CN-122016232-A - Intelligent towing tank device based on Gaussian process regression and control method

Abstract

The invention relates to the field of hydrodynamic experimental equipment and discloses an intelligent towing tank device based on Gaussian process regression and a control method thereof, wherein the intelligent towing tank device comprises an experiment execution module, a control module and a control module, wherein the experiment execution module is used for carrying an experiment model in a towing tank and executing motion corresponding to experiment parameters; the intelligent experimental device comprises a data acquisition module, an intelligent decision module and a control module, wherein the data acquisition module is used for acquiring hydrodynamic parameters of the experimental model in a motion process in real time, the intelligent decision module is used for constructing and updating a Gaussian process regression model based on the hydrodynamic parameters acquired by the data acquisition module and corresponding experimental parameters, deciding a next group of experimental parameters to be executed based on the updated Gaussian process regression model, and the control module is used for controlling an experimental execution module to complete corresponding experiments according to the next group of experimental parameters output by the experimental parameter optimization unit. The device supports parameterized and non-parameterized operation modes, can efficiently cover a high-dimensional parameter space, and improves experimental efficiency by more than one order of magnitude compared with a traditional method.

Inventors

• FAN DIXIA
• Yuan Dehan
• LI RUIPENG

Assignees

• 西湖大学

Dates

Publication Date

20260512

Application Date

20260128

Claims (10)

1. 1. An intelligent towing tank device based on gaussian process regression, characterized in that the device comprises: The experiment execution module is used for carrying an experiment model in the towing tank and executing the motion corresponding to the experiment parameters; The data acquisition module is used for acquiring hydrodynamic parameters of the experimental model in the motion process in real time; The intelligent decision module is used for constructing and updating a Gaussian process regression model based on the hydrodynamic parameters and the corresponding experimental parameters acquired by the data acquisition module, and deciding a next group of experimental parameters to be executed based on the updated Gaussian process regression model; and the control module is used for controlling the experiment execution module to complete the corresponding experiment according to the next group of experiment parameters output by the experiment parameter optimization unit.
2. 2. The intelligent towing tank device according to claim 1, wherein the experiment execution module comprises a towing tank body, a guide rail carriage arranged along the length direction of the towing tank, a four-degree-of-freedom motion platform mounted on the guide rail carriage, and a driving unit connected with the four-degree-of-freedom motion platform, wherein the four-degree-of-freedom motion platform is used for carrying an experiment model and executing a combined motion track of forward flow, transverse flow and rotation defined by multi-dimensional experiment parameters, and the driving unit is used for driving the guide rail carriage and the four-degree-of-freedom motion platform to move and outputting corresponding actual motion track data.
3. 3. The intelligent towing tank device according to claim 2, wherein the data acquisition module comprises a multi-axis force sensor connected with the experimental model and a data acquisition card, the multi-axis force sensor is used for acquiring fluid force data born by the experimental model in the motion process, and the data acquisition card is used for synchronously acquiring the fluid force data and the actual motion trail data.
4. 4. The intelligent drag water pond apparatus of claim 3, wherein the intelligent decision module comprises a data preprocessing unit, a gaussian process regression modeling unit, an uncertainty evaluation unit, and an experimental parameter optimization unit, wherein: The data preprocessing unit is used for carrying out filtering and normalization processing on the collected fluid force data and the actual motion trail data, and constructing a multidimensional input-output data set for regression modeling; the Gaussian process regression modeling unit is used for constructing a Gaussian process regression model taking experimental parameters as input and fluid power as output based on the data set; the uncertainty evaluation unit is used for calculating a prediction mean value and prediction uncertainty of candidate experimental parameters in a parameter space based on the Gaussian process regression model; The experimental parameter optimization unit is used for determining the next group of experimental parameters in a feasible parameter space according to a preset active learning sampling criterion or a Bayesian optimization criterion based on the prediction mean value and the prediction uncertainty.
5. 5. The intelligent towing tank apparatus according to claim 4, wherein the experimental parameter optimizing unit operates according to an active learning sampling criterion that an experimental parameter having a largest predictive uncertainty value is selected as the next set of experimental parameters in a feasible experimental parameter space.
6. 6. The intelligent towing tank apparatus according to claim 4, wherein the experimental parameter optimizing unit calculates a confidence interval width of the predicted result corresponding to the candidate experimental parameter based on a gaussian process regression model, and determines the next set of experimental parameters by using the confidence interval width as a sampling index.
7. 7. The intelligent drag-water pond apparatus of claim 4, wherein the experimental parameter optimization unit operates according to bayesian optimization criteria that construct a desired improvement function of an objective function based on a gaussian process regression model and determine a next set of experimental parameters by solving the desired improvement function.
8. 8. The intelligent towing tank apparatus according to claim 4, wherein the experimental parameter optimizing unit considers both the prediction mean term and the prediction uncertainty term and constructs the evaluation function for experimental parameter selection in a weighted manner when determining the next set of experimental parameters.
9. 9. The intelligent drag water pond apparatus of claim 4, wherein the control module sets a convergence criterion based on the prediction uncertainty, and terminates the experimental parameter optimization process when the maximum prediction uncertainty of the gaussian process regression model in the parameter space is continuously below a preset threshold a plurality of times.
10. 10. An intelligent towing tank device control method based on Gaussian process regression is characterized by comprising the following steps: Carrying an experimental model in a towing tank and executing motion corresponding to experimental parameters; Collecting hydrodynamic parameters of the experimental model in the motion process in real time; Based on the hydrodynamic parameters and the corresponding experimental parameters acquired by the data acquisition module, constructing and updating a Gaussian process regression model, and deciding a next group of experimental parameters to be executed based on the updated Gaussian process regression model; and controlling an experiment execution module to complete a corresponding experiment according to the next group of experiment parameters output by the experiment parameter optimization unit.

