Body installation type real-time sea condition identification method for wave power generation device

Abstract

The invention discloses a wave energy power generation device body installation type real-time sea condition identification method. The method comprises the following steps: a data acquisition step: a pressure wave meter measures pressure change information in real time; the attitude sensor measures attitude information in real time; the gravity type wave measuring buoy measures the real-time effective wave height; a model establishment step: establishing a sea condition neural network model by a data processing unit, and establishing a mapping relationship between the pressure change information and the attitude information at each time and the significant wave height at the same time by the data processing unit; a model using step: inputting pressure change information and attitude information into the sea condition neural network model in real time so as to automatically output significant wave height; and attitude control: the device control system automatically controls the attitude of the wave-absorbing floater according to the output effective wave height. By adopting the above arrangement, the defect that the effective wave height is not accurately measured by a pressure wave meter due to the influence of high sea conditions is overcome, the effective wave height can be accurately measured under different sea conditions, and automatic control of the attitude of the wave absorbing floater is facilitated.

Description

Technical Field

[0001] This invention relates to the field of ocean wave energy utilization technology, and in particular to a real-time sea state identification method for wave energy power generation device mounted on the device body, providing data support for the adaptive attitude adjustment of wave-absorbing floats. Background Technology

[0002] The ocean possesses abundant wave energy reserves, with very rich exploitable resources. The development of ocean wave energy power generation equipment can provide clean and renewable energy for users such as coastal cities, remote islands and reefs, offshore facilities, and marine instruments. In recent years, the rapid development of offshore wind power, fully utilizing ocean energy, combining wind and wave power, reducing costs and increasing efficiency, has also brought new opportunities for the industrialization of wave energy power generation equipment.

[0003] Waves are characterized by their fluctuating, intermittent, and random nature, making wave measurement instruments a persistent challenge in the industry. These instruments are generally expensive, difficult to acquire data from, and prone to loss. Currently, wave measurement methods mainly include several types: line-of-sight wavemeters, wave rods, pressure wavemeters, acoustic wavemeters, gravity wavemeters, and remote sensing wavemeters. Line-of-sight wavemeters are shore-based instruments; wave rods can only be used for fixed-point observations, and both line-of-sight wavemeters and wave rods have relatively low data accuracy; pressure wavemeters cannot accurately measure short-period waves; acoustic wavemeters are installed underwater, making data reading inconvenient; gravity wavemeters are easily lost; and remote sensing wavemeters utilize remote sensing images and are suitable for large-area wave measurement.

[0004] Since wave-absorbing floats generally do not require attitude adjustment in low sea states, the inaccuracy of short-period wave measurements when the pressure wave meter is installed on the main body of the wave energy generation device can be ignored. However, in high sea states, the longitudinal and transverse movements of the device body greatly affect its measurement accuracy. Therefore, the wave measurement method with the main body of the wave energy generation device still needs to be solved in order to achieve automatic attitude control of the wave-absorbing float. Summary of the Invention

[0005] In order to overcome the shortcomings of the prior art, the purpose of this invention is to provide a real-time sea state identification method for wave energy power generation device body installation, which overcomes the problem that the measurement of effective wave height by pressure wave meter is inaccurate due to the influence of high sea state, and can accurately measure the effective wave height in real time under different sea states, thereby facilitating the automatic attitude control of wave-absorbing float.

[0006] The objective of this invention is achieved through the following technical solution:

[0007] A real-time sea state identification method for wave energy power generation device mounted on the device body includes data acquisition, model building, model use and attitude control steps.

[0008] The data acquisition steps are as follows: a pressure wave meter is installed on the underwater part of the wave energy power generation device to measure the pressure change information caused by the effective wave height in real time; an attitude sensor is installed in the wave energy power generation device to measure the attitude information of the wave energy power generation device in real time; and the effective wave height in real time is measured using a gravity wave meter buoy.

