predicted rudder effect, the adaptive data predicted rudder effect, and the kinetic parameters, to obtain a fault diagnosis result of the underwater vehicle.

[0087] In some optional embodiments, the multi-strategy rudder effect prediction module 1010 comprises: an initial confidence acquisition unit configured to input the theoretical driving predicted rudder effect and the data driving predicted rudder effect into the fault diagnosis network for confidence prediction, to obtain an initial theoretical confidence corresponding to the theoretical driving predicted rudder effect and an initial data confidence corresponding to the data driving predicted rudder effect.

[0088] a confidence iteration unit configured to perform weight distribution adjustment on the theoretical driving predicted rudder effect according to the initial theoretical confidence to obtain an iterative theoretical predicted rudder effect, and perform weight distribution adjustment on the data driving predicted rudder effect according to the initial data confidence to obtain an iterative data predicted rudder effect.

[0089] a repeated iteration unit configured to input the iterative theoretical predicted rudder effect and the iterative data predicted rudder effect into the fault diagnosis network for confidence prediction, to update the initial theoretical confidence and the initial data confidence, and repeat the above steps of confidence prediction and weight distribution adjustment until a preset stopping condition is reached.

[0090] a prediction confidence acquisition unit configured to take the initial theoretical confidence as the theoretical prediction confidence and take the initial data confidence as the data prediction confidence.

[0091] In some optional embodiments, the vehicle fault diagnosis module 1020 comprises: a parameter optimization unit configured to search for initial network parameters of the fault diagnosis network by using an ant colony algorithm, to obtain optimized network parameters of the fault diagnosis network.

[0092] a model training unit configured to perform model training on the fault diagnosis network set according to the optimized network parameters based on a training data set, to update the optimized network parameters, to obtain target network parameters, and perform parameter setting on the fault diagnosis network according to the target network parameters, to complete training.

[0093] In some optional embodiments, the vehicle emergency decision module 1030 comprises: The fuzzy division unit is used to perform fuzzy division on the fault type, fault degree, remaining energy state, navigation mission type and environmental disturbance level to obtain fuzzy variables.

[0094] The fuzzy decision-making unit is used to perform fuzzy reasoning on fuzzy variables based on pre-set fuzzy rules to obtain a fault-tolerant control strategy.

[0095] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0096] The underwater vehicle fault diagnosis device based on the physical data hybrid modeling strategy in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0097] See also Figure 11 , Figure 11 1 is a structural diagram of a computer device provided by an embodiment of the present application. As shown in the figure, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are connected to each other using different buses for communication and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the computer device, including instructions stored in or on the memory to display graphical information of a GUI on an external input / output device (such as a display device coupled to an interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 11 A processor 10 is taken as an example.

[0098] The processor 10 may be a central processing unit, a network processor, or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic, or any combination thereof.

[0099] The memory 20 stores instructions that can be executed by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0100] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created based on the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0101] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0102] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0103] The embodiments of the present application also provide a computer-readable storage medium. The above-mentioned method according to the embodiment of the present application can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0104] An embodiment of the present application provides a computer program product, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform a method according to any embodiment of the present application.

[0105] Although the embodiments of the present application have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present application, and such modifications and variations shall fall within the scope defined by the appended claims.

[0106] The systems, devices, modules, or units described in the above embodiments may be implemented by computer chips or entities, or by products having certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0107] For the convenience of description, the above devices are described as being divided into various units according to their functions. Of course, when implementing this application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0108] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0109] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0110] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1The function specified in one or more boxes.

[0111] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0112] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0113] The various embodiments in this specification are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the description of the method embodiments.

[0114] The foregoing is merely an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

[0115] Although the embodiments of the present application have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present application, and such modifications and variations shall fall within the scope defined by the appended claims.

Claims

1. A method for underwater vehicle fault diagnosis based on physical data hybrid modeling strategy, characterized in that: The physical data hybrid modeling strategy adopts at least a steering efficiency prediction physical model and a steering efficiency prediction neural network; the method includes: Inputting the dynamic parameters of the underwater vehicle into the steering efficiency prediction physical model to predict the steering efficiency of the underwater vehicle and obtain a theoretical driven predicted steering efficiency of the underwater vehicle; inputting the dynamic parameters into the steering efficiency prediction neural network to predict the steering efficiency of the underwater vehicle and obtain a data driven predicted steering efficiency of the underwater vehicle; wherein the theoretical driven predicted steering efficiency corresponds to a theoretical prediction confidence, and the data driven predicted steering efficiency corresponds to a data prediction confidence; performing fault diagnosis on the underwater vehicle based on the theoretically driven predicted steering efficiency, the theoretically predicted confidence, the data driven predicted steering efficiency, the data predicted confidence, and the dynamic parameters to obtain a fault diagnosis result of the underwater vehicle; A fault emergency decision is made based on the fault diagnosis result and the operating status parameters of the underwater vehicle, and a fault-tolerant control strategy for handling the corresponding fault is output.

2. The method according to claim 1, characterized in that The dynamic parameters include the heading speed and rudder angle of the underwater vehicle; inputting the dynamic parameters of the underwater vehicle into the rudder efficiency prediction physical model to predict the rudder efficiency of the underwater vehicle and obtain the theoretical driving predicted rudder efficiency of the underwater vehicle includes: Calculating the torque exerted by each rudder on the underwater vehicle based on the rudder angles of each rudder on the underwater vehicle and the corresponding hydrodynamic coefficients to obtain a static rudder effect of the underwater vehicle; The static steering effect is nonlinearly corrected according to the heading speed to obtain the theoretically driven predicted steering effect.