Description

Intelligent towing tank device based on Gaussian process regression and control method Technical Field The invention relates to the field of fluid mechanics experimental equipment, in particular to an intelligent towing tank device based on Gaussian process regression and a control method. Background Fluid-structure interactions are the core research direction in the field of fluid mechanics, and many typical research objects have complex nonlinear response characteristics that need to be verified through a large number of experiments, such as vortex induced vibration and flutter wing trajectory parameter optimization. The traditional method is used for carrying out experiments through sequential hypothesis testing and trial-and-error methods, a high-dimensional parameter space is difficult to cover, vortex-induced vibration research relates to a plurality of independent parameters such as vibration frequency, amplitude, reynolds number and the like, system sampling in the high-dimensional parameter space is required to carry out massive experiments, the traditional method is not feasible, experimental parameter selection is dependent on manual experience, an experimental scheme cannot be adaptively adjusted according to existing data, data acquisition and analysis are disjointed, hidden physical rules are difficult to mine, manual errors are easy to introduce in manual operation, fluid disturbance cannot be effectively eliminated in experimental gaps, and cross contamination of experimental results is caused. Most of the existing laboratory automation equipment can only realize single repeated operation, lacks autonomous decision making capability based on machine learning, and cannot meet the requirements of complex fluid mechanics experiments on high-dimensional parameter exploration and high-efficiency data utilization. Therefore, development of an intelligent towing tank device with autonomous learning and decision making capability is needed to realize full automation and intellectualization of experimental procedures. Disclosure of Invention The invention aims to overcome the defects of low efficiency, parameter exploration limitation, insufficient intellectualization and the like of the traditional towing tank experiment, and provides a full-automatic intelligent towing tank device based on Gaussian process regression. An intelligent towing tank device based on gaussian process regression, comprising: The experiment execution module is used for carrying an experiment model in the towing tank and executing the motion corresponding to the experiment parameters; The data acquisition module is used for acquiring hydrodynamic parameters of the experimental model in the motion process in real time; The intelligent decision module is used for constructing and updating a Gaussian process regression model based on the hydrodynamic parameters and the corresponding experimental parameters acquired by the data acquisition module, and deciding a next group of experimental parameters to be executed based on the updated Gaussian process regression model; and the control module is used for controlling the experiment execution module to complete the corresponding experiment according to the next group of experiment parameters output by the experiment parameter optimization unit. The experiment execution module comprises a towing tank body, a guide rail sliding frame arranged along the length direction of the towing tank, a four-degree-of-freedom motion platform arranged on the guide rail sliding frame and a driving unit connected with the four-degree-of-freedom motion platform, wherein the four-degree-of-freedom motion platform is used for carrying an experiment model and executing a forward flow, a transverse flow and a rotation combined motion track defined by multidimensional experiment parameters, and the driving unit is used for driving the guide rail sliding frame and the four-degree-of-freedom motion platform to move and outputting corresponding actual motion track data. The data acquisition module comprises a multi-axis force sensor connected with the experimental model and a data acquisition card, wherein the multi-axis force sensor is used for acquiring fluid force data born by the experimental model in the motion process, and the data acquisition card is used for synchronously acquiring the fluid force data and the actual motion trail data. The intelligent decision module comprises a data preprocessing unit, a Gaussian process regression modeling unit, an uncertainty evaluation unit and an experimental parameter optimization unit, wherein: The data preprocessing unit is used for carrying out filtering and normalization processing on the collected fluid force data and the actual motion trail data, and constructing a multidimensional input-output data set for regression modeling; the Gaussian process regression modeling unit is used for constructing a Gaussian process regression model taking experimental pa