Multimodal obstacle avoidance navigation control method, system and device for automated mobile equipment

Abstract

The present application relates to the field of automation control technology, and discloses a multi-modal obstacle avoidance navigation control method, system and device for an automated mobile device. The method comprises: determining corresponding control modal characterization parameters based on the collision risk level of the automated mobile device and the target obstacle in the current navigation mission. According to the control modal characterization parameters, a target modal controller is selected from a plurality of pre-built modal controllers to perform a control mode switching operation. The target modal controller corresponds to a control objective function. Based on the solution result of the control objective function corresponding to the target modal controller, the thrusters and servos are driven to realize motion control and attitude adjustment of the automated mobile device. At this point, by dynamically selecting the obstacle avoidance strategy and cooperating with the corresponding modal controller, the autonomous obstacle avoidance capability and navigation safety of the automated mobile device in a complex and changing environment are effectively improved.

Description

Technical Field

[0001] The present application relates to the field of automation control technology, and in particular to a multi-modal obstacle avoidance navigation control method, system and device for an automated mobile device. Background Art

[0002] With the development of automated control technology, more and more automated mobile devices are emerging. Take underwater vehicles, for example. In recent years, as humanity's desire to explore the ocean continues to grow, their application areas have gradually expanded, from lakes to oceans, from nearshore to offshore, and from relatively safe areas to complex, unstructured marine environments. This has placed higher demands on vehicle design and control technology. However, when facing special operating conditions and complex tasks, the control technology of underwater vehicles in related technologies still has shortcomings. A new obstacle avoidance navigation control method is needed that can adapt to the complexity of the underwater environment and ensure the safety of underwater vehicles. Summary of the Invention

[0003] This application aims to solve, at least to some extent, one of the technical problems in the related art. To this end, this application proposes a multi-modal obstacle avoidance navigation control method, system, and device for an automated mobile device. The main technical solutions adopted in this application include:

[0004] In a first aspect, an embodiment of the present application provides a multi-modal obstacle avoidance navigation control method for an automated mobile device, the method comprising: determining corresponding control modal characterization parameters based on the collision risk level of the automated mobile device in a current navigation mission; wherein the collision risk level is used to describe the risk probability of a collision between the automated mobile device and a target obstacle in the current navigation mission; the control modal characterization parameters are used to refer to the obstacle avoidance control strategy required by the automated mobile device in the current navigation mission; a plurality of modal controllers suitable for a plurality of preset obstacle avoidance control strategies are pre-constructed, and a target modal controller is selected from the plurality of modal controllers according to the control modal characterization parameters to perform a control mode switching operation; wherein the target modal controller corresponds to a control objective function; and a thruster and a servo are driven based on the solution result of the control objective function corresponding to the target modal controller to realize motion control and attitude adjustment of the automated mobile device.

[0005] In a second aspect, an embodiment of the present application provides a multimodal obstacle avoidance navigation control system for an automated mobile device, the system comprising: an obstacle avoidance decision unit, for determining corresponding control modal characterization parameters based on the collision risk level of the automated mobile device in the current navigation mission; wherein the collision risk level is used to describe the risk probability of a collision between the automated mobile device and the target obstacle in the current navigation mission; the control modal characterization parameters are used to refer to the obstacle avoidance control strategy required by the automated mobile device in the current navigation mission; a multimodal navigation control unit, comprising multiple modal controllers suitable for multiple preset obstacle avoidance control strategies, for selecting a target modal controller from multiple modal controllers according to the control modal characterization parameters; wherein the target modal controller corresponds to a control objective function; a driving unit, for driving a thruster and a servo based on the solution result of the control objective function corresponding to the target modal controller, so as to realize motion control and attitude adjustment of the automated mobile device.

[0006] In a third aspect, an embodiment of the present application provides a multi-modal obstacle avoidance navigation control device for an automated mobile device, the device comprising: a control mode determination module for determining corresponding control mode characterization parameters based on the collision risk level of the automated mobile device in the current navigation mission; wherein the collision risk level is used to describe the risk probability of a collision between the automated mobile device and the target obstacle in the current navigation mission; the control mode characterization parameters are used to refer to the obstacle avoidance control strategy required by the automated mobile device in the current navigation mission; a control mode switching module for pre-building multiple mode controllers suitable for multiple preset obstacle avoidance control strategies, and selecting a target mode controller from the multiple mode controllers according to the control mode characterization parameters to perform a control mode switching operation; wherein the target mode controller corresponds to a control objective function; a drive control module for driving the thrusters and servos based on the solution result of the control objective function corresponding to the target mode controller to achieve motion control and attitude adjustment of the automated mobile device.

[0007] In a fourth aspect, the present application also provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of any of the above methods when the computer program is executed by a processor.

[0008] In a fifth aspect, the present application also provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any of the above methods when executing the computer program.

[0009] In a sixth aspect, the present invention provides a computer program product, comprising a computer program, which implements the steps of any of the above methods when executed by a processor.

[0010] In a seventh aspect, the present application also provides an underwater vehicle, the control method of which executes the contents of any of the above-mentioned multimodal obstacle avoidance navigation control methods for automated mobile devices.

[0011] In the above embodiment, the risk level of the automated mobile device colliding with an obstacle during the current mission is first assessed. Based on this risk level, the corresponding control modal characterization parameters are determined, thereby selecting the desired obstacle avoidance control strategy. A target modal controller corresponding to the selected strategy is then selected from multiple pre-built modal controllers. Each controller corresponds to a specific control objective function, which guides how the mobile device adjusts its motion and posture. Finally, based on the control objective function solved by the target modal controller, the thrusters and servos are driven to perform specific motion control and posture adjustments, thereby achieving effective obstacle avoidance. This allows the automated mobile device to dynamically select the most appropriate obstacle avoidance strategy based on the actual environment and mission requirements. Different controllers are also designed for different obstacle avoidance strategies, allowing the automated mobile device to flexibly manage obstacle avoidance actions during navigation, thereby improving its autonomous obstacle avoidance capabilities and navigation safety in complex and changing environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the specific implementation methods or the description of the prior art. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0013] Figure 1 This is a flow chart of a multi-modal obstacle avoidance navigation control method for an underwater vehicle according to one embodiment of the present application;

[0014] Figure 2 A flowchart of a method for determining a collision risk level according to an embodiment of the present application;

[0015] Figure 3 This is a schematic structural diagram of a multi-modal obstacle avoidance navigation control system for an underwater vehicle according to one embodiment of the present application;

[0016] Figure 4 This is a structural block diagram of a multi-modal obstacle avoidance navigation control device for an underwater vehicle according to one embodiment of the present application;

[0017] Figure 5 The figure is a diagram of the internal structure of a computer device provided according to one embodiment of the present application. 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 creative work are within the scope of protection of this application.

[0019] Taking underwater vehicles as an example, as a new force in the field of marine engineering, underwater vehicles play a vital role in various fields, including the national economy and marine defense, including but not limited to underwater detection, sampling, maritime search and rescue, seabed mapping, underwater early warning, and active attack. In recent years, with the increasing demand for ocean exploration, the application areas of underwater vehicles have gradually expanded, from lakes to oceans, from nearshore to offshore, and from relatively safe areas to complex, unstructured marine environments. This has placed higher demands on vehicle design and control technologies. Although research and application of underwater vehicles have matured both domestically and internationally, achieving a series of remarkable results, the control methods of underwater vehicles in related technologies often fail to meet the requirements and still exhibit serious limitations when encountering special operating conditions, such as extreme marine environments, performing complex missions, conducting detailed mapping in unknown waters, or operating in narrow and obstacle-filled spaces. These limitations may manifest in insufficient obstacle avoidance capabilities, limited autonomy, low navigation accuracy, or poor adaptability, restricting their application in a wider range of more challenging scenarios. Among the many technical bottlenecks in related technologies, the navigation safety of underwater vehicles (AUVs) is a particularly prominent issue, especially in unknown and unstructured ocean environments. In such environments, AUVs may encounter a variety of unexpected obstacles, and the diversity and unpredictability of these obstacles significantly increase the risks involved in navigation. Specifically, AUVs in related technologies often employ relatively simple strategies for obstacle avoidance control, which are quite limited in their ability and adaptability when faced with complex and changing obstacles. For example, some AUVs may only be equipped with basic sonar systems for obstacle detection, lacking more advanced sensor fusion technology and intelligent decision-making algorithms to effectively avoid potential collisions. Alternatively, due to their single control system, fixed steering angles may not be sufficient to avoid all obstacles in confined waters; or, when encountering fast-moving obstacles, pre-set stop and turn actions may not react quickly enough, resulting in obstacle avoidance failure. In other words, in emergency situations or when rapid obstacle avoidance decisions are required, the response speed and accuracy of these AUVs may not meet actual obstacle avoidance and safety requirements. Therefore, research on multimodal obstacle avoidance navigation control technologies is urgently needed.

