CN117404258B - Composite cylindrical foundation offshore construction cloud monitoring system - Google Patents
Composite cylindrical foundation offshore construction cloud monitoring system Download PDFInfo
- Publication number
- CN117404258B CN117404258B CN202311711262.4A CN202311711262A CN117404258B CN 117404258 B CN117404258 B CN 117404258B CN 202311711262 A CN202311711262 A CN 202311711262A CN 117404258 B CN117404258 B CN 117404258B
- Authority
- CN
- China
- Prior art keywords
- wind speed
- training
- construction
- tower
- sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000010276 construction Methods 0.000 title claims abstract description 117
- 239000002131 composite material Substances 0.000 title claims abstract description 21
- 238000012544 monitoring process Methods 0.000 title claims abstract description 17
- 238000012549 training Methods 0.000 claims abstract description 174
- 238000010801 machine learning Methods 0.000 claims abstract description 70
- 238000003062 neural network model Methods 0.000 claims abstract description 38
- 238000009434 installation Methods 0.000 claims abstract description 18
- 238000004364 calculation method Methods 0.000 claims abstract description 17
- 238000013480 data collection Methods 0.000 claims abstract description 9
- 238000000034 method Methods 0.000 claims description 18
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011900 installation process Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D13/00—Assembly, mounting or commissioning of wind motors; Arrangements specially adapted for transporting wind motor components
- F03D13/20—Arrangements for mounting or supporting wind motors; Masts or towers for wind motors
- F03D13/25—Arrangements for mounting or supporting wind motors; Masts or towers for wind motors specially adapted for offshore installation
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/04—Ageing analysis or optimisation against ageing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Sustainable Development (AREA)
- Sustainable Energy (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Mechanical Engineering (AREA)
- Geometry (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Wind Motors (AREA)
Abstract
The invention discloses a composite barrel-type foundation offshore construction cloud monitoring system, which relates to the technical field of tower installation, wherein construction information for installing a tower to be installed is collected in advance through a construction information collection module, wind speed proportion training data, life prediction training data and supporting force training data are collected through a history data collection module, a first neural network model for evaluating wind speed proportion, a first machine learning model for predicting the service life of a blade and a second machine learning model for predicting a supporting label are trained by a prediction model training module, a construction inclination angle calculation module is arranged, and the inclination angle of the installation of the tower is selected based on the construction information and the first neural network model, the first machine learning model and the second machine learning model; before the tower is installed, the inclination angle of the installation of the tower is planned, so that the total flow of the corresponding fan in the life cycle of the blade is improved.
Description
Technical Field
The invention belongs to the technical field of offshore tower construction, and particularly relates to a composite cylindrical foundation offshore construction cloud monitoring system.
Background
Wind power generation is one of important forms of clean energy, and has wide development prospect. In offshore wind power projects, the tower is one of the key structures supporting the wind turbines and blades. The tower is generally arranged on a composite cylinder type foundation, and the composite cylinder type foundation is arranged at the bottom of the sea for fixing; therefore, the tower installation process is performed on a composite cylinder basis; the inclination angle of the tower barrel when being installed on the basis of the composite barrel has important influence on the performance of the fan and the service life of the blades.
The prior arrangement of the installation inclination angle of the tower is often carried out by experience of engineers, but quantitative analysis is lacking;
therefore, the invention provides a composite cylindrical foundation offshore construction cloud monitoring system.
Disclosure of Invention
The present invention aims to solve at least one of the technical problems existing in the prior art. Therefore, the invention provides the composite barrel type foundation offshore construction cloud monitoring system, and the inclination angle of the installation of the tower barrel is planned before the installation of the tower barrel, so that the total flow of the corresponding fan in the life cycle of the blade is improved.