3. The method according to claim 1, characterized in that The dynamic parameters include the navigation speed and navigation angular velocity of the underwater vehicle; inputting the dynamic parameters into the steering efficiency prediction neural network to predict the steering efficiency of the underwater vehicle to obtain the data-driven predicted steering efficiency of the underwater vehicle, including: performing acceleration calculations at set time intervals based on the navigation speed to obtain navigation acceleration; performing acceleration calculations at set time intervals based on the navigation angular velocity to obtain navigation angular acceleration; Inputting the navigation acceleration and the navigation angular acceleration into the steering effect prediction neural network, and linearly combining the navigation acceleration and the navigation angular acceleration using basis functions corresponding to different weights to obtain a parameter feature map; The steering efficiency is predicted according to the parameter feature mapping to obtain the data-driven predicted steering efficiency.

4. The method according to claim 1, wherein The performing of fault diagnosis on the underwater vehicle based on the theoretical drive predicted steering efficiency, the theoretical prediction confidence, the data driven predicted steering efficiency, the data prediction confidence, and the dynamic parameters to obtain a fault diagnosis result of the underwater vehicle includes: performing confidence correction on the theoretical driven predicted steering effect according to the theoretical prediction confidence to obtain an adaptive theoretical predicted steering effect; performing confidence correction on the data driven predicted steering effect according to the data prediction confidence to obtain an adaptive data predicted steering effect; The rudder effect is predicted according to the adaptive theory, the rudder effect and the dynamic parameters are predicted according to the adaptive data, and the fault diagnosis network is used to perform fault diagnosis on the underwater vehicle to obtain a fault diagnosis result of the underwater vehicle.

5. The method according to claim 4, characterized in that The theoretical prediction confidence and the data prediction confidence are obtained by: Inputting the theoretically driven predicted steering efficiency and the data driven predicted steering efficiency into the fault diagnosis network for confidence prediction, thereby obtaining an initial theoretical confidence corresponding to the theoretically driven predicted steering efficiency and an initial data confidence corresponding to the data driven predicted steering efficiency; performing weight distribution adjustment on the theoretical driven predicted steering efficiency according to the initial theoretical confidence to obtain an iterative theoretical predicted steering efficiency; performing weight distribution adjustment on the data-driven predicted steering efficiency according to the initial data confidence level to obtain an iterative data predicted steering efficiency; Inputting the iterative theoretical predicted rudder efficiency and the iterative data predicted rudder efficiency into the fault diagnosis network for confidence prediction to update the initial theoretical confidence and the initial data confidence; repeating the above steps of confidence prediction and weight distribution adjustment until a preset stop condition is reached; The initial theoretical confidence is used as the theoretical prediction confidence, and the initial data confidence is used as the data prediction confidence.

6. The method according to claim 4, characterized in that The fault diagnosis network is trained in the following way: Using an ant colony algorithm to search for initial network parameters of the fault diagnosis network to obtain optimized network parameters of the fault diagnosis network; Based on the training data set, the fault diagnosis network set according to the optimized network parameters is model trained to update the optimized network parameters to obtain target network parameters; the fault diagnosis network is parameterized according to the target network parameters to complete the training.

7. The method according to any one of claims 1 to 6, characterized in that The fault diagnosis result includes the fault type and fault severity, and the operating status parameters include the remaining energy state of the underwater vehicle, the navigation mission type, and the environmental disturbance level; The performing of a fault emergency decision based on the fault diagnosis result and the operating status parameters of the underwater vehicle and outputting a fault-tolerant control strategy for handling the corresponding fault includes: Performing fuzzy classification on the fault type, the fault degree, the remaining energy state, the navigation mission type, and the environmental disturbance level to obtain fuzzy variables; The fuzzy variables are subjected to fuzzy reasoning based on pre-set fuzzy rules to obtain the fault-tolerant control strategy.

8. An underwater vehicle fault diagnosis device based on a physical data hybrid modeling strategy, characterized in that: The physical data hybrid modeling strategy adopts at least a rudder efficiency prediction physical model and a rudder efficiency prediction neural network; the device includes: a multi-strategy steering efficiency prediction module, configured to input the dynamic parameters of the underwater vehicle into the steering efficiency prediction physical model to perform steering efficiency prediction on the underwater vehicle, thereby obtaining a theoretical driven predicted steering efficiency of the underwater vehicle; and input the dynamic parameters into the steering efficiency prediction neural network to perform steering efficiency prediction on the underwater vehicle, thereby obtaining a data driven predicted steering efficiency of the underwater vehicle; wherein the theoretical driven predicted steering efficiency corresponds to a theoretical prediction confidence, and the data driven predicted steering efficiency corresponds to a data prediction confidence; a vehicle fault diagnosis module, configured to perform fault diagnosis on the underwater vehicle based on the theoretical drive predicted steering efficiency, the theoretical prediction confidence, the data driven predicted steering efficiency, the data prediction confidence, and the dynamic parameters, and obtain a fault diagnosis result for the underwater vehicle; The vehicle emergency decision module is used to make a fault emergency decision based on the fault diagnosis result and the operating status parameters of the underwater vehicle, and output a fault-tolerant control strategy for handling the corresponding fault.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 7.

Images