[0020] Based on this, according to an embodiment of the present application, an embodiment of a multi-modal obstacle avoidance navigation control method for an automated mobile device 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 a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here. Specifically, a multi-modal obstacle avoidance navigation control method for an automated mobile device includes: determining a corresponding control mode characterization parameter based on the collision risk level of the automated mobile device in the current navigation mission; wherein the collision risk level is used to describe the risk probability of collision between the automated mobile device and the target obstacle in the current navigation mission; the control mode characterization parameter is used to refer to the obstacle avoidance control strategy required by the automated mobile device in the current navigation mission; multiple modal controllers suitable for multiple preset obstacle avoidance control strategies are pre-constructed, and a target modal controller is selected from the multiple modal controllers according to the control mode characterization parameter to perform a control mode switching operation; wherein the target modal controller corresponds to a control objective function; based on the solution result of the control objective function corresponding to the target modal controller, the thruster and the servo are driven to achieve motion control and attitude adjustment of the automated mobile device.

[0021] In this application, the risk level of the automated mobile device colliding with obstacles in the current mission is first evaluated. Then, based on this risk level, the corresponding control modal characterization parameters are determined to select the required obstacle avoidance control strategy. Then, a target modal controller corresponding to the selected strategy is selected from multiple pre-built modal controllers, where each controller corresponds to a specific control objective function, which is used to guide how the vehicle adjusts its motion and attitude. Finally, based on the solution of the control objective function of the target modal controller, the thrusters and servos are driven to perform specific motion control and attitude adjustment, thereby achieving effective obstacle avoidance. At this point, based on the actual navigation environment and mission requirements, the automated mobile device can dynamically select the most appropriate obstacle avoidance strategy. Different obstacle avoidance strategies are also designed with different controllers, allowing the automated mobile device to flexibly manage obstacle avoidance actions during navigation, thereby improving its autonomous obstacle avoidance capability and navigation safety in complex and changing environments.

[0022] It should be noted that the automated mobile equipment can be a ship on the water surface, an underwater vehicle, or a self-moving equipment on the road.

[0023] Taking an underwater vehicle as an example, this embodiment provides a multi-modal obstacle avoidance navigation control method for an underwater vehicle. Figure 1 FIG. 1 is a flow chart of a multi-modal obstacle avoidance navigation control method for an underwater vehicle according to an embodiment of the present application. Figure 1 As shown, the process includes the following steps:

[0024] S110: Determine corresponding control mode characterization parameters based on the collision risk level of the underwater vehicle in the current navigation mission.

[0025] The underwater vehicle's current navigation mission may refer to the type of mission the underwater vehicle is performing, such as exploration, mapping, search and rescue, or monitoring. Specifically, the current navigation mission may correspond to information such as the underwater vehicle's current movement behavior pattern, speed, heading, intended path, and target location. The collision risk level describes the risk of a collision between the underwater vehicle and the target obstacle during the current navigation mission. Specifically, the collision risk level can be categorized based on factors such as the distance between the underwater vehicle and the obstacle, the relative speed, the vehicle's maneuverability, and the characteristics of the obstacle. For example, the collision risk levels may be low, medium, or high. A low risk level may indicate a situation where the distance between the underwater vehicle and the obstacle is long, or the obstacle is moving slowly, posing a minimal direct threat to the vehicle. At this level, the vehicle can maintain its current navigation path and speed, requiring only minor adjustments to avoid the obstacle. A medium risk level may indicate a situation where the distance between the underwater vehicle and the obstacle is moderate, or the obstacle is moving rapidly, posing a moderate threat to the vehicle. At this level, the vehicle needs to adjust its path, possibly changing speed or direction, to ensure safe passage through the obstacle area. The high risk level can be used to indicate situations where the underwater vehicle is very close to an obstacle, or the obstacle is moving very quickly, posing a serious threat to the vehicle. At this level, the vehicle needs to take immediate avoidance measures, which may include emergency stops, rapid changes in direction, or the use of special avoidance strategies to avoid collision with the obstacle.

[0026] The control mode characterization parameters are used to refer to the obstacle avoidance control strategy required by the underwater vehicle in the current navigation mission. Specifically, the control mode characterization parameters are a set of parameters used to define and distinguish different control strategies or control modes. Exemplarily, the control mode characterization parameters can be specific parameter values ​​that identify the type of control strategy, or they can be text parameters defined for each mode, such as "normal navigation", "obstacle avoidance mode", "emergency obstacle avoidance", etc. Specifically, the corresponding control mode characterization parameters are determined based on the collision risk level of the underwater vehicle in the current navigation mission, which can be determined by pre-constructed risk-control mode relationship data. Exemplarily, the risk-control mode relationship data can be characterized in the form of a function or a table. It should be noted that the relationship between the collision risk level and the control mode characterization parameters can be flexible and diverse, including but not limited to one-to-one, many-to-one or many-to-many relationships. For example, if the collision risk level is divided into ten levels, then the risk level from level 1 to level 4 can correspond to the "normal navigation" of the control modal characterization parameter; the risk level from level 5 to level 7 can correspond to the "obstacle avoidance mode" of the control modal characterization parameter; and the risk level from level 8 to level 10 can correspond to the "emergency obstacle avoidance" of the control modal characterization parameter. Similarly, if the collision risk level is divided into three levels: low, medium, and high, then the risk-control modal relationship data can also be shown as follows:

[0027]

[0028] Where, represents the control mode characterization parameter is 1, represents the control mode characterization parameter is 2, The control modal characterization parameter is 3, that is, the collision risk level is low risk, and the corresponding control modal characterization parameter value can be 1; the collision risk level is medium risk, and the corresponding control modal characterization parameter value can be 2; the collision risk level is high risk, and the corresponding control modal characterization parameter value can be 3.

[0029] S120 , pre-building a plurality of modal controllers suitable for a plurality of preset obstacle avoidance control strategies, and selecting a target modal controller from the plurality of modal controllers according to control modal characterization parameters to perform a control mode switching operation.

[0030] Among them, the obstacle avoidance control strategy can refer to the control methods and steps designed to prevent the underwater vehicle from colliding with obstacles during navigation, and can include path replanning and attitude speed adjustment. The modal controller can refer to a pre-designed controller that responds to different obstacle avoidance control strategies and is used to guide the underwater vehicle in how to respond to specific environmental conditions and complete different control tasks. It should be understood that due to different obstacle avoidance control strategies, the underwater vehicle may issue control instructions through different control modal characterization parameters based on the complexity of the environment and the density of obstacles, allowing the control system to activate modal controllers with different control accuracies and adopt different obstacle avoidance control strategies to deal with emergencies with different collision risk levels.

[0031] Similarly, the relationship between control modal characterization parameters and modal controllers can be flexible and diverse, including but not limited to one-to-one, many-to-one, or many-to-many relationships. That is, the target modal controller can be single or multiple. For example, if there is a single target modal controller, one control modal characterization parameter can correspond to only one specific modal controller. For example, if the collision risk level is low, the corresponding control modal characterization parameter value can be 1. Upon receiving a control instruction containing this control modal characterization parameter, the control system will activate the first modal controller with the highest control accuracy and employ a high-precision tracking control obstacle avoidance strategy to control the underwater vehicle to avoid obstacles while completing the planned navigation path and posture. Similarly, multiple control modal characterization parameters can also point to the same modal controller. For example, if the collision risk level is any one of level 1 to level 4, the corresponding control modal characterization parameter value can be any one of 1 to 4. Upon receiving a control instruction containing any of the control modal characterization parameters from 1 to 4, the control system may only trigger the "first modal controller" with the highest control accuracy. However, in response to different control mode characterization parameters, the control parameter values ​​of this modal controller may be different to adjust the strength or details of the control strategy to control the underwater vehicle to avoid obstacles ahead while completing the original navigation path and posture.