To achieve the above object, an embodiment according to a first aspect of the present invention provides a composite cylindrical foundation offshore construction cloud monitoring system, including a construction information collection module, a historical data collection module, a prediction model training module, and a construction inclination calculation module; wherein, each module is electrically connected;
the construction information collection module is mainly used for collecting construction information for installing the tower to be installed in advance;
specifically, the construction information includes equipment data and wind data;
the equipment data comprises tower length, tower weight, fan weight and blade area;
the wind power data comprise wind speed sequences and wind direction sequences of different heights at the position of a seabed foundation of a tower to be installed;
the different heights are obtained by presetting the acquisition basic height and the acquisition height step length; the wind speed sequence and the wind direction sequence are collected from the basic height, the height is increased in sequence to collect the height step length, and the wind speed sequence and the wind direction sequence are collected again;
the wind speed sequence is a sequence in which collected real-time wind speeds are arranged in time sequence, and the wind direction sequence is a sequence in which collected angles of wind are arranged in time sequence; the angle of the wind is the inclination angle of the wind direction and the horizontal plane;
the construction information collection module sends the collected construction information to the construction inclination calculation module;
the historical data collection module is mainly used for collecting wind speed proportion training data, life prediction training data and supporting force training data;
the wind speed proportion training data, the life prediction training data and the supporting force training data are obtained in real time from other installed fans in the same sea area; representing other installed fans in the same sea area as training fans;
the wind speed proportion training data comprise wind speed proportion training sequences of all training fans;
the wind speed proportion training data comprise average wind speed sequences of each training fan at different heights;
wherein the life prediction training data comprises life characteristics of all training fan blades and blade life;
the life characteristic is blade average pressure; the average pressure is calculated in the following way:
the serial number of each unit time is marked as T, and the total duration of wind speed acquisition is marked as T; the method comprises the steps of marking the wind speed of a t unit time of the height of a fan as vt, marking the angle between a fan blade and the wind direction as theta, marking the area of the blade as s, and marking the air density as rho; the calculation formula of the average pressure P of the blade is:the method comprises the steps of carrying out a first treatment on the surface of the The service life of the blade is the time from installation to abnormal operation of the blade;
the supporting force training data comprise supporting feature sets and supporting labels of all training fan blades;
wherein the support feature set comprises blade average pressure, tower weight, fan weight, and tower inclination;
the support label is one of 0 or 1; wherein a support tag of 0 indicates that the tower is not in an inclined state; the support label is 1 to indicate that the tower barrel is inclined; the inclination state is that the difference value between the inclination angle of the tower barrel of the training fan and the inclination angle of the training fan when the training fan is installed is larger than a preset inclination angle threshold value after the training fan is operated;
the historical data collection module sends wind speed proportion training data, life prediction training data and supporting force training data to the prediction model training module;
the prediction model training module is mainly used for training a first neural network model for estimating the wind speed proportion based on wind speed proportion training data, training a first machine learning model for predicting the service life of the blade based on service life prediction training data and training a second machine learning model for predicting a support label based on support force training data;
the method for training the first neural network model for evaluating the wind speed proportion based on the wind speed proportion training data comprises the following steps:
presetting a first prediction time step, a first sliding step and a first sliding window length according to practical experience, converting a wind speed proportion training sequence of each training fan into a plurality of first training samples by using a sliding window method, taking each group of first training samples as input of a first neural network model, taking a wind speed proportion training sequence of a future first prediction time step as output of the first neural network model, taking a wind speed proportion training sequence of a first prediction time step in a follow-up first prediction time step of each first training sample as a prediction target, and training the first neural network model; generating a first neural network model of a predicted wind speed ratio; the first neural network model is an RNN neural network model;
the method for training the first machine learning model for predicting the service life of the blade based on the service life prediction training data comprises the following steps:
taking the life characteristics of each training fan blade as input of a first machine learning model, wherein the first machine learning model takes the predicted blade life of each life characteristic as output, takes the blade life corresponding to the life characteristic in a life prediction training data set as a prediction target, and takes the sum of prediction errors of all life characteristics as a training target; training the first machine learning model until the sum of the prediction errors reaches convergence, and stopping training to train the first machine learning model for evaluating the service life of the blade; the first machine learning model is any one of a polynomial regression model or an SVM model;
the second machine learning model for predicting the support label is trained based on the support force training data in the following manner:
taking a support feature set of each training fan blade as input of a second machine learning model, wherein the second machine learning model takes a predicted support label of each support feature set as output, takes a support label corresponding to the support feature set in a support force training data set as a predicted target, and takes the sum of prediction errors of all the support feature sets as a training target; training the second machine learning model until the sum of the prediction errors reaches convergence, and stopping training to train the second machine learning model of the prediction support label; the second machine learning model is any one of a polynomial regression model or an SVM model;
the prediction model training module sends the first neural network model, the first machine learning model and the second machine learning model to the construction dip angle calculation module;
the construction inclination angle calculation module is mainly used for selecting the inclination angle of tower installation based on construction information, a first neural network model, a first machine learning model and a second machine learning model;
selecting the inclination angle of the tower installation comprises the following steps:
step one: the length of the tower barrel is marked as L, the acquisition basic height is marked as B, and the acquisition height step length is marked as c;
calculating the wind speed average value of the wind speed sequence of each height in the construction information, and sequencing all the wind speed average values from low to high to obtain a construction average wind speed sequence;
step two: taking the last sequence of the first sliding window length of the construction average wind speed sequence as the input of a first neural network model to obtain the prediction of the average wind speed of the subsequent first prediction time step; adding the predicted average wind speed of the subsequent first predicted time step to the construction average wind speed sequence, and repeatedly executing the second step until the number of elements in the construction average wind speed sequence reaches；
Step three: randomly selecting the inclination angle of a tower barrel to be installed; marking the inclination angle of the tower to be installed as x, and reading the first wind speed sequence from the construction average wind speed sequence when the installed height L1 of the tower is L x cosxThe average wind speed is divided by the first average wind speed in the construction average wind speed sequence to be used as the construction wind speed proportion;
step four: multiplying each element in the wind speed sequence corresponding to the acquired foundation height by the construction wind speed proportion to obtain a wind speed sequence born by the blade after the tower barrel is installed, and expressing the wind speed sequence as a construction wind speed sequence;
step five: calculating a flow angle corresponding to each wind direction in the wind direction sequence based on the inclination angle x of the tower and the wind direction sequence in the construction information and based on geometric common knowledge, so as to obtain a flow angle sequence; the flow angle is the angle between the wind direction and the blade;
step six: calculating the average wind speed of a construction wind speed sequence, taking the average wind speed as the construction average wind speed, calculating the average pressure of the blades corresponding to the tower barrel to be installed based on the blade area and the construction average wind speed in the construction information, and marking the average pressure as the construction average pressure;
step seven: inputting the average pressure of the tower to be installed, the weight of the tower, the weight of the fan and the inclination angle x of the tower into a second machine learning model to obtain a predicted support label; if the support label is 1, re-executing the third to seventh steps; if the support label is 0, executing the step eight;
step eight: inputting construction average pressure into a first machine learning model to obtain predicted blade life, and marking the predicted life as M;
step nine: calculating the flow F in unit time which can be obtained by the blade corresponding to the tower to be installed; the flow rate per unit time is calculated by the following steps:
marking the area of the blade corresponding to the tower to be installed with a mark N;
marking the number of each element in the construction wind speed sequence as I, and marking the length of the construction wind speed sequence as I; marking the construction wind speed corresponding to the ith element as vi, and marking the ith flow angle in the flow angle sequence as ri;
flow per unit timeThe method comprises the steps of carrying out a first treatment on the surface of the Step ten: calculating the comprehensive weight Wx of the dip angle x, wherein Wx=F×M; circularly executing the third step to the tenth step until the cycle times reach a preset cycle times threshold value;
step eleven: and selecting the inclination angle with the largest comprehensive weight Wx from all the circulating inclination angles x as the installation inclination angle of the tower to be installed.