[0009] The model establishment steps are as follows: the data processing unit establishes a sea state neural network model based on the pressure change information, attitude information and significant wave height obtained in the data acquisition steps; the data processing unit establishes a mapping relationship between the pressure change information and attitude information at each time and the significant wave height at the same time based on the sea state neural network model.

[0010] The model is used in the following steps: Based on the established sea state neural network model, pressure change information and attitude information are input into the sea state neural network model in real time to automatically output the effective wave height in real time.

[0011] The attitude control steps are as follows: The device control system automatically controls the attitude of the wave-absorbing buoy based on the effective wave height output in real time by the sea state neural network model.

[0012] Furthermore, in the data acquisition step, the attitude information measured by the attitude sensor includes: direction, angle, angular velocity, rotational velocity angle, acceleration, and rotational acceleration.

[0013] Furthermore, in the model building step, during the process of building the sea state neural network model in the data processing unit, the following method is adopted: multiple input layers and output layers are built in the sea state neural network model in chronological order. Pressure change information and attitude information at the same time are input into the input layer, and the corrected effective wave height at the corresponding time is input into the output layer, so that the pressure change information and attitude information at the same time are mapped with the corresponding effective wave height.

[0014] Furthermore, in the model building step, after the data processing unit builds the sea state neural network model, it also includes data preprocessing, model training, model verification, and model solidification steps.

[0015] The data preprocessing involves selecting pressure change information, attitude information, and significant wave height over a period of time as verification samples, matching the pressure change information, attitude information, and significant wave height according to time, and dividing the verification samples into a first part and a second part in chronological order.

[0016] The model training involves training the sea state neural network model using the validation samples from the previous part.

[0017] The model validation: The sea state neural network model is validated using the latter part of the validation samples;

[0018] The model is solidified: After the sea state neural network model has been trained and verified, it is solidified.

[0019] Furthermore, in the model verification process, when using the latter part of the verification samples to verify the sea state neural network model, the following method is adopted: the pressure change information and attitude information in the latter part of the verification samples are input into the input layer of the sea state neural network model in chronological order, so that the output layer of the sea state neural network model automatically outputs the corresponding significant wave height, and compares it with the actual significant wave height to determine whether the accuracy of the significant wave height output by the sea state neural network model reaches 85%. If so, the trained sea state neural network model is considered reliable; if not, the model training, model verification, and model solidification steps are repeated until the trained sea state neural network model is considered reliable.

[0020] Furthermore, in the model building step, a classification layer is built after the output layer in the sea state neural network model. The classification layer divides the sea state level into low sea state, medium sea state, high sea state, and extreme sea state based on the significant wave height. Therefore, in the model usage step, when the sea state neural network model automatically outputs the significant wave height, it simultaneously obtains the sea state level through the classification layer and outputs the result.

[0021] Furthermore, in the classification layer, the classification layer includes a wave height increasing module and a wave height decreasing module. The wave height increasing module is used to detect sea states where the significant wave height gradually increases, and the wave height decreasing module is used to detect sea states where the significant wave height gradually decreases. The classification layer determines the position and position change of the significant wave height in the wave height increasing module and the wave height decreasing module in chronological order to obtain the real-time changes in sea state.

[0022] Furthermore, the wave height gradually increasing module has critical wave heights for low-to-medium sea states, medium-to-high sea states, and high extreme sea states; the wave height gradually decreasing module has critical wave heights for extreme high sea states, medium-to-high sea states, and medium-to-low sea states; the critical wave heights for medium-to-low sea states are: < critical wave heights for low-to-medium sea states < critical wave heights for medium-to-high sea states < critical wave heights for medium-to-high sea states < critical wave heights for extreme high sea states < critical wave heights for high extreme sea states. When the significant wave height gradually increases, and the significant wave height is greater than the critical wave heights for low-to-medium sea states, medium-to-high sea states, and high extreme sea states, respectively, the sea state level sequentially changes from low sea states to medium sea states, high sea states, and extreme sea states; conversely, when the significant wave height continuously decreases, and the significant wave height is less than the critical wave heights for medium-to-low sea states, medium-to-high sea states, and extreme high sea states, respectively, the sea state level sequentially changes from extreme sea states to high sea states, medium sea states, and low sea states.