[0032] Furthermore, in highly complex obstacle avoidance environments, multiple target modal controllers can be used. This means that multiple control modal characterization parameters can simultaneously influence multiple modal controllers, collaboratively adjusting the underwater vehicle's path planning and attitude control to adapt to the changing underwater environment, improving its autonomous obstacle avoidance capabilities and mission success rate. Specifically, a library containing multiple modal controllers can be pre-designed and constructed, each of which can correspond to a specific obstacle avoidance control strategy. The target modal controller can then be selected from the multiple modal controllers based on the control modal characterization parameters to perform control mode switching operations.

[0033] S130 , driving the thruster and the steering gear based on the solution of the control objective function corresponding to the target modal controller to achieve motion control and attitude adjustment of the underwater vehicle.

[0034] Specifically, after determining the control mode characterization parameters, the target mode controller can be selected and matched accordingly, and the target mode controller corresponds to different control objective functions, which can be mathematical functions used to guide and evaluate control performance, including tracking error and control energy consumption. For example, if the target mode controller is a normal navigation mode controller, then the control objective function can be an objective function that minimizes the tracking error between the current position of the aircraft and the predetermined path, and includes weight factors for adjusting the importance of tracking errors in different dimensions; if the target mode controller is a high-precision mode controller, then the control objective function can be an objective function that achieves high-precision tracking control of the six degrees of freedom of the position and attitude in the original route while taking into account energy consumption; if the target mode controller is an emergency obstacle avoidance mode controller, then the control objective function can be an objective function that achieves turning the aircraft away from obstacles in the shortest time without considering energy consumption.

[0035] Furthermore, after the control system of the underwater vehicle executes the switching operation from the current modal controller to the target modal controller, the control objective function can be used to solve the solution result including the operating parameters after the switch, and the solution result can be output to the thruster and servo to drive and control them to complete the corresponding actions, thereby realizing flexible and precise control of the underwater vehicle.

[0036] It is understood that during the operation of the target modal controller, environmental changes within the underwater vehicle and the real-time control effectiveness of the current target modal controller can be continuously monitored. If environmental conditions or mission requirements change, new control modal characterization parameters can be reassessed and determined. Based on these parameters, a new target modal controller can be reselected from the multiple modal controllers to execute a control mode switching operation. This allows the control strategy to be flexibly adjusted based on the new solution to the control objective function to cope with complex underwater environments.

[0037] In the above-described embodiment, the risk level of the vehicle colliding with obstacles during the current mission is first assessed. Based on this risk level, the corresponding control modal characterization parameters are determined, thereby selecting the desired obstacle avoidance control strategy. A target modal controller corresponding to the selected strategy is then selected from multiple pre-built modal controllers. Each controller corresponds to a specific control objective function, which guides how the vehicle adjusts its motion and attitude. Finally, based on the control objective function solution of the target modal controller, the thrusters and servos are driven to perform specific motion control and attitude adjustments, thereby achieving effective obstacle avoidance. This allows the underwater vehicle to dynamically select the most appropriate obstacle avoidance strategy based on the actual navigation environment and mission requirements. Different controllers are also designed for different obstacle avoidance strategies, allowing the underwater vehicle to flexibly manage obstacle avoidance actions during navigation, thereby improving its autonomous obstacle avoidance capabilities and navigation safety in complex and changing underwater environments.

[0038] In some embodiments, the automated mobile device is an underwater vehicle. The method further includes: performing risk prediction based on the relative positional relationship between the target obstacle and the underwater vehicle and the current navigation mission performed by the underwater vehicle to determine a collision risk level of the underwater vehicle.

[0039] The relative positional relationship between the target obstacle and the underwater vehicle can refer to the spatial positional relationship between the target obstacle and the underwater vehicle, including information such as the obstacle's bearing, distance, and altitude relative to the vehicle. Specifically, this multi-source information can be obtained through sonar, radar, vision systems, or other types of sensors and used to calculate the accurate positional relationship between the obstacle and the vehicle. Furthermore, after obtaining the relative positional relationship between the target obstacle and the underwater vehicle, as well as information about the underwater vehicle's current navigation mission, data analysis and decision-making algorithms can be used to predict the likelihood of a collision between the vehicle and the obstacle based on current environmental perception information, the vehicle's state, and mission requirements, thereby determining the underwater vehicle's collision risk level.

[0040] Optionally, when obtaining the relative positional relationship between the target obstacle and the underwater vehicle, the obstacle position, obstacle velocity, vehicle position, and vehicle velocity can be predicted separately to determine the relative positional relationship between the target obstacle and the underwater vehicle at that moment and within a certain period of time. For example, using obstacle position prediction as an example, this can be implemented using a long short-term memory (LSTM) algorithm. When constructing the LSTM model, the input layer step number can be set to β+1, where β∈{2,3,4,5,6}. The number of neuron nodes in the LSTM layer and the fully connected layer can be a power of 2, and the number of iterations can be set to θ, where θ∈{1,500,1000,1500,2000,2500,3000}. Initially, β=5, the number of neuron nodes in the LSTM layer and the fully connected layer can be 64, and the number of iterations θ=1500. Furthermore, during model training, preprocessed obstacle location time series data is used as the training set. This data is first fed into the constructed LSTM model. Specifically, the processed obstacle location time series data for β+1 consecutive moments (Y[t–β], Y[t–(β–1)], …, Y[t]) is used as the network input, and the predicted location at time t+1, Y[t+1], is used as the network output. Through multiple iterations of training, the model learns the changing patterns of obstacle locations. Furthermore, a cross-entropy loss function is used to evaluate the discrepancy between the current training results and the true distribution. An optimization algorithm is then used to adjust model parameters to minimize the loss function. Finally, after training is complete, the model outputs predicted obstacle location data.

[0041] Similarly, similar to obstacle position prediction, after obtaining the obstacle velocity prediction data, the aircraft position prediction data, and the aircraft velocity prediction data, these data can also be spliced. That is, the pre-processed obstacle position time series data and the obstacle position prediction data are spliced ​​together to obtain obstacle position splicing data; the pre-processed obstacle velocity time series data and the obstacle velocity prediction data are spliced ​​together to obtain obstacle velocity splicing data; similarly, the pre-processed aircraft position time series data and the aircraft position prediction data are spliced ​​together to obtain aircraft position splicing data; the pre-processed aircraft velocity time series data and the aircraft velocity prediction data are spliced ​​together to obtain aircraft velocity splicing data. Based on the obstacle position splicing time series, obstacle velocity splicing time series, aircraft position splicing time series, aircraft velocity splicing time series, and the target position information and target attitude information required by the current navigation mission, a risk level classification is performed to obtain a collision risk level. It should be noted that the collision risk level determination process is realized through a three-level data processing process. At the first level, the obstacle position, obstacle speed, aircraft position and aircraft speed are predicted respectively, and the corresponding obstacle position prediction data, obstacle speed prediction data, aircraft position prediction data and aircraft speed prediction data are obtained; at the second level, the corresponding pre-processed data is spliced ​​with the predicted data to realize the combination of actual data and predicted data, providing a more comprehensive data basis for risk level classification; at the third level, the risk level classification is carried out based on the obstacle position splicing timing sequence, obstacle speed splicing timing sequence, aircraft position splicing timing sequence, aircraft speed splicing timing sequence and the target position information and target attitude information required by the current navigation mission to obtain the collision risk level.

