Structural object active control method for fixed-point wave absorption in wave water tank

Abstract

The invention discloses a structure active control method for fixed-point wave absorption in a wave water tank. The control method comprises the following steps: acquiring time sequence motion features of feature points of a structure and time sequence radiation wave heights at different measuring points on two sides of the structure, and constructing a data set; training a machine learning model by using the data set; and collecting the wave height of a point to be subjected to wave absorption and the immersion depth of the feature point of the structure in real time, and predicting the real-time motion amplitude of the structure by using the trained machine learning model according to the real-time wave height of the point to be subjected to wave absorption, the real-time immersion depth of the feature point of the structure and the distance of the point to be subjected to wave absorption. And controlling the structure to move according to the real-time motion amplitude of the structure, and then carrying out wave absorption on the to-be-absorbed wave point. By means of the control method, efficient wave absorption of waves at the single-point position can be achieved, structural parameter mapping radiation wave relation construction is achieved so as to achieve the active wave absorption function, and the control method has important significance in the field of wave absorption and disaster reduction.

Description

Technical Field

[0001] The present invention belongs to the field of wave dissipation and disaster reduction, and particularly relates to a method for realizing active control of a fixed-point wave dissipation of a structure based on machine learning. Background Art

[0002] The ocean occupies seven-tenths of the earth's surface area and contains a large amount of resources. With the increasing demand for social resources, the development and utilization of ocean resources have gradually become a key area. At present, the design of offshore structures often uses the test results of two-dimensional cross-section flume experiments to evaluate the reliability of the structure and the feasibility of the selected scheme.

[0003] Among them, limited by the flume size, when the incident wave generated at the wave-making end of the flume propagates in the flume, even if there is a wave dissipation facility / equipment arranged at the end, a reflected wave will still inevitably be generated, and this phenomenon will affect the long-term high-fidelity test results.

[0004] On the other hand, at present, breakwaters are mainly divided into two categories: fixed breakwaters and floating breakwaters. The construction and maintenance costs of fixed breakwaters are high, and they are not applicable to deep-sea areas and are difficult to adapt to the influence caused by water level changes. Floating breakwaters are applicable to the deep and far sea, but due to drifting with the waves, the wave dissipation effect is poor. Summary of the Invention

[0005] In view of the deficiencies of the prior art, the present invention proposes an active control method for a structure for fixed-point wave dissipation in a wave flume. This method constructs a parameter mapping relationship between the immersion depth, motion amplitude, and target point distance of the structure and the wave height at the target point through a machine learning method, combines the advantages of time prediction, predicts the motion amplitude of the structure, and controls the motion of the structure according to the predicted motion amplitude to generate the required radiation wave, thereby eliminating the waves at the wave dissipation point to be eliminated.

[0006] The technical solution adopted by the present invention is as follows:

[0007] I. An active control method for a structure for fixed-point wave dissipation in a wave flume

[0008] The active control method for the structure includes the following steps:

[0009] S1) Obtain the time-series immersion depth and time-series motion amplitude of the characteristic points of the structure, obtain the time-series radiation wave height on the first side at different measurement points on the first side of the structure, and the time-series radiation wave height on the second side at different measurement points on the second side opposite to the first side, and use the obtained data to construct a data set.

[0010] In the step S1), at each time step, the immersion depth at the start time of the time step, the motion amplitude within the time step, the measurement point distance of each measurement point, and the radiation wave height at the measurement point are constructed into a set of sample data.

[0011] S2) Construct a machine learning model and train the machine learning model using the data set obtained in step S1).

[0012] In the said step S2), the machine learning model maps the following relationships:

[0013] ① The relationship between the measurement point distance, the motion amplitude within the time step, the immersion depth at the starting moment of the time step, and the radiation wave height at the measurement point on the first side of the structure (i.e., the first-side radiation wave height);

[0014] ② The relationship between the measurement point distance, the motion amplitude within each time step, the immersion depth at the starting moment of the time step, and the radiation wave height at the measurement point on the second side of the structure (i.e., the second-side radiation wave height).

[0015] S3) Obtain the distance between the wave-dissipating point to be obtained and the characteristic point of the structure as the wave-dissipating point distance to be obtained, obtain the initial immersion depth of the characteristic point of the structure and the initial wave height of the wave-dissipating point to be obtained, use the trained machine learning model to predict the motion amplitude of the structure at the first time step according to the collected wave-dissipating point distance to be obtained, the initial immersion depth of the characteristic point of the structure, and the initial wave height of the wave-dissipating point to be obtained, and control the structure to start moving according to the predicted motion amplitude.