[0023] Furthermore, in the attitude control step, after the sea state neural network model outputs the corresponding sea state level based on the significant wave height, the device control system automatically controls the attitude of the wave-absorbing buoy based on the acquired sea state level.

[0024] The present invention has the following beneficial effects:

[0025] This invention addresses the error inherent in existing technologies that use pressure wave meters to measure significant wave height. It addresses this issue by implementing a real-time sea state identification method. This method comprises four main steps: data acquisition, model building, model usage, and attitude control. The data acquisition step collects relevant data required for the sea state neural network model, primarily pressure change information, attitude information, and significant wave height. The model building step establishes a sea state neural network model, mapping the pressure change and attitude information to the significant wave height. In the model usage step, the real-time collected pressure change and attitude information is input into the sea state neural network model during the operation of the wave energy generator, automatically outputting the significant wave height in real time. Finally, in the attitude control step, the device control system automatically controls the attitude of the wave-absorbing buoy based on the real-time significant wave height. Therefore, the wave energy generator of this invention can measure significant wave height in real time using the sea state neural network model and can resist the influence of low or high sea states, thereby improving the accuracy of significant wave height measurement. This facilitates the device control system's automatic attitude control of the wave-absorbing buoy based on the real-time significant wave height, thus improving the accuracy of real-time wave energy monitoring. Attached Figure Description

[0026] Figure 1 The present invention relates to a real-time sea state level identification system mounted on the wave energy power generation device body.

[0027] Figure 2 This is a schematic diagram of the real-time sea state classification method based on significant wave height according to the present invention. Detailed Implementation

[0028] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Terms such as “upper,” “inner,” “middle,” “left,” “right,” and “one” used in this specification are merely for clarity of description and are not intended to limit the scope of the invention. Changes or adjustments to their relative relationships, without substantially altering the technical content, should also be considered within the scope of the invention.

[0029] A method for real-time sea state identification mounted on the wave energy power generation device, such as... Figure 1As shown, the main system adopted is a real-time sea state level identification system installed on the wave energy power generation device body. This real-time sea state level identification system mainly includes the wave energy power generation device body, a pressure wave meter, an attitude sensor, and a data processing unit. The pressure wave meter is installed in the underwater part of the wave energy power generation device body, and the attitude sensor is installed in the wave energy power generation device body. The pressure wave meter and the attitude sensor are electrically connected to the data processing unit through wired or wireless means.

[0030] The real-time sea state identification method of the present invention includes data acquisition, model building, model use, and attitude control steps.

[0031] Data acquisition steps: Install a pressure wave meter on the underwater part of the wave energy generator body to measure the pressure change information caused by the effective wave height in real time; install an attitude sensor on the wave energy generator body to measure the attitude information of the wave energy generator body in real time, including: direction, angle, angular velocity, rotational speed angle, acceleration and rotational acceleration; use a gravity wave meter buoy to measure the real-time effective wave height.

[0032] Model building steps: The data processing unit builds a sea state neural network model based on the pressure change information, attitude information and significant wave height obtained from the data acquisition steps. Based on the sea state neural network model, the data processing unit establishes the mapping relationship between the pressure change information and attitude information at each time and the significant wave height at the same time.

[0033] Model usage steps: Based on the established sea state neural network model, input pressure change information and attitude information into the sea state neural network model in real time to automatically output the significant wave height in real time.

[0034] Attitude control steps: The device control system automatically controls the attitude of the wave-absorbing buoy based on the effective wave height output in real time by the sea state neural network model.