[0042] Optionally, further filtering can be performed on the obstacle position data, obstacle velocity data, aircraft position data, and aircraft velocity data to extract relevant features, such as relative distance, relative velocity, and relative angle. For example, the obstacle position data, obstacle velocity data, aircraft position data, and aircraft velocity data can be input into a gated residual network (GRN) to extract key features that best reflect attitude information, such as the relative distance, relative velocity, and relative angle between the obstacle and the aircraft. Furthermore, after obtaining these key features, the Euclidean distance between the obstacle and the aircraft can be directly calculated to obtain relative distance information. Similarly, the velocity difference between the obstacle and the aircraft can be directly calculated to obtain relative velocity information. The relative angle between the obstacle and the aircraft can also be directly calculated using a four-quadrant inverse tangent function. Finally, based on this attitude information, such as relative distance, relative velocity, and relative angle, as well as the target position and attitude information required by the current navigation mission, a risk classification is performed to determine the collision risk level.

[0043] In addition, the principal component analysis (PCA) method can also be used. First, in order to eliminate the influence of different feature dimensions, the obstacle position splicing data, obstacle speed splicing data, aircraft position splicing data, and aircraft speed splicing data can be standardized. Then calculate the covariance matrix of each data, and perform eigenvalue decomposition on its covariance matrix to obtain eigenvalues ​​and eigenvectors. Among them, the eigenvalue represents the variance of each data in the direction of its corresponding eigenvector, and the eigenvector represents the direction of the principal component of each data. Sort by eigenvalue from large to small, select the principal components corresponding to the first k eigenvectors, and construct the transformation matrix W. Further, the original data X of each data is mapped to the low-dimensional space by the following formula:

[0044]

[0045] Where Z is the data after dimensionality reduction; X is the normalized data matrix; and W is the matrix consisting of the first k eigenvectors.

[0046] Finally, relevant features, such as relative distance, relative speed, and relative angle, are extracted from the reduced data. Furthermore, after extracting the relevant features, a risk level classification model can be constructed based on the extracted features and the target position and attitude information required by the current navigation mission. Exemplarily, this risk level classification model can be a support vector machine (SVM), a decision tree, or a neural network. During model training, the training set (i.e., relevant features extracted based on the relative positional relationship between the target obstacle and the underwater vehicle and the current navigation mission being performed by the underwater vehicle) and the corresponding collision risk level labels are input into the classification model for training. This allows the model to learn the mapping relationship between different features and collision risk levels. Once training is complete, the model outputs the corresponding collision risk level, thereby enabling the prediction and classification of underwater vehicle collision risks.

[0047] In the above-described embodiment, by comprehensively utilizing multi-source sensor information, the relative positional relationship between the target obstacle and the underwater vehicle can be effectively determined. Combined with the vehicle's current navigation mission, the likelihood of a collision between the vehicle and the obstacle can be accurately predicted, thereby determining the collision risk level. This not only improves the underwater vehicle's perception of complex environments but also enhances the quality of its autonomous obstacle avoidance decisions, significantly enhancing the safety and reliability of the vehicle during mission execution and enabling it to more intelligently and flexibly respond to various potential collision threats.

[0048] In some embodiments, please refer to the attached Figure 2, based on the relative positional relationship between the target obstacle and the underwater vehicle and the current navigation mission performed by the underwater vehicle, risk prediction is performed to determine the collision risk level of the underwater vehicle, including:

[0049] S210: Acquire multi-source sensor data.

[0050] The multi-source sensor data includes the location information, size information, and speed information of the target obstacle, the current speed information, current location information, and current attitude information of the underwater vehicle, as well as the target location information and target attitude information required by the current navigation mission.

[0051] For example, multi-source sensors may include an inertial navigation system (INS) and a global positioning system (GPS) installed inside the aircraft (with the GPS receiver installed inside the aircraft and the antenna outside the hull), as well as a Doppler speed meter, collision avoidance sonar, and depth gauge installed outside the aircraft. Specifically, by using multiple collision avoidance sonars installed on the bow of the aircraft, obstacle distance information detected by each collision avoidance sonar can be obtained. Through fusion processing and analysis, the obstacle's location, size, and velocity information can be obtained.

[0052] Furthermore, the internal inertial navigation system can determine the vehicle's position, angular velocity, and angular acceleration. It is understood that the vehicle's position information can be expressed in a three-axis (x, y, z) carrier coordinate system. Based on the vehicle's position, angular velocity, and angular acceleration information, combined with the Doppler velocimeter, the vehicle's forward velocity can be determined. Ultimately, through comprehensive processing and analysis of this information, the underwater vehicle's current velocity, position, and attitude can be accurately determined.

[0053] Specifically, since GPS signals are lost underwater, it is necessary to first obtain initial position information using a GPS receiver on the surface to achieve high-precision initial position information. After the initial position is obtained, attitude calibration can be performed using an INS. The INS measures the vehicle's angular velocity and acceleration using gyroscopes and accelerometers, and uses this information to perform track recursion, providing real-time position, velocity, and attitude information. Underwater, INS data can be fused with data from a Doppler Velocity Log (DVL) and a depth gauge. The DVL provides forward velocity information, while the depth gauge provides depth information, both of which can be used to calibrate INS errors to improve navigation accuracy. Ultimately, the underwater vehicle's current velocity, position, and attitude information are obtained.

[0054] Similarly, based on the requirements of the current mission, the continuous position information provided by the INS, supplemented by the high-precision calibration information provided by the GPS, can determine the target position information required by the current mission. Furthermore, the INS can measure the angular velocity of the vehicle through the gyroscope, and through integration, obtain attitude angle information (heading angle, pitch angle, and roll angle). Based on this attitude information, the navigation parameters are then converted from the vehicle coordinate system to the navigation coordinate system, thereby determining the target attitude information required by the current mission.

[0055] S220: Utilize Kalman filtering to perform fusion processing on multi-source sensor data to obtain a fusion processing result.

[0056] Specifically, the state vector and observation vector are first determined based on the navigation requirements of the underwater vehicle. Furthermore, a system model and observation model are established. The system model can be a linear or nonlinear differential equation or difference equation that describes how the state vector changes over time. Specifically, the system model can be based on the kinematic and dynamic equations of the underwater vehicle. Similarly, the observation model can be a linear or nonlinear equation that describes the relationship between the observation vector and the state vector. After establishing the system and observation models, the Kalman filter is initialized. An initial estimate of the state vector is determined based on sensor data or other a priori information at the initial moment. The initial error covariance matrix for the state estimate is then set based on the sensor accuracy and the uncertainty of the initial state. The current state and error covariance are then predicted in a time update step. The Kalman gain is then calculated in a measurement update step. Simultaneously, the predicted state is corrected using observed data and the error covariance is updated. This iterative process of sensor data from each sampling period ultimately yields a fusion result. The fusion processing results include the current position, speed, and attitude of the underwater vehicle, the position, size, and speed of the target obstacle, and the target position and attitude required by the current navigation mission.

[0057] S230: Perform risk level classification based on the fusion processing result to obtain a collision risk level.

[0058] Specifically, after obtaining the fused data, the motion parameters of the obstacle and the vehicle are first normalized to eliminate the influence of different parameter dimensions and numerical ranges. Furthermore, risk classification can be implemented using a neural network model. For example, a SOM-BP neural network can be used to classify the input fusion results into a collision risk level. SOM-BP is a hybrid algorithm that combines a self-organizing map (SOM) and a back propagation (BP) neural network. Specifically, after training, the SOM network forms a two-dimensional grid with multiple neurons, each representing a cluster center. The input data, including the normalized position, size, and velocity of the target obstacle, the current velocity, position, and attitude of the underwater vehicle, and the target position and attitude required for the current navigation mission, are mapped to the neurons with the closest distance to them, thereby achieving dimensionality reduction and clustering of the data, and then extracting key features. Furthermore, the data clustered by the SOM network is used as the input of the BP neural network for nonlinear mapping, and combined with some features of the original input data, the obstacle avoidance risk level assessment result is finally output.