Compared with the prior art, the invention has the beneficial effects that:
according to the invention, construction information for installing a tower to be installed is collected in advance, then wind speed proportion training data, life prediction training data and supporting force training data of other installed fans in the same sea area are collected, a first neural network model for evaluating wind speed proportion is trained based on the wind speed proportion training data, a first machine learning model for predicting the life of a blade is trained based on the life prediction training data, a second machine learning model for predicting a supporting label is trained based on the supporting force training data, and a proper inclination angle for installing the tower is selected based on the construction information and the first neural network model, the first machine learning model and the second machine learning model; the optimal tower inclination angle is calculated before the tower is installed, so that the total flow of the corresponding fan in the life cycle of the blade is improved.
Drawings
Fig. 1 is a connection diagram of a module of the composite cylindrical foundation offshore construction cloud monitoring system.
Detailed Description
The technical solutions of the present invention will be clearly and completely described in connection with the embodiments, and it is obvious that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1, the composite cylindrical foundation offshore construction cloud monitoring system comprises a construction information collection module, a historical data collection module, a prediction model training module and a construction inclination calculation module; wherein, each module is electrically connected;
the construction information collection module is mainly used for collecting construction information for installing the tower to be installed in advance;
specifically, the construction information includes equipment data and wind data;
the equipment data comprises tower length, tower weight, fan weight and blade area;
it will be appreciated that the tower is mounted vertically on the seabed foundation, so that the length of the tower is related to the height of the last tower, and the wind power is different at different heights, so that the flow rate of the fan and the pressure born by the blades are different;
the weight of the tower barrel and the weight of the fan further determine the weight born by the tower barrel under different inclination angles, so that the supporting force of the tower barrel is determined;
the wind power data comprise wind speed sequences and wind direction sequences of different heights at the position of a seabed foundation of a tower to be installed;
the different heights are obtained by presetting the acquisition basic height and the acquisition height step length; the wind speed sequence and the wind direction sequence are collected from the basic height, the height is increased in sequence to collect the height step length, and the wind speed sequence and the wind direction sequence are collected again;
the wind speed sequence is a sequence in which collected real-time wind speeds are arranged in time sequence, and the wind direction sequence is a sequence in which collected angles of wind are arranged in time sequence; the angle of the wind is the inclination angle of the wind direction and the horizontal plane; the wind speed and the wind direction can be obtained in real time through a wind speed sensor and a wind direction sensor respectively;
it will be appreciated that since the tower is high, collecting wind speed and direction from the tower's height is a difficult task when the tower is not installed, and therefore wind speed and direction sequences can be collected from lower heights according to practical experience, respectively;
the construction information collection module sends the collected construction information to the construction inclination calculation module;
the historical data collection module is mainly used for collecting wind speed proportion training data, life prediction training data and supporting force training data;
the wind speed proportion training data, the life prediction training data and the supporting force training data are obtained in real time from other installed fans in the same sea area; representing other installed fans in the same sea area as training fans;
the wind speed proportion training data comprise wind speed proportion training sequences of all training fans;
in a preferred embodiment, the wind speed ratio training data comprises an average wind speed sequence at different heights for each training fan; the average proportion of the high-level wind speed to the low-level wind speed is estimated through the average wind speeds of different heights; it can be understood that the wind speed sensor and the wind direction sensor are arranged at the height position of the tower drum of the training fan, so that the average wind speed at the height of the tower drum can be obtained; the wind speed and the wind direction of the training fan are acquired based on the acquisition basic height and the acquisition height step length;
wherein the life prediction training data comprises life characteristics of all training fan blades and blade life;
in a preferred embodiment, the life characteristic is blade average pressure; the average pressure is calculated in the following way:
the serial number of each unit time is marked as T, and the total duration of wind speed acquisition is marked as T; the method comprises the steps of marking the wind speed of a t unit time of the height of a fan as vt, marking the angle between a fan blade and the wind direction as theta, marking the area of the blade as s, and marking the air density as rho; the calculation formula of the average pressure P of the blade is:it is understood that the angle between the fan blade and the wind direction can be calculated according to the inclination angle of the fan blade and the horizontal plane and the angle of the wind direction by utilizing the common geometric sense;
it will be appreciated that when wind speeds are high, the greater the pressure to which the blade is subjected, resulting in damage to the blade;
the service life of the blade is the time from installation to abnormal operation of the blade; it should be noted that, the standard that the blade cannot normally operate is defined according to actual experience by a professional;
the supporting force training data comprise supporting feature sets and supporting labels of all training fan blades;
wherein the support feature set comprises blade average pressure, tower weight, fan weight, and tower inclination;
the support label is one of 0 or 1; wherein a support tag of 0 indicates that the tower is not in an inclined state; the support label is 1 to indicate that the tower barrel is inclined; the inclination state is that the difference value between the inclination angle of the tower barrel of the training fan and the inclination angle of the training fan when the training fan is installed is larger than a preset inclination angle threshold value after the training fan is operated;
the historical data collection module sends wind speed proportion training data, life prediction training data and supporting force training data to the prediction model training module;
the prediction model training module is mainly used for training a first neural network model for estimating the wind speed proportion based on wind speed proportion training data, training a first machine learning model for predicting the service life of the blade based on service life prediction training data and training a second machine learning model for predicting a support label based on support force training data;
in a preferred embodiment, the first neural network model for estimating the wind speed ratio is trained based on the wind speed ratio training data in the following manner:
presetting a first prediction time step, a first sliding step and a first sliding window length according to practical experience, converting a wind speed proportion training sequence of each training fan into a plurality of first training samples by using a sliding window method, taking each group of first training samples as input of a first neural network model, taking a wind speed proportion training sequence of a future first prediction time step as output of the first neural network model, taking a wind speed proportion training sequence of a first prediction time step in a follow-up first prediction time step of each first training sample as a prediction target, and training the first neural network model; generating a first neural network model of a predicted wind speed ratio; preferably, the first neural network model is an RNN neural network model;
it should be noted that, the sliding window method is used as a conventional technical means of a cyclic neural network model or a time sequence prediction model, and the invention is not described in principle here; but for the purpose of facilitating the implementation of the invention, the invention provides the following examples regarding sliding window methods:
assuming we want to train a time prediction model with history data 1,2,3,4,5,6, set the prediction time step to 1, the sliding step to 1 and the sliding window length to 3; then 3 sets of training data and corresponding predicted target data are generated: [1,2,3], [2,3,4] and [3,4,5] are used as training data, and [4], [5] and [6] are respectively used as prediction targets;
the method for training the first machine learning model for predicting the service life of the blade based on the service life prediction training data comprises the following steps:
taking the life characteristics of each training fan blade as input of a first machine learning model, wherein the first machine learning model takes the predicted blade life of each life characteristic as output, takes the blade life corresponding to the life characteristic in a life prediction training data set as a prediction target, and takes the sum of prediction errors of all life characteristics as a training target; training the first machine learning model until the sum of the prediction errors reaches convergence, and stopping training to train the first machine learning model for evaluating the service life of the blade; the first machine learning model is any one of a polynomial regression model or an SVM model;
the second machine learning model for predicting the support label is trained based on the support force training data in the following manner:
taking a support feature set of each training fan blade as input of a second machine learning model, wherein the second machine learning model takes a predicted support label of each support feature set as output, takes a support label corresponding to the support feature set in a support force training data set as a predicted target, and takes the sum of prediction errors of all the support feature sets as a training target; training the second machine learning model until the sum of the prediction errors reaches convergence, and stopping training to train the second machine learning model of the prediction support label; the second machine learning model is any one of a polynomial regression model or an SVM model;