[0016] The said step S3) is specifically as follows:

[0017] S3.1) Generate an incident wave in the wave flume, and make the incident wave incident from the first side; the incident wave generates an incident wave and a reflected wave on the first side of the structure, and the propagation directions of the reflected wave and the incident wave are opposite; the incident wave generates a transmitted wave on the second side of the structure, and the propagation directions of the transmitted wave and the incident wave are the same;

[0018] S3.2) Obtain the distance between the wave-dissipating point to be obtained and the structure, as well as the initial immersion depth of the characteristic point of the structure, use the distance between the wave-dissipating point to be obtained and the structure as the wave-dissipating point distance to be obtained, and collect the wave height at the wave-dissipating point to be obtained at the initial moment as the initial wave height.

[0019] In specific implementation, the process of obtaining the initial wave height in the said step S3.2) can be replaced by the following steps: determine whether the wave-dissipating point to be obtained is located on the first side or the second side of the structure. If the wave-dissipating point to be obtained is located on the first side of the structure, use the goda two-point method to separate the incident wave height and the reflected wave height in the wave field, and set the sum of the incident wave height and the reflected wave height at the initial moment as the initial wave height; if the wave-dissipating point to be obtained is located on the second side of the structure, collect the transmitted wave height at the wave-dissipating point to be obtained, and set the transmitted wave height at the initial moment as the initial wave height.

[0020] S4) During the wave breaking process, the wave height of the wave point to be broken and the immersion depth of the characteristic point of the structure are collected in real time to obtain the real-time wave height of the wave point to be broken and the real-time immersion depth of the characteristic point of the structure. The trained machine learning model is used to predict the motion amplitude of the structure according to the real-time wave height of the wave point to be broken, the real-time immersion depth of the characteristic point of the structure and the distance of the wave point to be broken. The motion of the structure is controlled according to the predicted motion amplitude, and then the wave is broken at the wave point to be broken.

[0021] Specifically, in step S4), the trained machine learning model is used to predict the motion amplitude of the structure in the next time step according to the distance of the wave point to be eliminated, the immersion depth at the starting time of the current time step, and the wave height of the wave point to be eliminated at the end time of the current time step, and the motion of the structure in the next time step is controlled according to the motion amplitude of the structure in the next time step.

[0022] Furthermore, in the steps S3 and S4), the pre-trained wave height prediction model can be used to process the initial wave height of the wave point to be eliminated and the wave height of the wave point to be eliminated at the end of the current time step to obtain the predicted wave height after considering the time delay, and then the predicted wave height after considering the time delay is input into the trained machine learning model. Specifically, the wave height prediction model is a prediction model about the relationship between wave height and time.

[0023] Furthermore, the structure active control method may also include the following process: accumulating the time series data collected during the wave absorption process of step S4), constructing a data set based on the time series data, and using the data set to optimize the trained machine learning model.

[0024] 2. An active control system for structures used for wave breaking at fixed points in a wave tank

[0025] The structure active control system comprises:

[0026] The structure is arranged in the water body and is used to eliminate the waves at the point to be eliminated by generating radiation waves.

[0027] The motor is connected to the structure and is used to drive the structure to move.

[0028] Several wave height meters are arranged at the wave point to be dissipated, and are used to collect the initial wave height and real-time wave height of the wave point to be dissipated. The initial wave height is the wave height at the wave point to be dissipated at the initial moment, the transmitted wave height, or the incident wave height and reflected wave height separated by the Goda two-point method.

[0029] The training module is used to construct a data set through theoretical solution, experimental measurement or numerical simulation, train a machine learning model using the data set, and use the trained machine learning model to predict the real-time motion amplitude of the structure according to the real-time wave height of the wave elimination point to be received from the active wave elimination control module, the real-time immersion depth of the characteristic point of the structure, and the distance of the wave elimination point to be received, and output it. Specifically, when the training module constructs a data set through experimental measurement, the active control system of the structure further includes a plurality of wave height meters respectively arranged at each measurement point along the propagation direction of the radiation wave, and these wave height meters are used to collect the first-side sequential radiation wave height and the second-side sequential radiation wave height; the training module includes a data acquisition unit and a machine learning unit; the input end of the data acquisition unit is respectively communicatively connected to the output end of the wave height meters respectively arranged at each measurement point and the motor, the output end of the data acquisition unit is communicatively connected to the input end of the machine learning unit, and the machine learning unit is coupled to the active wave elimination control module.

[0030] The active wave elimination control module is mutually coupled with the training module, and its output end is also connected to the control end of the motor, and is used to obtain the real-time wave height of the wave elimination point to be received and the real-time immersion depth of the characteristic point of the structure, input the real-time wave height of the wave elimination point to be received, the real-time immersion depth of the characteristic point of the structure, and the distance of the wave elimination point to be received into the trained machine learning model, receive the real-time motion amplitude of the structure from the trained machine learning model, and control the motion of the structure by controlling the motor according to the real-time motion amplitude of the structure.

[0031] The beneficial effects of the present invention are as follows:

[0032] 1. The present invention eliminates waves at any position point in the two-dimensional cross-section flume by actively controlling the motion amplitude of the structure in the next time step.

[0033] 2. The method of the present invention can achieve long-time and high-precision wave generation for the wave front wave elimination point of the structure.