[0035] To provide a detailed explanation of the model building process in the model connection step of this embodiment:

[0036] In the model building process, the following method is used in the data processing unit to build the sea state neural network model: Multiple input and output layers are constructed in chronological order within the sea state neural network model. Pressure change and attitude information at the same time are input into the input layers, and the corrected significant wave height at the corresponding time is input into the output layers, thus mapping the pressure change and attitude information at the same time to the corresponding significant wave height. Therefore, by constructing input and output layers in the sea state neural network model, the pressure change and attitude information at the same time are mapped to the corresponding significant wave height. Thus, even when the wave energy photoelectric device is in actual use and a gravity buoy is not available to accurately measure the significant wave height, the corresponding significant wave height can still be calculated using this sea state neural network model. Similarly, when pressure change and attitude information is missing, the missing pressure change and / or attitude information can still be calculated using this neural network model. This demonstrates that by establishing a sea state neural network model, the required data can be obtained more effectively based on the mapping relationship between data, resulting in high work efficiency.

[0037] In the model building process, after the data processing unit builds the sea state neural network model, it also includes data preprocessing, model training, model validation, and model solidification. The following describes each step after model building:

[0038] Data preprocessing: Pressure change information, attitude information, and significant wave height over a period of time are selected as validation samples. The pressure change information, attitude information, and significant wave height are matched according to time. At the same time, the validation samples are divided into a first part and a second part according to time order, with the first part accounting for 80% of all validation samples and the second part accounting for 20%.

[0039] Model Training: The sea state neural network model is trained using the validation samples from the previous part. During the training process, the pressure change information and attitude information from the validation samples are input into the input layer of the sea state neural network model in chronological order. Simultaneously, the significant wave height from the validation samples is input into the output layer of the sea state neural network model in chronological order, so that the pressure change information and attitude information and the significant wave height have a mapping relationship at the same point in time. Therefore, it is convenient to obtain the missing or detected information through the model in the future.

[0040] Model Validation: The sea state neural network model is validated using the latter part of the validation samples. During model validation, the following method is employed: the pressure change and attitude information from the latter part of the validation samples are input into the input layer of the sea state neural network model in chronological order. This causes the output layer of the model to automatically output the corresponding significant wave height, which is then compared with the actual significant wave height. The accuracy of the output significant wave height is determined to be at least 85%. If so, the trained sea state neural network model is considered reliable. If not, the model training, validation, and solidification steps are repeated until the trained model is deemed reliable. Therefore, model validation can effectively determine whether the established model is accurate and reliable in actual use, thereby improving the accuracy of the established model.

[0041] Model solidification: After the sea state neural network model has been trained and validated, it is solidified to reduce interference or errors in subsequent use.

[0042] To improve real-time sea state identification and accurately determine and output corresponding sea state information based on real-time significant wave height, thereby enhancing real-time monitoring quality, this invention includes a classification layer after the output layer in the sea state neural network model during model building. This classification layer categorizes sea state levels into low, medium, high, and extreme sea states (e.g., high wave height) based on significant wave height. Figure 2 (As shown); Therefore, in the model usage steps, when the sea state neural network model automatically outputs the significant wave height, it simultaneously obtains the sea state level through the classification layer and outputs the result, so as to realize the function of accurately judging the real-time sea state through the real-time significant wave height and improving the monitoring quality.

[0043] The classification layer includes a wave height increasing module and a wave height decreasing module. The wave height increasing module detects sea states where the significant wave height gradually increases, while the wave height decreasing module detects sea states where the significant wave height gradually decreases. The classification layer sequentially determines the position and positional changes of the significant wave height within the wave height increasing and decreasing modules to obtain real-time sea state changes. Therefore, by setting up wave height increasing and decreasing modules in the classification layer to determine whether the wave height is gradually increasing or decreasing, the specific sea state conditions can be efficiently identified.