[0059] It's important to understand that when constructing a SOM-BP neural network, the network topology must first be determined. The number of neurons in the input layer is determined by the normalized parameter dimensions. For example, if the normalized parameters have 10 dimensions, the input layer will have 10 neurons. The number of neurons in the output layer corresponds to the number of categories of obstacle avoidance risk. Assuming the risk levels are categorized as low, medium, and high, the output layer will have 3 neurons. The number of neurons and layers in the hidden layer can be determined experimentally or empirically to find the optimal network structure. During training, the input dataset can be divided into a training set, a validation set, and a test set. The training set is used to train the network, adjusting the network weights and thresholds through the back propagation (BP) algorithm. The validation set is used to adjust network parameters to prevent overfitting, where the network performs well on the training set but poorly on new data. The test set is used to evaluate the network's final performance, ensuring that it performs well on unseen data. Furthermore, to optimize the weights and thresholds of the SOM-BP neural network, the particle swarm optimization (PSO) algorithm can be used. The PSO algorithm is a machine learning algorithm that simulates the foraging behavior of bird flocks to find the optimal solution. Specifically, when applying the PSO algorithm, the particle swarm is first initialized, including the particle position, velocity, and initial solution. The fitness value of each particle is then calculated, and the individual optimal solution (pbest) and global optimal solution (gbest) are updated. The particle velocity and position are updated based on pbest and gbest, and this process is repeated until a termination condition is met, such as reaching the maximum number of iterations or the fitness value reaching a preset threshold. Ultimately, the optimal solution found by the PSO algorithm serves as the initial parameters for the BP network, thereby improving training efficiency and prediction accuracy.

[0060] In the above-mentioned implementation, the integrated utilization of multiple sensors, including an inertial navigation system (INS), a global positioning system (GPS), a Doppler velocimeter, a collision avoidance sonar, and a depth gauge, achieves precise perception of the underwater vehicle and its surroundings. The Kalman filter algorithm is used to fuse this multi-source sensor data, not only improving the accuracy of the vehicle's position, velocity, and attitude, but also providing detailed dynamic information about obstacles. Furthermore, a neural network is used to classify the fused data into risk levels, enabling real-time assessment and quantification of collision risk. This provides decision support for the underwater vehicle and ultimately enhances its autonomous obstacle avoidance capabilities and navigation safety in complex environments.

[0061] In some embodiments, the plurality of modal controllers include a first modal controller, a second modal controller, and a third modal controller.

[0062] The first modal controller, the second modal controller, and the third modal controller may also refer to pre-designed controllers responsive to different obstacle avoidance control strategies, used to guide the underwater vehicle in responding to specific environmental conditions and completing different control tasks. Furthermore, the first modal controller may correspond to a first preset level, the second modal controller may correspond to a second preset level, and the third modal controller may correspond to a third preset level, with the collision risk probabilities corresponding to the first, second, and third preset levels increasing in sequence.

[0063] Specifically, the first modal controller is adapted to, when the collision risk level falls within a first preset level, perform high-precision tracking control of the underwater vehicle while simultaneously taking into account rudder angle constraints, energy consumption constraints, and rudder speed constraints. For example, if the first preset collision risk level is low, and the corresponding control modal characterization parameters determine that the first modal controller is used as the target modal controller, the first modal controller can be used to implement a normal path tracking navigation strategy. Specifically, Model Predictive Control (MPC) or Adaptive Control can be used to calculate the error between the vehicle's current state and the intended path, including information such as position error, velocity error, and attitude error. A control law is then configured to adjust the inputs to the propellers and servos in real time based on the error calculation results to reduce the error and guide the vehicle back to the intended path. Furthermore, when adjusting the propellers and servos, the rudder angle constraint (maximum rotation angle of the servo), energy consumption constraint (energy consumption of the vehicle), and rudder speed constraint (maximum rotation speed of the servo) are also considered to ensure vehicle safety and efficiency. Ultimately, while ensuring that the vehicle can track the predetermined path with high precision, it is also possible to optimize energy consumption and the use of servos, taking into account rudder angle constraints, energy consumption constraints, and rudder speed constraints.

[0064] The second modal controller is adapted to implement downgraded tracking control for the underwater vehicle while also taking into account rudder angle and rudder speed constraints when the collision risk level falls within the second preset level. Similarly, for example, if the second preset collision risk level is medium, and the corresponding control modal characterization parameters determine that the second modal controller is the target modal controller, the second modal controller can be used to implement a navigation strategy that downgrades tracking of the obstacle avoidance path. However, it should be noted that, unlike the first modal controller, the second modal controller primarily focuses on position tracking of the new obstacle avoidance path, rather than pursuing high-precision attitude tracking and energy optimization. Specifically, the second modal controller can combine information about obstacles and the vehicle and apply a path planning algorithm to develop a path that avoids obstacles. It then evaluates the deviation between the vehicle's current position and the newly planned path and, through control methods, adjusts the thruster and servo inputs to ensure the vehicle follows the obstacle avoidance path. Specifically, this process imposes constraints on the vehicle's pitch value to prevent excessive pitch angles from causing dangerous situations. Simultaneously, rudder angle and rudder speed constraints are also adhered to, ensuring the safety and effectiveness of servo operation.

[0065] The third modal controller is adapted to directly drive the thrusters and servos to avoid the target obstacle if the collision risk level falls within a third preset level. For example, if the third preset collision risk level is a high risk level and the third modal controller is determined as the target modal controller based on the corresponding control modal characterization parameters, it should be noted that in high-risk situations, the obstacle may pose a serious threat to the aircraft. Therefore, conventional constraints such as rudder angle constraints, rudder speed constraints, energy optimization, and high-precision attitude tracking are no longer considered. Instead, the third modal controller directly implements the emergency steering and obstacle avoidance navigation strategy. Specifically, the third modal controller can rapidly assess the relative state of the obstacle and the aircraft and immediately send commands to the servos and thrusters to adjust the rudder angle or maximum thrust to change the course to the maximum allowable value, allowing the aircraft to quickly escape the danger zone. Furthermore, during the emergency obstacle avoidance process, the controller continuously monitors the dynamics of both parties and adjusts the rudder angle and thrust in real time as needed until the aircraft completely avoids the obstacle and returns to a safe and stable state to address the serious threat.

[0066] In the above-mentioned embodiment, by designing three different levels of modal controllers, precise control of the underwater vehicle under different collision risk levels is achieved, significantly improving the vehicle's autonomous obstacle avoidance capability and the flexibility of mission execution. In low-risk situations, the first modal controller ensures the stable navigation of the vehicle along the predetermined path through high-precision tracking control, while optimizing energy consumption and servo use. When facing medium-risk situations, the second modal controller focuses on rapid obstacle avoidance, guiding the vehicle to avoid obstacles through path planning and adjustment of thruster and servo inputs, while taking into account pitch constraints and rudder angle and rudder speed limits. In high-risk emergency situations, the third modal controller takes emergency obstacle avoidance measures, directly changing the vehicle's heading rapidly with the maximum rudder angle or thrust to ensure a quick escape from danger. This multi-modal control strategy not only enhances the adaptability of the vehicle to complex environments, but also improves its safety and efficiency when performing diverse tasks.

[0067] In some implementations, if the target modal controller is a first modal controller, determining a control objective function corresponding to the first modal controller includes:

[0068] The tracking error of the underwater vehicle is determined based on the target position information required by the current navigation mission, the target attitude information required by the current navigation mission, the current position information of the underwater vehicle, the current attitude information of the underwater vehicle and the adjustable weight parameters.

[0069] The first tracking error may be error data that quantifies the deviation between the current position and attitude of the vehicle and the predetermined target position and attitude. The adjustable weight parameter may be data used to describe the relative importance of controlling energy consumption of the underwater vehicle when performing the current navigation mission.

[0070] Furthermore, the control energy consumption of the underwater vehicle is determined based on the four steering angles of the steering gear, the steering angle upper limit of the steering gear, the steering speed upper limit of the steering gear, and the adjustable weight parameter. Finally, a control objective function corresponding to the first modal controller is determined based on the first tracking error of the underwater vehicle and the control energy consumption.

[0071] For example, the control objective function corresponding to the first modal controller can be expressed in the following manner:

[0072]

[0073] in, Represents the control objective function corresponding to the first modal controller; Represents the target position information and target attitude information required by the current navigation mission; Represents the current position information and current attitude information of the underwater vehicle; Represents the four rudder angles of the servo; Represents the upper limit of the rudder angle of the servo; is the rate of change of the rudder angle per unit time, which represents the rudder speed; Represents the upper limit of the servo's rudder speed; It represents the adjustable weight parameter for controlling the energy consumption of the underwater vehicle when performing the current navigation mission.