[0034] 3. The method of the present invention can eliminate waves for the wave back wave elimination point of the structure, and as an actively controlled breakwater structure, it can protect the facilities at the target point.

[0035] 4. The method of the present invention is applicable to application scenarios considering the influence of time processing delay phenomenon on control during the implementation process of the method, making the control decision more convenient for practical engineering applications such as the control decision of efficient wave elimination of breakwaters. Description of the Drawings

[0036] Figure 1 It is a schematic diagram of the active control method and system of the structure in the present invention;

[0037] Figure 2 It is a schematic diagram of the training module in the present invention;

[0038] Figure 3 Schematic diagram of the active wave attenuation control module in the present invention;

[0039] Figure 4 Schematic diagram of the motion form of the structure in the present invention. Detailed implementation manners

[0040] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0041] The first aspect of the present invention provides an active control method for a structure for fixed-point wave attenuation in a wave flume. The active control method for the structure of the present invention includes the following steps:

[0042] S1) Obtain the time-series immersion depth and time-series motion amplitude of the characteristic points of the structure during the motion process, obtain the time-series radiation wave heights at different measurement points on the first side of the structure, and the time-series radiation wave heights at different measurement points on the second side opposite to the first side, obtain the distances between each measurement point and the structure as their respective measurement point distances, and construct a data set using the obtained measurement point distances and time-series data.

[0043] Among them, the motion amplitude represents the displacement distance within a time step. As Figure 4 shown, for swaying and heaving, the motion amplitude is expressed as the motion displacement of the object; for rolling, the motion amplitude is expressed as the rocking angle of the object.

[0044] In specific implementation, the characteristic points of the structure are selected according to the usage requirements. For example, for regular or irregular objects, their corners or centers of gravity can be selected as the characteristic points of the structure for convenient measurement.

[0045] Among them, the acquisition method of the above-mentioned time series data includes but is not limited to theoretical solution, experimental measurement or numerical simulation. The process of obtaining time series data by experimental measurement is specifically as follows: first, the motor is connected to the structure by transmission, the structure is arranged in the water body in the wave tank, and the initial immersion depth of the characteristic point of the structure is measured. Then, the time series immersion depth and time series motion amplitude of the characteristic point of the structure during the movement are obtained by combining the preset movement form, preset movement mode, preset movement process and initial immersion depth of the structure. Subsequently, several wave height meters are respectively installed at different preset measuring point positions on the first side and the second side of the structure, and the distance between each measuring point and the structure is used as the measuring point distance corresponding to the measuring point. The structure is driven to move by the motor, and the time series radiation wave height data at the respective measuring points are collected by each wave height meter, and the time series radiation wave height data collected by each wave height meter arranged on the first side is used as the first side time series radiation wave height, and the time series radiation wave height data collected by each wave height meter arranged on the second side is used as the second side time series radiation wave height. In the specific implementation, the number and location of wave height meters or wave collection points for collecting training data should be selected according to the needs. It should be noted that all wave height meters are arranged on the same straight line, which is parallel to the line connecting the two sides of the structure and the propagation direction of the radiated wave.

[0046] In step S1), the process of constructing a data set using the acquired measuring point distance and time series data is specifically as follows: constructing a set of sample data corresponding to the measuring point by the immersion depth at the start of the time step, the motion amplitude within the time step, the measuring point distance of each measuring point and the radiation wave height at the measuring point under each time step, and combining the sample data corresponding to all measuring points under each time step into a data set. Among them, the structure completes a movement in each time step, the immersion depth at the start of the time step refers to the immersion depth of the characteristic point of the structure under water before the start of this movement, and the radiation wave height at the measuring point is the wave height data collected at the measuring point within the preset wave response period after the end of this movement.

[0047] Furthermore, each set of sample data may also include the length of the time step.

[0048] S2) constructing a machine learning model and training the machine learning model using the data set obtained in step S1).

[0049] The machine learning model maps the following relationship:

[0050] ① For the measuring point on the first side of the structure, the machine learning model maps the relationship between the measuring point distance, the motion amplitude in each time step, the immersion depth at the start of the time step, and the radiation wave height at the measuring point on the first side of the structure (i.e., the radiation wave height on the first side); this mapping relationship is set according to the following formula:

[0051] a l :xb , L, d - η l

[0052] Wherein, a l represents the mapping relation formula of the first side, x b represents the motion amplitude within one time step, L represents the measuring point distance, d represents the immersion depth at the starting moment of the time step, and η l represents the radiation wave height of the first side of the structure.

[0053] ②. For the measuring points on the second side of the structure, the machine learning model maps the relationship between the measuring point distance, the motion amplitude within each time step, the immersion depth at the starting moment of the time step, and the radiation wave height (i.e., the second - side radiation wave height) at the measuring points on the second side of the structure. This mapping relation formula is set according to the following formula:

[0054] a r : x b , L, d - η r

[0055] Wherein, a r represents the mapping relation formula of the second side, x b represents the motion amplitude within one time step, L represents the measuring point distance, d represents the immersion depth at the starting moment of the time step, and η r represents the radiation wave height of the second side of the structure.