[0044] The wave height increment module has a critical wave height for low to medium sea states (defined as H). Im Critical wave height for medium and high sea states (defined as H) mh ) and the critical wave height for high extreme sea states (defined as H he The wave height descending module has a critical wave height for extreme high sea states (defined as H). ehCritical wave height for high sea state (defined as H) hm ) and the critical wave height for medium and low sea states (defined as H mI Critical wave height for medium and low sea states (H) mI < Critical wave height for low to medium sea states (H) Im < High sea state critical wave height (H) hm < Critical wave height for medium to high sea states (H) mh ) < Critical wave height for extreme high sea states (H) eh < Critical wave height for high extreme sea states (H) he The wave height parameter can be set via the data processing unit or manually. In practice, as the significant wave height gradually increases, and when it exceeds the critical wave heights for low-to-medium sea states, medium-to-high sea states, and high-extreme sea states, the sea state level sequentially progresses from low to medium, high, and extreme sea states. Conversely, as the significant wave height decreases, and when it falls below the critical wave heights for medium-to-low, medium-to-high, and extreme high sea states, the sea state level sequentially progresses from extreme to high, medium, and low sea states.

[0045] In the attitude control step, after the sea state neural network model outputs the corresponding sea state level based on the significant wave height, the device control system automatically controls the attitude of the wave-absorbing buoy based on the acquired sea state level. Compared with adjusting the attitude of the wave-absorbing buoy based on a single significant wave height and its mapped attitude and pressure change information, this invention can divide the sea state level into four levels. Therefore, when the corresponding sea state level is obtained, the attitude of the wave-absorbing buoy can be automatically adjusted, thereby avoiding repeated adjustments to the attitude of the wave-absorbing buoy, which may increase other motion disturbances or instability, thus improving operational stability.

[0046] In summary, this invention addresses the error inherent in measuring significant wave height using pressure wave meters in existing technologies. It addresses this issue by implementing a real-time sea state identification method, which comprises four main steps: data acquisition, model building, model usage, and attitude control. The data acquisition step collects relevant data required for the sea state neural network model, primarily pressure change information, attitude information, and significant wave height. The model building step establishes a sea state neural network model, mapping pressure change and attitude information to significant wave height. In the model usage step, the real-time collected pressure change and attitude information is input into the sea state neural network model during the use of the wave energy generator, automatically outputting the significant wave height in real time. Finally, in the attitude control step, the device control system automatically controls the attitude of the wave-absorbing buoy based on the real-time significant wave height. Therefore, the wave energy generator of this invention can measure significant wave height in real time using a sea state neural network model and can resist the influence of low or high sea states, thereby improving the accuracy of significant wave height measurement. This facilitates the device control system's automatic control of the wave-absorbing buoy's attitude based on the real-time significant wave height, thus improving the accuracy of real-time wave energy monitoring.

[0047] The embodiments of the present invention are not limited thereto. Based on the above description of the present invention, and using common technical knowledge and conventional means in the field, the present invention can be modified, replaced or combined in various other forms without departing from the basic technical idea of ​​the present invention, and all such modifications, replacements or combinations fall within the scope of protection of the present invention.

Claims

1. A method for real-time sea state identification mounted on the main body of a wave energy power generation device, characterized in that, This includes data acquisition, model building, model usage, and attitude control steps; The data acquisition steps are as follows: a pressure wave meter is installed on the underwater part of the wave energy power generation device to measure the pressure change information caused by the effective wave height in real time; an attitude sensor is installed in the wave energy power generation device to measure the attitude information of the wave energy power generation device in real time; and the effective wave height in real time is measured using a gravity wave meter buoy. The model establishment steps are as follows: the data processing unit establishes a sea state neural network model based on the pressure change information, attitude information and significant wave height obtained in the data acquisition steps; the data processing unit establishes a mapping relationship between the pressure change information and attitude information at each time and the significant wave height at the same time based on the sea state neural network model. The model is used in the following steps: Based on the established sea state neural network model, pressure change information and attitude information are input into the sea state neural network model in real time to automatically output the effective wave height in real time; The attitude control steps are as follows: The device control system automatically controls the attitude of the wave-absorbing buoy based on the effective wave height output in real time by the sea state neural network model.

2. The wave energy power generation device-mounted real-time sea state identification method as described in claim 1, characterized in that, In the data acquisition step, the attitude information measured by the attitude sensor includes: direction, angle, angular velocity, rotational velocity angle, acceleration, and rotational acceleration.