[0074] It can be understood that the control objective function of the first modal controller is equivalent to defining an optimization problem, which aims to achieve accurate path tracking of the underwater vehicle through the first modal controller while also considering the control energy consumption. Specifically, the control objective function It consists of two parts. The first part is the tracking error which quantifies the deviation between the current position and attitude of the spacecraft and the predetermined target position and attitude, including position deviation 、 and Deviation from posture 、 and The second part is the control energy consumption, which is set with a weight parameter adjusted between 0 and 1 to balance the relative importance of tracking accuracy and control energy consumption in the objective function. It is related to the four rudder angles of the servo, and is determined by Furthermore, constraints are set to ensure that the rudder angle and its rate of change do not exceed a preset upper limit, thereby ensuring the safety of the vehicle's operation and the durability of the servo. Ultimately, by minimizing this control objective function, the underwater vehicle can accurately track the planned path and attitude while effectively controlling its energy consumption.

[0075] In the above implementation, the control objective function for the first modal controller is designed to ensure that the underwater vehicle accurately tracks its intended path and attitude while also effectively controlling energy consumption. Specifically, an optimization problem is defined that includes tracking error and control energy consumption, and weight parameters are introduced to balance the two components, achieving the dual goals of high-precision navigation and energy conservation.

[0076] In some embodiments, if the target modal controller is a second modal controller, determining the control objective function corresponding to the second modal controller includes:

[0077] A pitch constraint value of the underwater vehicle is determined, and a control objective function corresponding to the second modal controller is determined based on the current position information of the underwater vehicle, the obstacle avoidance position information of the adjusted underwater vehicle, and the pitch constraint value.

[0078] The underwater vehicle's pitch constraint value may include a tilt range defined by the vehicle's minimum and maximum pitch constraints, which is used to limit the vehicle's tilt within the vertical plane, ensuring its attitude remains within safe, stable, and efficient operating limits. It is also understood that the control objective function of the second modal controller also defines an optimization problem, but its purpose is to achieve obstacle avoidance path tracking for the underwater vehicle at a moderate collision risk level. In this control objective function, only the position tracking problem of the new obstacle avoidance path is considered, and high-precision attitude tracking and energy optimization are no longer considered. Specifically, this control objective function focuses on minimizing the three-dimensional distance error between the vehicle's current position and the adjusted obstacle avoidance target position to guide the vehicle along the updated obstacle avoidance path. Furthermore, the vehicle's pitch constraint (defined by minimum and maximum values) and physical limitations of servo operation, such as upper limits on rudder angle and rudder speed, are also designed to ensure the vehicle's stability and safety during obstacle avoidance maneuvers.

[0079] For example, the control objective function corresponding to the second modal controller can be expressed in the following manner:

[0080]

[0081] in, Represents the control objective function corresponding to the second modal controller; Represents the current position information of the underwater vehicle; Represents the obstacle avoidance position information of the underwater vehicle after adjustment; Represents the minimum value of the pitch constraint of the underwater vehicle; Represents the maximum value of the longitudinal constraint of the underwater vehicle; Represents the four rudder angles of the servo; is the rate of change of the rudder angle per unit time, which represents the rudder speed; Represents the upper limit of the rudder angle of the servo; Represents the upper limit of the servo's steering speed.

[0082] In the above implementation, the control objective function for the second modal controller is designed to enable the underwater vehicle to achieve effective path tracking under moderate collision risk levels. This control objective function no longer pursues high-precision attitude tracking and energy optimization, but instead prioritizes ensuring that the vehicle can follow the updated obstacle avoidance path. It also imposes a certain degree of constraint on the vehicle's pitch value to avoid dangerous conditions where the pitch exceeds the limit.

[0083] In some embodiments, if the target modal controller is a third modal controller, determining the control objective function corresponding to the third modal controller includes:

[0084] The control objective function corresponding to the third modal controller is determined based on the steering angle value calculated at the previous moment and the obstacle avoidance steering angle deviation value of the underwater vehicle.

[0085] The steering angle value calculated at the previous moment can refer to the rudder angle command value calculated by the control algorithm based on the vehicle state (such as position, attitude, or velocity) at the previous time step (previous moment), while the obstacle avoidance rudder angle deviation value can refer to the additional rudder angle adjustment introduced to avoid obstacles. It can be understood that, unlike the control objective functions of the first and second modal controllers, the third modal controller is more often used in situations with high collision risk, where the detected obstacle may pose a serious threat to the vehicle. Its control objective function prioritizes emergency obstacle avoidance, temporarily ignoring conventional requirements such as rudder angle constraints, rudder speed limits, energy consumption optimization, and high-precision attitude tracking. Specifically, the control objective function of the third modal controller can use the rudder angle value at the previous moment as the initial condition, superimpose the obstacle avoidance rudder angle deviation value on this basis, generate a new rudder angle command, and implement the emergency steering obstacle avoidance navigation strategy through the third modal controller.

[0086] For example, the control objective function corresponding to the third modal controller can be expressed in the following manner:

[0087]

[0088] in, Represents the obstacle avoidance rudder angle value of the servo after adjustment. Represents the steering angle value calculated at the previous moment, Represents the obstacle avoidance rudder angle deviation value of the underwater vehicle.

[0089] In the above implementation, the control objective function for the third modal controller is designed to focus solely on calculating new rudder angle values ​​to minimize the relative distance to obstacles. This allows the underwater vehicle to quickly and effectively perform obstacle avoidance maneuvers in high-risk emergency situations, enabling rapid response to sudden collision threats. This improves the vehicle's survivability and safety.

[0090] It should be understood that, although the various steps in the above flowchart are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the above flowchart may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily to be performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0091] This specification also provides a multi-modal obstacle avoidance navigation control system for an automated mobile device. Figure 3 The multimodal obstacle avoidance navigation control system 300 includes: an obstacle avoidance decision unit 310, a multimodal navigation control unit 320 and a driving unit 330.

[0092] The obstacle avoidance decision unit 310 is configured to determine corresponding control mode characterization parameters based on the collision risk level of the automated mobile device in the current navigation mission.

[0093] Specifically, the obstacle avoidance decision unit 310 also includes a risk analyzer 311 and an obstacle avoidance decision maker 313. The risk analyzer 311 is responsible for assessing the collision risk level of the automated mobile device during the current navigation mission. The collision risk level describes the probability of a collision between the automated mobile device and a target obstacle during the current navigation mission. The obstacle avoidance decision maker 313 is responsible for determining the required obstacle avoidance control strategy, namely, the control mode characterization parameters, based on the collision risk level. The control mode characterization parameters represent the obstacle avoidance control strategy required for the automated mobile device during the current navigation mission.

[0094] The multi-modal navigation control unit 320 includes multiple modal controllers suitable for multiple preset obstacle avoidance control strategies, and is configured to select a target modal controller from the multiple modal controllers based on control modal characterization parameters. The target modal controller corresponds to a control objective function.

[0095] For example, please refer to Figure 3The multimodal navigation control unit 320 may include a switching manager 321, a first mode controller 323, a second mode controller 325, and a third mode controller 327. The switching manager 321 may select a corresponding mode controller to perform a control mode switching operation based on the control mode characterization parameters determined by the obstacle avoidance decision unit 310. The first mode controller 323, the second mode controller 325, and the third mode controller 327 may be controllers designed to respond to different control objective functions and implement different obstacle avoidance control strategies under different collision risk levels.

[0096] The driving unit 330 is used to drive the propeller and the servo based on the solution of the control objective function corresponding to the target mode controller, so as to realize the motion control and attitude adjustment of the automated mobile device.

[0097] For example, the drive unit 330 may include a propeller 331 and servos, wherein the number of servos may be four, namely a first servo 332, a second servo 334, a third servo 336, and a fourth servo 338. It is understood that after receiving the solution results of the modal controller in the multimodal navigation control unit 320, the drive unit 330 may drive the propeller 331 and the four servos in accordance with the control objective function solution results based on the target modal controller to achieve motion control and attitude adjustment of the automated mobile device.