[0056] The machine learning model includes an input layer, a hidden layer, and an output layer. In the machine learning model, the parameters of the input layer should be determined by the parameters that can be represented and collected in step S1). As an optional implementation manner of the present invention, in each group of sample data, the measuring point distance of the same measuring point, the immersion depth at the starting moment of the same time step, and the motion amplitude within the same time step are used as inputs, and the radiation wave height within the same time step at the same measuring point is used as the output. As another optional implementation manner of the present invention, in each group of sample data, the measuring point distance of the same measuring point, the immersion depth at the starting moment of the same time step, and the radiation wave height at the same measuring point within the same time step are used as inputs, and the motion amplitude within the same time step is used as the output. In the machine learning model, the number of hidden layers and the number of parameters should be selected under the condition of the best efficiency to ensure a suitable learning architecture for fast and accurate analysis and prediction.

[0057] S3) Obtain the distance between the wave - dissipating point to be processed and the structure as the wave - dissipating point distance to be processed, obtain the initial immersion depth of the characteristic point of the structure and the initial wave height at the wave - dissipating point to be processed, use the trained machine learning model to predict the motion amplitude of the structure in the first time step according to the collected wave - dissipating point distance, the initial immersion depth of the characteristic point of the structure, and the initial wave height at the wave - dissipating point to be processed, and control the structure to start moving according to the predicted motion amplitude, and the wave - dissipation starts.

[0058] Step S3) is specifically as follows:

[0059] S3.1) Generate an incident wave, i.e., the wave to be eliminated, in the wave flume, and make the incident wave incident from the first side of the structure and propagate to the second side;

[0060] In the present invention, the propagation direction of the radiation wave in step S1) is parallel to the propagation direction of the incident wave (the wave to be eliminated) in step S3). Taking the propagation direction of the incident wave as the x-axis, the horizontal direction perpendicular to the propagation direction of the incident wave as the y-axis, and the vertical direction as the z-axis to establish a coordinate system, then the first side and the second side in step S1) are the negative and positive directions of the x-axis respectively; the measuring point distance in step S1) and the wave elimination point distance in step S3) are both relative distances on the x-axis, and the y-axis coordinates do not need to be considered. The water depth in the wave flume, the shape and position of the structure are all the same as those in step 1);

[0061] S3.2) Measure the distance between the wave elimination point and the structure and the initial immersion depth of the characteristic point of the structure. Take the distance between the wave elimination point and the structure as the wave elimination point distance. Keep the structure stationary, and use a wave height meter to collect the wave height at the wave elimination point at the initial moment as the initial wave height.

[0062] Further, the process of collecting the wave height at the wave elimination point at the initial moment as the initial wave height in step S3.2) can be replaced by the following steps: Determine whether the wave elimination point is located on the first side or the second side of the structure; if the wave elimination point is located on the first side of the structure, use the goda two-point method to separate the incident wave height and the reflected wave height in the wave field, and set the sum of the incident wave height and the reflected wave height at the wave elimination point at the initial moment as the initial wave height; if the wave elimination point is located on the second side of the structure, use a wave height meter to collect the transmitted wave height at the wave elimination point, and set the transmitted wave height at the initial moment as the initial wave height.

[0063] S4) During the wave elimination process, real-time collect the wave height of the wave elimination point and the immersion depth of the characteristic point of the structure to obtain the real-time wave height of the wave elimination point and the real-time immersion depth of the characteristic point of the structure. Use the trained machine learning model to predict the real-time motion amplitude of the structure according to the real-time wave height of the wave elimination point, the real-time immersion depth of the characteristic point of the structure and the wave elimination point distance, and control the motion of the structure according to the real-time motion amplitude of the structure, so as to eliminate the wave at the wave elimination point.

[0064] In step S4), the process of using the trained machine learning model to predict the real-time motion amplitude of the structure based on the real-time wave height of the wave dissipation point to be processed, the real-time immersion depth of the characteristic point of the structure, and the distance of the wave dissipation point to be processed is specifically as follows: Using the trained machine learning model, predict the motion amplitude of the structure within the next time step based on the distance of the wave dissipation point to be processed, the immersion depth of the structure at the starting moment of the current time step, and the real-time wave height of the wave dissipation point to be processed at the end moment of the current time step, and control the motion of the structure in the next time step according to the motion amplitude of the structure within the next time step.

[0065] Further, the active control method for this structure further includes step S5):

[0066] S5) Accumulate the time-series data collected during the wave dissipation process in step S4). When the number of data reaches the preset quantity, use the time-series data collected in step S4) to construct a data set containing all relevant variables and use the data set to optimize the trained machine learning model.