3. The wave energy power generation device-mounted real-time sea state identification method as described in claim 1, characterized in that, In the model building step, the following method is used in the process of building the sea state neural network model in the data processing unit: multiple input layers and output layers are built in the sea state neural network model in chronological order. Pressure change information and attitude information at the same time are input into the input layer, and the corrected effective wave height at the corresponding time is input into the output layer, so that the pressure change information and attitude information at the same time are mapped with the corresponding effective wave height.

4. The wave energy power generation device body-mounted real-time sea state identification method as described in claim 3, characterized in that, In the model building step, after the data processing unit builds the sea state neural network model, it also includes data preprocessing, model training, model verification and model solidification steps. The data preprocessing involves selecting pressure change information, attitude information, and significant wave height over a period of time as verification samples, matching the pressure change information, attitude information, and significant wave height according to time, and dividing the verification samples into a first part and a second part in chronological order. The model training involves training the sea state neural network model using the validation samples from the previous part. The model validation: The sea state neural network model is validated using the latter part of the validation samples; The model is solidified: After the sea state neural network model has been trained and verified, it is solidified.

5. The wave energy power generation device body-mounted real-time sea state identification method as described in claim 4, characterized in that, In the model validation process, when using the latter part of the validation samples to validate the sea state neural network model, the following method is adopted: the pressure change information and attitude information in the latter part of the validation samples are input into the input layer of the sea state neural network model in chronological order, so that the output layer of the sea state neural network model automatically outputs the corresponding significant wave height, and compares it with the actual significant wave height to determine whether the accuracy of the significant wave height output by the sea state neural network model reaches 85%. If so, the trained sea state neural network model is considered reliable; if not, the model training, model validation and model solidification steps are repeated until the trained sea state neural network model is considered reliable.

6. The wave energy power generation device body-mounted real-time sea state identification method as described in claim 3, characterized in that, In the model building step, a classification layer is built after the output layer in the sea state neural network model. The classification layer divides the sea state level into low sea state, medium sea state, high sea state and extreme sea state based on the significant wave height. Therefore, in the model usage step, when the sea state neural network model automatically outputs the significant wave height, it simultaneously obtains the sea state level through the classification layer and outputs the result.

7. The wave energy power generation device body-mounted real-time sea state identification method as described in claim 6, characterized in that, In the classification layer, the classification layer includes a wave height increasing module and a wave height decreasing module. The wave height increasing module is used to detect sea states where the significant wave height gradually increases, and the wave height decreasing module is used to detect sea states where the significant wave height gradually decreases. The classification layer determines the position and position change of the significant wave height in the wave height increasing module and the wave height decreasing module in chronological order to obtain the real-time changes in sea state.

8. The wave energy power generation device body-mounted real-time sea state identification method as described in claim 7, characterized in that, The wave height gradually increasing module has critical wave heights for low-to-medium sea states, medium-to-high sea states, and high extreme sea states; the wave height gradually decreasing module has critical wave heights for extreme high sea states, medium-to-high sea states, and medium-to-low sea states. The critical wave heights for medium-to-low sea states are: < critical wave heights for low-to-medium sea states < critical wave heights for medium-to-high sea states < critical wave heights for medium-to-high sea states < critical wave heights for extreme high sea states < critical wave heights for high extreme sea states. When the significant wave height gradually increases, and is greater than the critical wave heights for low-to-medium, medium-to-high, and high extreme sea states respectively, the sea state level sequentially progresses from low sea states to medium, high, and extreme sea states. Conversely, when the significant wave height continuously decreases, and is less than the critical wave heights for medium-to-low, medium-to-high, and extreme high sea states respectively, the sea state level sequentially progresses from extreme sea states to high, medium, and low sea states.

9. The wave energy power generation device body-mounted real-time sea state identification method as described in claim 6, characterized in that, In the attitude control step, after the sea state neural network model outputs the corresponding sea state level based on the significant wave height, the device control system automatically controls the attitude of the wave-absorbing buoy based on the acquired sea state level.

Images