[0098] Optionally, the multimodal obstacle avoidance navigation control system 300 may further include a sensor 340. It will be appreciated that the sensor 340 may include multi-source sensor data, allowing the control system to obtain various environmental and vehicle status information through the sensors, providing necessary data support for the obstacle avoidance decision unit 310. Furthermore, the multimodal obstacle avoidance navigation control system 300 may also compare the preset target navigation path (i.e., ideal trajectory) of the automated mobile device during the current navigation mission with the automated mobile device's real-time route during the current navigation mission, similarly providing necessary data support to the obstacle avoidance decision unit 310 to assist in obstacle avoidance decisions.

[0099] In this embodiment, the multimodal obstacle avoidance navigation control system for an automated mobile device is presented in the form of functional units, where the units refer to an ASIC (Application Specific Integrated Circuit), a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the aforementioned functions. The specific definition of the multimodal obstacle avoidance navigation control system for an automated mobile device can be found in the definition of the multimodal obstacle avoidance navigation control method for an underwater vehicle described above and will not be repeated here.

[0100] Taking an underwater vehicle as an example, the embodiment of this specification also provides a multi-modal obstacle avoidance navigation control device 400 for an underwater vehicle, such as Figure 4 As shown, it includes: a control mode determination module 410, a control mode switching module 420 and a drive control module 430, wherein:

[0101] The control mode determination module 410 is used to determine the corresponding control mode characterization parameters based on the collision risk level of the underwater vehicle in the current navigation mission; wherein the collision risk level is used to describe the risk probability of the underwater vehicle colliding with the target obstacle in the current navigation mission; the control mode characterization parameters refer to the obstacle avoidance control strategy required by the underwater vehicle in the current navigation mission.

[0102] The control mode switching module 420 is used to pre-build multiple modal controllers suitable for multiple preset obstacle avoidance control strategies, and select a target modal controller from the multiple modal controllers according to the control mode characterization parameters to perform the control mode switching operation; wherein, the target modal controller corresponds to the control objective function.

[0103] The drive control module 430 is used to drive the propeller and the steering gear based on the solution of the control objective function corresponding to the target mode controller to achieve motion control and attitude adjustment of the underwater vehicle.

[0104] In some embodiments, a multimodal obstacle avoidance navigation control device 400 for an underwater vehicle further includes a risk prediction module for performing risk prediction based on the relative positional relationship between the target obstacle and the underwater vehicle and the current navigation mission performed by the underwater vehicle to determine the collision risk level of the underwater vehicle.

[0105] In some embodiments, the risk prediction module is also used to perform risk prediction based on the relative position relationship between the target obstacle and the underwater vehicle and the current navigation mission performed by the underwater vehicle, and determine the collision risk level of the underwater vehicle, including: obtaining multi-source sensor data; wherein the multi-source sensor data includes the position information, size information, and speed information of the target obstacle, the current speed information, current position information, and current attitude information of the underwater vehicle, and the target position information and target attitude information required by the current navigation mission; using Kalman filtering to fuse the multi-source sensor data to obtain a fusion processing result; and performing risk level classification based on the fusion processing result to obtain a collision risk level.

[0106] In some embodiments, the multiple modal controllers include a first modal controller, a second modal controller, and a third modal controller. The control mode switching module 420 is further configured to determine a first modal controller, adapted to, when the collision risk level falls within a first preset level, perform high-precision tracking control on the underwater vehicle while taking into account rudder angle constraints, energy consumption constraints, and rudder speed constraints; determine a second modal controller, adapted to, when the collision risk level falls within a second preset level, perform degraded tracking control on the underwater vehicle while taking into account rudder angle constraints and rudder speed constraints; and determine a third modal controller, adapted to, when the collision risk level falls within a third preset level, directly drive the propeller and steering gear to avoid the target obstacle; wherein the collision risk probabilities corresponding to the first preset level, the second preset level, and the third preset level increase in sequence.

[0107] In some embodiments, if the target mode controller is the first mode controller; the control mode switching module 420 is also used to determine the control objective function corresponding to the first mode controller, including: determining the first tracking error of the underwater vehicle based on the target position information required by the current navigation mission, the target attitude information required by the current navigation mission, the current position information of the underwater vehicle, the current attitude information of the underwater vehicle and the adjustable weight parameters; wherein the adjustable weight parameters are used to describe the relative importance of the control energy consumption of the underwater vehicle when performing the current navigation mission; determining the control energy consumption of the underwater vehicle according to the four rudder angles of the servo, the rudder angle upper limit of the servo, the rudder speed upper limit of the servo and the adjustable weight parameters; determining the control objective function corresponding to the first mode controller based on the tracking error and control energy consumption of the underwater vehicle.

[0108] In some embodiments, if the target mode controller is a second mode controller; the control mode switching module 420 is also used to determine the control objective function corresponding to the second mode controller, including: determining the longitudinal tilt constraint value of the underwater vehicle; determining the control objective function corresponding to the second mode controller based on the current position information of the underwater vehicle, the obstacle avoidance position information of the adjusted underwater vehicle and the longitudinal tilt constraint value.

[0109] In some embodiments, if the target mode controller is a third mode controller; the control mode switching module 420 is also used to determine the control objective function corresponding to the third mode controller, including: determining the control objective function corresponding to the third mode controller based on the steering angle value calculated at the previous moment and the obstacle avoidance rudder angle deviation value of the underwater vehicle.

[0110] The specific definitions of a multimodal obstacle avoidance navigation control device for an underwater vehicle can be found in the definitions of a multimodal obstacle avoidance navigation control method for an underwater vehicle described above and will not be repeated here. Each module in the aforementioned multimodal obstacle avoidance navigation control device for an underwater vehicle can be implemented in whole or in part via software, hardware, or a combination thereof. Each of these modules can be embedded in or independent of a processor in a computer device in hardware form, or stored in a computer device memory in software form, allowing the processor to call and execute the corresponding operations of each module.

[0111] In this embodiment, a multimodal obstacle avoidance navigation control device for an underwater vehicle 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.

[0112] The present application also provides an underwater vehicle, wherein the underwater vehicle control method implements any of the aforementioned multimodal obstacle avoidance navigation control methods for underwater vehicles. The specific definition of the underwater vehicle can be found in the definition of the multimodal obstacle avoidance navigation control method for underwater vehicles described above, and will not be further elaborated here.

[0113] The embodiment of the present application further provides a computer device, which may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 5 As shown. The computer device includes a processor, memory, communication interface, display screen and input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a multimodal obstacle avoidance navigation control method for an underwater vehicle is implemented. The display screen of the computer device can be a liquid crystal display or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad provided on the computer device housing, or an external keyboard, touchpad or mouse, etc. Those skilled in the art can understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or have a different component arrangement.

[0114] 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 on a storage medium, or downloaded via a network and originally stored in a remote storage medium or a non-transitory machine-readable storage medium and then 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. It is understood that a computer 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, processor, or hardware, the method shown in the above embodiment is implemented.

[0115] 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.

[0116] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, features specified as "first" or "second" may explicitly or implicitly include at least one of such features. In the description of this application, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined. The various embodiments in this specification are described in a progressive manner, and similar parts between the various embodiments can be referenced across them. Each embodiment focuses on the differences from other embodiments. Because they are generally similar to the method embodiments, the description is relatively simple, and relevant parts can be referenced to the description of the method embodiments. The foregoing is merely an example of the present application and is not intended to limit the present application. Various modifications and variations are readily apparent to those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of this application are intended to be encompassed by the claims of this application. Although the embodiments of this 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 this application, and such modifications and variations are intended to be within the scope of the appended claims.