[0067] Among the time-series data collected in step S4), each piece of data includes the distance of the wave dissipation point to be processed, the motion amplitude within the same time step, the immersion depth at the starting moment of the same time step, and the radiation wave height at the end moment of the same time step. Among them, the radiation wave height at the end moment of each time step is the difference between the initial wave height and the wave height at the end moment.

[0068] In steps S1) and S3) - S4), the special-shaped structure should maintain a consistent shape, and the shape of the structure includes but is not limited to horizontal plates, vertical plates, wedges, etc., and the specific shape is selected according to actual needs. The motion form of the special-shaped structure should be consistent, that is, the translation motion direction or the fulcrum of the rotational motion should be consistent. The motion can adopt a single degree of freedom or a form of multi-degree of freedom coupling, and the specific motion form is selected according to actual needs. The motion process of the special-shaped structure should be consistent, including but not limited to sinusoidal motion and other motion processes that can confirm the motion amplitude of the controlled object at all times according to the motion curve.

[0069] Further, when the model speed and the motor response speed are slow and the delay time is large, the influence of the delay time needs to be considered. To solve this problem, the method of the present invention introduces the delay time and compensates for the delay time to meet the requirements of closed-loop control. The delay time is the time period from reading the wave height at the target point to the final motion of the motor control board. The specific process is as follows:

[0070] After step S2), after controlling the structure to move one step, the time-series curves of the radiation waves changing with time at each measurement point on both sides of the structure are collected. Each time-series curve is used as a set of training data, and a wave height prediction model is trained through Deep Reinforcement Learning (DRL) for the relationship between wave height and time, obtaining a trained wave height prediction model.

[0071] The trained wave height prediction model is used to predict the wave height value of waves, especially irregular waves, after a delay time. The input of the trained wave height prediction model is the wave height value collected at the current moment t and the delay time △t, and the output is the predicted wave height value considering the time delay at the moment t + △t after the delay time.

[0072] In step S3), the initial wave height of the wave point to be eliminated is processed using the pre-trained wave height prediction model to obtain the predicted initial wave height after the delay time △t, and then the predicted initial wave height is input into the trained machine learning model. The trained machine learning model predicts the motion amplitude of the structure in the first time step according to the distance of the wave point to be eliminated, the initial immersion depth of the characteristic point of the structure, and the predicted initial wave height of the wave point to be eliminated.

[0073] In step S4), the wave height at the end moment of the current time step of the wave point to be eliminated is processed using the pre-trained wave height prediction model to obtain the predicted wave height considering the time delay after the delay time △t, and then the predicted wave height considering the time delay is input into the trained machine learning model. The trained machine learning model predicts the motion amplitude of the structure in the next time step according to the distance of the wave point to be eliminated, the immersion depth at the start moment of the current time step, and the predicted wave height considering the time delay at the end moment of the current time step of the wave point to be eliminated.

[0074] The second aspect of the present invention provides an active control system for a structure for fixed-point wave elimination in a wave flume. The active control system for the structure of the present invention includes:

[0075] A structure arranged in the water body, which is used to generate radiation waves in the wave flume through its own movement and use the radiation waves to eliminate the waves at the wave points to be eliminated. Specifically, the structure is a horizontal plate, a vertical plate or a wedge. The motion mode of the structure is at least one of vertical translation motion, horizontal translation motion and rotational motion; the motion form of the structure is single-degree-of-freedom or multi-degree-of-freedom coupling. When the motion form of the structure is single-degree-of-freedom, one of the above motion modes is adopted; when the motion form of the structure is multi-degree-of-freedom coupling, a combination of at least two of the above motion modes is adopted. The specific motion mode and motion form can be selected according to actual needs. The motion curve of the structure is a sine motion. Specifically, the wave flume can be a two-dimensional numerical or physical flume.

[0076] The motor has an output end that is transmission-connected to the structure and is used to drive the structure to move.

[0077] The training module is used to construct a data set through theoretical solution, experimental measurement or numerical simulation, use the data set to train the machine learning model, use the trained machine learning model to predict the real-time motion amplitude of the structure according to the real-time wave height of the wave point to be eliminated, the real-time immersion depth of the characteristic point of the structure and the distance to the wave point to be eliminated, and output the real-time motion amplitude of the structure.

[0078] Several wave height meters are arranged at the wave point to be eliminated, and are used to collect the initial wave height and real-time wave height of the wave point to be eliminated. The initial wave height is the wave height at the wave point to be eliminated at the initial moment, the transmitted wave height, or the incident wave height and the reflected wave height separated by the Goda two-point method. When the training module constructs a data set through experimental measurement, it also includes several wave height meters arranged at each measuring point on the first side and the second side of the structure respectively along the propagation direction of the radiation wave, and are used to collect the time-series radiation wave height of the first side and the time-series radiation wave height of the second side. The output ends of these wave height meters arranged at each measuring point are all connected to the training module in communication, so as to transmit the collected time-series wave height to the training module.

