Abstract: The disclosure provides a method and system for trajectory tracking control of a vehicle-manipulator coupling system with finite time prescribed performance. Specifically, a coupling weaken trajectory planning method is designed to reduce the system's coupling effects. A finite time performance function is designed to constrain the trajectory tracking error. In a case the constraint conditions corresponding to the finite time performance function are satisfied, the trajectory tracking error is converted to obtain a transformed error. The sliding mode surface is designed based on the transformed error to control the transformed error to converge in a finite time, and the external disturbance of the vehicle-manipulator coupling system is observed based on non-linear disturbance observer. The control input of the vehicle-manipulator coupling system is designed based on the sliding mode surface and the non-linear disturbance observer output.

Abstract: The disclosure provides a method and system for trajectory tracking control of a vehicle-manipulator coupling system with finite time prescribed performance. Specifically, the trajectory tracking error of the vehicle-manipulator coupling system is obtained. A finite time performance function is designed to constrain the trajectory tracking error. In a case the constraint conditions corresponding to the finite time performance function are satisfied, the trajectory tracking error is converted to obtain a transformed error. The sliding mode surface is designed based on the transformed error to control the transformed error to converge in a finite time, and the external disturbance of the vehicle-manipulator coupling system is observed based on non-linear disturbance observer. The control input of the vehicle-manipulator coupling system is designed based on the sliding mode surface and the non-linear disturbance observer output.

Abstract: A method and a system for ship stability prediction by weighted fusion of RBFNN and random forest based on GD are provided. Firstly, input characteristics when predicting failure probabilities under different failure modes are determined through prior knowledge. Secondly, a mean square error of k-fold cross-validation is used as performance evaluation criterion of the RBFNN and the RF to search for model capacities of the RBFNN and the RF. Then, network parameters of the RBFNN are updated. Multiple random sample sets are generated using a bootstrap sampling method and are parallelly trained to generate multiple regression trees. A Gini index is used as an attribute division index, and a prediction result of the random forest is obtained. Finally, weight coefficients are introduced for weighted fusion of prediction results of the RBFNN and the RF. The weight coefficient is obtained by solving through iterative optimization of the gradient descent.

Abstract: The invention discloses an integrated detection method of electromagnetic searching, locating and tracking for subsea cables. After being launched into water, the cable-tracking AUV carries out primary Z-shaped reciprocating sailing to search the electromagnetic signal of the target subsea cable, when the electromagnetic signal reaches a preset threshold value, the AUV executes the cable-tracking detection. In the tracking process, if the target electromagnetic signal intensity is lower than the preset threshold, it is determined that subsea cable tracking is lost. At this time, the secondary Z-shaped cable-researching route planning and tracking are performed based on the lost point. In the process that the AUV autonomously tracks and detects the subsea cable, relative locating between AUV and subsea cable is performed based on the electromagnetic signal radiated by the subsea cable, and autonomous tracking control under the guidance of the electromagnetic locating signal is performed.

Abstract: The disclosure provides a method and system for hierarchical disturbance rejection depth tracking control of an underactuated underwater vehicle, and the depth tracking of the underactuated underwater vehicle is divided into kinematic layer guidance and dynamic layer pitch tracking. Adaptive line of sight guidance is used in the kinematic layer to convert a depth error into a desired pitch angle and to estimate and compensate an angle of attack to reject disturbance introduced by an unmeasurable true angle of attack. Based on the above, in the dynamic layer, the active disturbance rejection sliding mode pitch tracking method is used to observe a composite disturbance including an unknown dynamic model and an environmental disturbance by using the active disturbance rejection framework. The model is compensated as a unified integral series type, a sliding mode control law is finally designed to resist an observation error, and a control elevator angle is calculated.

Abstract: Disclosed are a non-singular finite-time control method and system for prescribed performance dynamic positioning of unmanned boat, which belong to the field of automatic control of unmanned boats. The method includes obtaining a difference between the actual measured position and the desired position of the unmanned boat to obtain a position error of the unmanned boat; performing prescribed performance transformation on the unmanned boat position error; constructing a non-singular finite-time virtual velocity law as a reference velocity for unmanned boats; obtaining the difference between the reference speed and the actual measured speed to obtain the speed tracking error; calculating the fuzzy supervisory saturation compensated law and adaptive fuzzy approximation term; constructing a non-singular finite-time dynamical controller to output control commands; applying a force or moment to the unmanned boat to adjust the propeller speed of the unmanned boat, thereby realizing positioning of the unmanned boat.

Abstract: The disclosure discloses a trajectory optimization method and device for accurately deploying marine sensors under water. The method includes the following steps. 1. Randomly select N sets of initial control variables within a range. 2. Input all of N sets of xi to the sensor's underwater glide kinematics and dynamics models, and calculate the smallest distance between N actual deployment positions and target deployment positions. 3. Determine whether the number of iterative operations is less than the preset value, if yes, perform global random walk and local random walk on N sets of xi, obtain N sets of xi again, and return to step 2; otherwise, go to step 4. 4. Output the control variable xi corresponding to ?s(x)nminmin and the corresponding trajectory as the optimal control variable and optimal trajectory. The disclosure can improve the accuracy of prediction on the underwater three-dimensional trajectory of the marine sensor.

Abstract: The invention discloses an integrated detection method of electromagnetic searching, locating and tracking for subsea cables. After being launched into water, the cable-tracking AUV carries out primary Z-shaped reciprocating sailing to search the electromagnetic signal of the target subsea cable, when the electromagnetic signal reaches a preset threshold value, the AUV executes the cable-tracking detection. In the tracking process, if the target electromagnetic signal intensity is lower than the preset threshold, it is determined that subsea cable tracking is lost. At this time, the secondary Z-shaped cable-researching route planning and tracking are performed based on the lost point. In the process that the AUV autonomously tracks and detects the subsea cable, relative locating between AUV and subsea cable is performed based on the electromagnetic signal radiated by the subsea cable, and autonomous tracking control under the guidance of the electromagnetic locating signal is performed.

Abstract: The disclosure discloses a trajectory optimization method and device for accurately deploying marine sensors under water. The method includes the following steps. 1. Randomly select N sets of initial control variables within a range. 2. Input all of N sets of xi to the sensor's underwater glide kinematics and dynamics models, and calculate the smallest distance between N actual deployment positions and target deployment positions. 3. Determine whether the number of iterative operations is less than the preset value, if yes, perform global random walk and local random walk on N sets of xi, obtain N sets of xi again, and return to step 2; otherwise, go to step 4. 4. Output the control variable xi corresponding to ?s(x)nminmin and the corresponding trajectory as the optimal control variable and optimal trajectory. The disclosure can improve the accuracy of prediction on the underwater three-dimensional trajectory of the marine sensor.