Claims

1. A multi-modal obstacle avoidance navigation control method for an automated mobile device, characterized in that: The automated mobile device is an underwater vehicle; the method comprises: Determining corresponding control mode characterization parameters based on the collision risk level of the automated mobile device in the current navigation mission; wherein the collision risk level is used to describe the risk probability of the automated mobile device colliding with a target obstacle in the current navigation mission; and the control mode characterization parameters are used to refer to the obstacle avoidance control strategy required by the automated mobile device in the current navigation mission; A plurality of modal controllers suitable for a plurality of preset obstacle avoidance control strategies are pre-constructed, and a target modal controller is selected from the plurality of modal controllers according to the control modal characterization parameters to perform a control mode switching operation; wherein the target modal controller corresponds to a control objective function; the plurality of modal controllers include a first modal controller, a second modal controller, and a third modal controller; the first modal controller is adapted to, if the collision risk level belongs to a first preset level, perform high-precision tracking control on the underwater vehicle while taking into account rudder angle constraints, energy consumption constraints, and rudder speed constraints; the second modal controller is adapted to, if the collision risk level belongs to a second preset level, perform degraded tracking control on the underwater vehicle while taking into account rudder angle constraints and rudder speed constraints; the third modal controller is adapted to, if the collision risk level belongs to a third preset level, directly drive the thrusters and steering gear to avoid the target obstacle; wherein the collision risk probabilities corresponding to the first preset level, the second preset level, and the third preset level increase in sequence; If the target modal controller is the first modal controller, the control objective function corresponding to the first modal controller is determined in the following manner, including: determining a first tracking error of the underwater vehicle based on the target position information required by the current navigation mission, the target attitude information required by the current navigation mission, the current position information of the underwater vehicle, the current attitude information of the underwater vehicle, and an adjustable weight parameter; wherein the adjustable weight parameter is used to describe the relative importance of control energy consumption of the underwater vehicle when performing the current navigation mission; determining the control energy consumption of the underwater vehicle based on the four rudder angles of the servo, the rudder angle upper limit of the servo, the rudder speed upper limit of the servo, and the adjustable weight parameter; determining the control objective function corresponding to the first modal controller based on the tracking error and the control energy consumption of the underwater vehicle; and driving the thruster and the servo based on the solution result of the control objective function corresponding to the target modal controller to realize motion control and attitude adjustment of the automated mobile device.

2. The method according to claim 1, characterized in that The method further comprises: A risk prediction is performed based on the relative positional relationship between the target obstacle and the underwater vehicle and the current navigation mission performed by the underwater vehicle to determine the collision risk level of the underwater vehicle.

3. The method according to claim 2, characterized in that The performing risk prediction based on the relative positional relationship between the target obstacle and the underwater vehicle and the current navigation mission performed by the underwater vehicle to determine the collision risk level of the underwater vehicle includes: Acquiring multi-source sensor data; wherein the multi-source sensor data includes the position information, size information, and speed information of the target obstacle, the current speed information, current position information, and current attitude information of the underwater vehicle, and the target position information and target attitude information required by the current navigation mission; Performing fusion processing on the multi-source sensor data using Kalman filtering to obtain a fusion processing result; Risk level classification is performed based on the fusion processing result to obtain the collision risk level.

4. The method according to claim 1, wherein If the target modal controller is the second modal controller, determining a control objective function corresponding to the second modal controller includes: determining a pitch constraint value of the underwater vehicle; A control objective function corresponding to the second modal controller is determined based on the current position information of the underwater vehicle, the adjusted obstacle avoidance position information of the underwater vehicle, and the pitch constraint value.

5. The method according to claim 1, wherein If the target modal controller is the third modal controller, determining a control objective function corresponding to the third modal controller includes: The control objective function corresponding to the third modal controller is determined based on the steering angle value calculated at the previous moment and the obstacle avoidance steering angle deviation value of the underwater vehicle.

6. A multi-modal obstacle avoidance navigation control system for an automated mobile device, wherein the automated mobile device is an underwater vehicle; characterized in that: The system comprises: an obstacle avoidance decision unit, configured to determine corresponding control mode characterization parameters based on a collision risk level of the automated mobile device in a current navigation mission; wherein the collision risk level is used to describe the risk probability of a collision between the automated mobile device and a target obstacle in the current navigation mission; and the control mode characterization parameters are used to indicate an obstacle avoidance control strategy required by the automated mobile device in the current navigation mission; A multi-modal navigation control unit, comprising a plurality of modal controllers suitable for a plurality of preset obstacle avoidance control strategies, for selecting a target modal controller from the plurality of modal controllers according to the control modal characterization parameters; wherein the target modal controller corresponds to a control objective function; the plurality of modal controllers include a first modal controller, a second modal controller, and a third modal controller; the first modal controller is suitable for, if the collision risk level belongs to the first preset level, performing high-precision tracking control on the underwater vehicle while taking into account rudder angle constraints, energy consumption constraints, and rudder speed constraints; the second modal controller is suitable for, if the collision risk level belongs to the second preset level, performing degraded tracking control on the underwater vehicle while taking into account rudder angle constraints and rudder speed constraints; the third modal controller is suitable for, if the collision risk level belongs to the third preset level, directly driving the thruster and the steering gear to avoid the target obstacle; wherein, the first preset level , the collision risk probabilities corresponding to the second preset level and the third preset level increase in sequence; if the target modal controller is the first modal controller, the control objective function corresponding to the first modal controller is determined in the following manner, including: determining a first tracking error of the underwater vehicle based on the target position information required by the current navigation mission, the target attitude information required by the current navigation mission, the current position information of the underwater vehicle, the current attitude information of the underwater vehicle and an adjustable weight parameter; wherein the adjustable weight parameter is used to describe the relative importance of the control energy consumption of the underwater vehicle when performing the current navigation mission; determining the control energy consumption of the underwater vehicle according to the four rudder angles of the servo, the rudder angle upper limit of the servo, the rudder speed upper limit of the servo and the adjustable weight parameter; determining the control objective function corresponding to the first modal controller based on the tracking error of the underwater vehicle and the control energy consumption; A driving unit is used to drive a propeller and a steering gear based on the solution result of the control objective function corresponding to the target mode controller, so as to realize the motion control and attitude adjustment of the automated mobile device.

7. A multi-modal obstacle avoidance navigation control device for an automated mobile device, characterized in that: The automated mobile device is an underwater vehicle; the device comprises: a control mode determination module, configured to determine corresponding control mode characterization parameters based on a collision risk level of the automated mobile device in a current navigation mission; wherein the collision risk level is used to describe the risk probability of a collision between the automated mobile device and a target obstacle in the current navigation mission; and the control mode characterization parameters are used to indicate an obstacle avoidance control strategy required by the automated mobile device in the current navigation mission; A control mode switching module is used to pre-construct multiple mode controllers suitable for multiple preset obstacle avoidance control strategies, and select a target mode controller from the multiple mode controllers according to the control mode characterization parameters to perform a control mode switching operation; wherein, the target mode controller corresponds to the control objective function; the multiple mode controllers include a first mode controller, a second mode controller, and a third mode controller; the first mode controller is suitable for, if the collision risk level belongs to the first preset level, performing high-precision tracking control on the underwater vehicle while taking into account the rudder angle constraint, energy consumption constraint and rudder speed constraint; the second mode controller is suitable for, if the collision risk level belongs to the second preset level, performing degraded tracking control on the underwater vehicle while taking into account the rudder angle constraint and rudder speed constraint; the third mode controller is suitable for, if the collision risk level belongs to the third preset level, directly driving the thruster and the steering gear to avoid the target obstacle; wherein, The collision risk probabilities corresponding to the first preset level, the second preset level, and the third preset level increase in sequence; if the target modal controller is the first modal controller, determining the control objective function corresponding to the first modal controller in the following manner, including: determining a first tracking error of the underwater vehicle based on the target position information required by the current navigation mission, the target attitude information required by the current navigation mission, the current position information of the underwater vehicle, the current attitude information of the underwater vehicle, and an adjustable weight parameter; wherein the adjustable weight parameter is used to describe the relative importance of control energy consumption of the underwater vehicle when performing the current navigation mission; determining the control energy consumption of the underwater vehicle based on the four rudder angles of the servo, the rudder angle upper limit of the servo, the rudder speed upper limit of the servo, and the adjustable weight parameter; and determining the control objective function corresponding to the first modal controller based on the tracking error of the underwater vehicle and the control energy consumption; A drive control module is used to drive the propeller and the steering gear based on the solution result of the control objective function corresponding to the target mode controller, so as to realize the motion control and attitude adjustment of the automated mobile device.

8. 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 5.

Images