[0079] The active wave-breaking control module is coupled with the training module to realize the rapid transmission of information, thereby realizing the real-time control and correction of the movement of the structure; the input end is also connected to the output end of the wave height meter arranged at the wave-breaking point, and the output end is also connected to the control end of the motor. The active wave-breaking control module is used to obtain the real-time wave height of the wave-breaking point and the real-time immersion depth of the characteristic point of the structure, input the real-time wave height of the wave-breaking point, the real-time immersion depth of the characteristic point of the structure and the distance of the wave-breaking point into the trained machine learning model, receive the real-time movement amplitude of the structure from the trained machine learning model, and control the movement of the structure by controlling the motor according to the real-time movement amplitude of the structure.

[0080] like Figure 2 As shown, when the training module constructs a data set through experimental measurement, the training module includes a data acquisition unit and a machine learning unit. The input end of the data acquisition unit is respectively connected to the output end of the wave height meter and the motor arranged at each measuring point, the output end of the data acquisition unit is connected to the input end of the machine learning unit, and the machine learning unit is coupled to the active wave-breaking control module. The coupling connection between the machine learning unit and the active wave-breaking control module means that the machine learning unit receives the time series data collected during the wave-breaking process from the active wave-breaking control module, and outputs the predicted motion amplitude of the structure in the next time step to the active wave-breaking control module.

[0081] In the training module, the data acquisition unit is used to receive the time-series radiation wave height of the first side from the wave height meter arranged on the first side, receive the time-series radiation wave height of the second side from the wave height meter arranged on the second side, obtain the time-series motion amplitude and time-series immersion depth of the structure from the motor, and output the time-series radiation wave height of the first side, the time-series radiation wave height of the second side, the time-series motion amplitude and time-series immersion depth of the structure to the machine learning unit.

[0082] In the training module, the machine learning unit is used to construct a data set using data received from the data acquisition unit or the active wave-breaking control module, and use the data set to train or optimize the machine learning model in the machine learning unit; and use the trained machine learning model to predict the motion amplitude of the structure in the next time step based on the distance of the wave-breaking point to be received from the active wave-breaking control module, the immersion depth at the starting time of the current time step, and the real-time wave height of the wave-breaking point to be received at the end time of the current time step, and output it to the active wave-breaking control module.

[0083] The technical principle and technical effect of the present invention are further described below in conjunction with specific embodiments:

[0084] like Figure 1 As shown in the module 1, in this embodiment, a horizontal plate is used as the structure, and the distance between the center of the structure and the water surface is the initial immersion depth of the structure. The motion mode of the structure is vertical translation motion, and the displacement in the vertical direction within each time step is the motion amplitude x within the time step. b A motor is arranged above the structure, and the motor is in driving connection with the structure to drive the structure to move. The delay time is not considered in this embodiment.

[0085] like Figure 2 As shown, the motion process of the structure is sinusoidal motion, which can be expressed as:

[0086] x b =A×sin(ω b ×t)

[0087] In the formula, x b represents the motion amplitude of the structure in one time step, A represents the maximum distance that the structure deviates from its equilibrium position, ω b It represents the angular frequency of the structure in one time step, and t represents the length of one time step.

[0088] In this embodiment, the first side and the second side of the structure correspond to Figure 1 When the structure starts to move, the first side radiation wave η is generated on the left side. l , generating a second side radiation wave η on the right r. When the shape of the structure is 30° wedge, 50° wedge and circular arc, the curves of the first-side time-series radiation wave height collected at the wave height meter W3 are as Figure 1 shown in the training curves in

[0089] Figure 1 . Module 1 in Figure 1 and Figure 2 is a training module. As shown in l1 and l3 , the training module includes a data acquisition unit (upper) and a machine learning unit (lower). The input ends of the data acquisition unit are respectively communicatively connected to the outputs of the wave height meters (W1, W2, W3) installed at different preset measuring point positions on the left side of the structure, the wave height meters (W4, W5, W6) installed at different preset measuring point positions on the right side of the structure, and the motor. The data acquisition unit receives the first-side time-series radiation wave height (η r1 ~η r3 ) from the wave height meters arranged on the left side (W1, W2, W3), receives the second-side time-series radiation wave height (η 1 ~η 6 ) from the wave height meters arranged on the right side (W4, W5, W6), and receives the time-series motion amplitude and time-series immersion depth of the structure from the motor. The data acquisition unit also receives the measuring point distances (L 1 ~L 6 ) of each measuring point. The data acquisition unit outputs the measuring point distances (L 1 ~L 6 ), the first-side time-series radiation wave height (η l1 ~η l3 ), the second-side time-series radiation wave height (η r1 ~η r3 ), the time-series motion amplitude and time-series immersion depth of the structure to the machine learning unit.