it should be noted that, the calculation formula of the prediction error is:wherein k is the number of the characteristic data, zk is the prediction error, ak is the predicted value corresponding to the k-th group of characteristic data, wk is the actual value corresponding to the k-th group of cable characteristic data; for example: for the first machine learning model, the life characteristic corresponds to characteristic data, the predicted life of the blade corresponds to a predicted value, and the life of the blade corresponding to the life characteristic is an actual value; for the second machine learning model, the support feature set corresponds to feature data, the predicted support label corresponds to a predicted value, and the support label corresponding to the support feature set is an actual value;
the prediction model training module sends the first neural network model, the first machine learning model and the second machine learning model to the construction dip angle calculation module;
the construction inclination angle calculation module is mainly used for selecting the inclination angle of tower installation based on construction information, a first neural network model, a first machine learning model and a second machine learning model;
in a preferred embodiment, selecting the inclination of the tower installation comprises the steps of:
step one: the length of the tower barrel is marked as L, the acquisition basic height is marked as B, and the acquisition height step length is marked as c;
step one: calculating the wind speed average value of the wind speed sequence of each height in the construction information, and sequencing all the wind speed average values from low to high to obtain a construction average wind speed sequence;
step two: taking the last sequence of the first sliding window length of the construction average wind speed sequence as the input of a first neural network model to obtain the prediction of the average wind speed of the subsequent first prediction time step; adding the predicted average wind speed of the subsequent first predicted time step to the construction average wind speed sequence, and repeatedly executing the second step until the number of elements in the construction average wind speed sequence reachesThe method comprises the steps of carrying out a first treatment on the surface of the It will be appreciated that each element in the sequence of construction average wind speeds represents a height, the difference between each two successive heights being B;
step three: randomly selecting the inclination angle of a tower barrel to be installed; marking the inclination angle of the tower to be installed as x, and reading the first wind speed sequence from the construction average wind speed sequence when the installed height L1 of the tower is L x cosxThe average wind speed is divided by the first average wind speed in the construction average wind speed sequence to be used as the construction wind speed proportion;
step four: multiplying each element in the wind speed sequence corresponding to the acquired foundation height by the construction wind speed proportion to obtain a wind speed sequence born by the blade after the tower barrel is installed, and expressing the wind speed sequence as a construction wind speed sequence;
step five: calculating a flow angle corresponding to each wind direction in the wind direction sequence based on the inclination angle x of the tower and the wind direction sequence in the construction information and based on geometric common knowledge, so as to obtain a flow angle sequence; the flow angle is the angle between the wind direction and the blade; it will be appreciated that the angle of the blades is obtained from the pitch angle of the tower;
step six: calculating the average wind speed of a construction wind speed sequence, taking the average wind speed as the construction average wind speed, calculating the average pressure of the blades corresponding to the tower barrel to be installed based on the blade area and the construction average wind speed in the construction information, and marking the average pressure as the construction average pressure;
step seven: inputting the average pressure of the tower to be installed, the weight of the tower, the weight of the fan and the inclination angle x of the tower into a second machine learning model to obtain a predicted support label; if the support label is 1, re-executing the third to seventh steps; if the support label is 0, executing the step eight;
step eight: inputting construction average pressure into a first machine learning model to obtain predicted blade life, and marking the predicted life as M;
step nine: calculating the flow F in unit time which can be obtained by the blade corresponding to the tower to be installed; the flow rate per unit time is calculated by the following steps:
marking the area of the blade corresponding to the tower to be installed with a mark N;
marking the number of each element in the construction wind speed sequence as I, and marking the length of the construction wind speed sequence as I; marking the construction wind speed corresponding to the ith element as vi, and marking the ith flow angle in the flow angle sequence as ri;
flow per unit timeThe method comprises the steps of carrying out a first treatment on the surface of the Step ten: calculating the comprehensive weight Wx of the dip angle x, wherein Wx=F×M; circularly executing the third step to the tenth step until the cycle times reach a preset cycle times threshold value;
step eleven: and selecting the inclination angle with the largest comprehensive weight Wx from all the circulating inclination angles x as the installation inclination angle of the tower to be installed.