[0090] The machine learning unit uses the received data to construct a data set, uses the data set to train or optimize a machine learning model, and then obtains the parameter mapping relationship between the immersion depth, motion amplitude of the structure, the distance of the target point and the wave height η t of the target point:

[0091] a l : x b , L, d - η l

[0092] a r : x b , L, d - η r

[0093] In the formula, al represents the mapping relation formula on the left side, a rThe mapping relation on the right side is shown. $x_b$ represents the motion amplitude within a time step, $L$ represents the distance to the target point, $d$ represents the immersion depth at the start of the time step, and $\eta$ l represents the radiation wave height on the left side of the structure, and $\eta$ r represents the radiation wave height on the right side of the structure.

[0094] Figure 1 Module 2 in Figure 3 is the active wave dissipation control module. The principle of the active wave dissipation control module to control the motion of the structure according to the real-time wave height at the wave dissipation point to be processed is further described below in conjunction with

[0095] In this embodiment, an incident wave $\eta$ i is generated from the left side of the structure, and the incident wave propagates from left to right (i.e., from the first side to the second side). The incident wave generates an incident wave $\eta$ i and a reflected wave $\eta$ rr on the left side (wave front) of the structure, and a transmitted wave $\eta$ t on the right side (wave back).

[0096] When the wave dissipation point to be processed is located on the left side (the first side) of the structure, before wave dissipation, two wave gauges are arranged on the left side of the structure, and the incident wave and the reflected wave at the wave dissipation point to be processed are separated using the Goda two-point method. During the wave dissipation process, the wave height at the left wave dissipation point to be processed is:

[0097] $\eta$ t1 $=$ $\eta$ i $-$ $\eta$ rr $-$ $\eta$ l

[0098] In the formula, $\eta$ t1 represents the wave height at the left wave dissipation point to be processed, $\eta$ i represents the incident wave height at the left wave dissipation point to be processed, $\eta$ rr represents the reflected wave height at the left wave dissipation point to be processed, and $\eta$ l represents the radiation wave height at the left wave dissipation point to be processed.

[0099] For wave dissipation at the left wave dissipation point to be processed, that is, to make the wave height $\eta$ t1 at the left wave dissipation point to be processed $= 0$. At this time:

[0100] $\eta$ l $=$ $\eta$ i $-$ $\eta$ rr

[0101] In the formula, $\eta$ l represents the radiation wave height at the left wave dissipation point to be processed, $\eta$ i represents the incident wave height at the left wave dissipation point to be processed, and $\eta$ rr represents the reflected wave height at the left wave dissipation point to be processed.

[0102] When the wave dissipation point to be dissipated is on the right side (the second side) of the structure, before wave dissipation, a wave gauge is arranged at the wave dissipation point to be dissipated to obtain the transmitted wave at the wave dissipation point to be dissipated. During the wave dissipation process, the wave height at the right wave dissipation point to be dissipated is:

[0103] η t2 = η t + η r

[0104] In the formula, η t2 represents the wave height at the right wave dissipation point to be dissipated, η t represents the transmitted wave height at the wave dissipation point to be dissipated, and η r represents the radiated wave height at the right wave dissipation point to be dissipated.

[0105] Wave dissipation is performed at the right wave dissipation point to be dissipated, that is, the wave height η t2 at the right wave dissipation point to be dissipated is made to be 0. At this time:

[0106] η r = -η t

[0107] In the formula, η t represents the transmitted wave height at the right wave dissipation point to be dissipated, and η r represents the radiated wave height at the right wave dissipation point to be dissipated.

[0108] Taking the wave dissipation at the left wave dissipation point to be dissipated as an example, the output results of not using the structure for wave dissipation, using a horizontal plate for wave dissipation with a fixed motion process, and using the system of the present invention for active wave dissipation are as Figure 1 shown in the output results. It can be seen that during the long-term wave dissipation process, the system of the present invention can achieve a more accurate active wave dissipation effect.

[0109] The above specific embodiments are used to explain and illustrate the present invention, rather than to limit the present invention. Any modifications and changes made to the present invention within the spirit and scope of the claims of the present invention fall within the protection scope of the present invention.

Claims

1. A method for actively controlling a structure for fixed-point wave breaking in a wave tank, characterized in that: The structure active control method comprises the following steps: S1) obtaining the time-series immersion depth and time-series motion amplitude of the characteristic points of the structure, obtaining the time-series radiation wave height of the first side at different measuring points on the first side of the structure, and the time-series radiation wave height of the second side at different measuring points on the second side opposite to the first side, and constructing a data set; S2) constructing a machine learning model, and training the machine learning model using the data set obtained in step S1); S3) obtaining the distance between the wave point to be eliminated and the characteristic point of the structure as the distance of the wave point to be eliminated, using the trained machine learning model to predict the motion amplitude of the structure in the first time step according to the distance of the wave point to be eliminated, the initial immersion depth of the characteristic point of the structure and the initial wave height of the wave point to be eliminated, and controlling the structure to start moving according to the predicted motion amplitude; S4) During the wave breaking process, the trained machine learning model is used to predict the motion amplitude of the structure according to the real-time wave height of the wave breaking point, the real-time immersion depth of the characteristic point of the structure and the distance of the wave breaking point. The motion of the structure is controlled according to the predicted motion amplitude, and then the wave is broken at the wave breaking point.

2. The active control method for a structure for fixed-point wave breaking in a wave tank according to claim 1, characterized in that: In the step S1), the immersion depth at the start of each time step, the motion amplitude within the time step, the measuring point distance of each measuring point and the radiation wave height at the measuring point are constructed as a set of sample data.

3. The active control method for a structure for fixed-point wave breaking in a wave tank according to claim 2, characterized in that: In step S2), the machine learning model maps the following relationship: The relationship between the distance between the measuring points, the amplitude of the motion within the time step, the immersion depth at the start of the time step and the radiated wave height on the first side; and, the relationship between the distance between the measuring points, the amplitude of motion within each time step, the immersion depth at the start of the time step, and the radiated wave height on the second side.

4. The active control method for a structure for fixed-point wave breaking in a wave tank according to claim 1, characterized in that: The step S3) is specifically as follows: S3.1) generating an incident wave in a wave tank so that the incident wave is incident from a first side; S3.2) Obtain the distance between the wave point to be eliminated and the structure, and the initial immersion depth of the characteristic point of the structure, take the distance between the wave point to be eliminated and the structure as the distance of the wave point to be eliminated, and collect the wave height of the wave point to be eliminated at the initial moment as the initial wave height.

5. The active control method for a structure for fixed-point wave breaking in a wave tank according to claim 4, characterized in that: In step S3.2), the process of obtaining the initial wave height is replaced by the following steps: It is determined whether the wave point to be eliminated is located on the first side or the second side of the structure. If the wave point to be eliminated is located on the first side of the structure, the Goda two-point method is used to separate the incident wave height and the reflected wave height, and the sum of the incident wave height and the reflected wave height at the initial moment is set as the initial wave height; if the wave point to be eliminated is located on the second side of the structure, the transmitted wave height at the wave point to be eliminated is collected, and the transmitted wave height at the initial moment is set as the initial wave height.

6. The active control method for a structure for fixed-point wave breaking in a wave tank according to claim 1, characterized in that: In the step S4), the trained machine learning model is used to predict the movement amplitude of the structure in the next time step according to the distance of the wave point to be eliminated, the immersion depth at the starting time of the current time step, and the wave height of the wave point to be eliminated at the end time of the current time step, and the movement of the structure in the next time step is controlled according to the movement amplitude of the structure in the next time step.

7. The active control method for a structure for fixed-point wave breaking in a wave tank according to claim 6, characterized in that: In step S3) and step S4), the pre-trained wave height prediction model is used to process the initial wave height of the wave point to be eliminated and the wave height of the wave point to be eliminated at the end of the current time step, so as to obtain the predicted wave height after considering the time delay, and then input it into the trained machine learning model; the wave height prediction model is a prediction model about the relationship between wave height and time.

8. The active control method for a structure for fixed-point wave breaking in a wave tank according to claim 1, characterized in that: The active control method for the structure also includes: accumulating the time series data collected during the wave absorption process of step S4), constructing a data set according to the time series data, and using the data set to optimize the trained machine learning model.

9. An active control system for a structure applied to the active control method for a structure as claimed in any one of claims 1 to 8, characterized in that: The structure active control system comprises: A structure is arranged in the water body and is used to eliminate waves at a point to be eliminated by generating radiation waves; A motor is connected to the structure and is used to drive the structure to move; Several wave height meters are arranged at the wave-to-break point to collect the initial wave height and real-time wave height of the wave-to-break point; The training module is used to construct a data set through theoretical solution, experimental measurement or numerical simulation, use the data set to train the machine learning model, and use the trained machine learning model to predict the real-time motion amplitude of the structure and output it; The active wave-breaking control module is coupled with the training module, and the output end is also connected to the control end of the motor, so as to obtain the real-time wave height of the wave-breaking point and the real-time immersion depth of the characteristic point of the structure, input the real-time wave height of the wave-breaking point, the real-time immersion depth of the characteristic point of the structure and the distance of the wave-breaking point into the trained machine learning model, receive the real-time motion amplitude of the structure from the trained machine learning model, and control the motion of the structure by controlling the motor according to the real-time motion amplitude of the structure.

10. The active control system for structures used for fixed-point wave breaking in a wave tank according to claim 9, characterized in that: When the training module constructs a data set through experimental measurement, the active control system of the structure also includes a plurality of wave height meters arranged at each measuring point along the propagation direction of the radiation wave; the training module includes a data acquisition unit and a machine learning unit; the input end of the data acquisition unit is respectively communicatively connected to the output end and the motor of the wave height meter arranged at each measuring point, the output end of the data acquisition unit is communicatively connected to the input end of the machine learning unit, and the machine learning unit is coupled to the active wave absorption control module.

Images