The above preset parameters or preset thresholds are set by those skilled in the art according to actual conditions or are obtained by mass data simulation.
The above embodiments are only for illustrating the technical method of the present invention and not for limiting the same, and it should be understood by those skilled in the art that the technical method of the present invention may be modified or substituted without departing from the spirit and scope of the technical method of the present invention.
Claims (8)
1. The composite cylindrical foundation offshore construction cloud monitoring system is characterized by comprising a construction information collection module, a historical data collection module, a prediction model training module and a construction inclination calculation module; wherein, each module is electrically connected;
the construction information collection module is used for pre-collecting construction information for installing the tower to be installed and sending the collected construction information to the construction inclination angle calculation module;
the historical data collection module is used for collecting wind speed proportion training data, life prediction training data and supporting force training data and sending the wind speed proportion training data, the life prediction training data and the supporting force training data to the prediction model training module;
the prediction model training module trains a first neural network model for evaluating the wind speed proportion based on the wind speed proportion training data, trains a first machine learning model for predicting the service life of the blade based on the service life estimation training data and trains a second machine learning model for predicting the support label based on the support force training data, and sends the first neural network model, the first machine learning model and the second machine learning model to the construction dip angle calculation module;
and the construction inclination angle calculation module is used for selecting the inclination angle of the tower barrel installation based on the construction information, the first neural network model, the first machine learning model and the second machine learning model.
2. The composite tubular foundation offshore construction cloud monitoring system of claim 1, wherein the construction information comprises equipment data and wind data;
the equipment data comprises tower length, tower weight, fan weight and blade area;
the wind power data comprise wind speed sequences and wind direction sequences of different heights at the position of a seabed foundation of a tower to be installed;
the wind speed sequence is a sequence in which collected real-time wind speeds are arranged in time sequence, and the wind direction sequence is a sequence in which collected angles of wind are arranged in time sequence; the angle of the wind is the inclination angle of the wind direction and the horizontal plane.
3. The composite cylindrical foundation offshore construction cloud monitoring system of claim 2, wherein the wind speed ratio training data, life prediction training data and supporting force training data are obtained in real time from other installed fans in the same sea area; representing other installed fans in the same sea area as training fans;
the wind speed proportion training data comprise wind speed proportion training sequences of all training fans;
the wind speed proportion training data comprise average wind speed sequences of each training fan at different heights;
wherein the life prediction training data comprises life characteristics of all training fan blades and blade life;
the life characteristic is blade average pressure; the average pressure is calculated in the following way:
the serial number of each unit time is marked as T, and the total duration of wind speed acquisition is marked as T; the method comprises the steps of marking the wind speed of a t unit time of the height of a fan as vt, marking the angle between a fan blade and the wind direction as theta, marking the area of the blade as s, and marking the air density as rho; the calculation formula of the average pressure P of the blade is:the service life of the blade is that the blade cannot be normally used from installationThe duration of the run;
the supporting force training data comprise supporting feature sets and supporting labels of all training fan blades;
wherein the support feature set comprises blade average pressure, tower weight, fan weight, and tower inclination;
the support label is one of 0 or 1; wherein a support tag of 0 indicates that the tower is not in an inclined state; the support label is 1 to indicate that the tower barrel is inclined; the inclination state is that after the training fan operates, the difference value between the inclination angle of the tower barrel of the training fan and the inclination angle during installation is larger than a preset inclination angle threshold value.
4. A composite cylindrical foundation offshore construction cloud monitoring system according to claim 3, wherein the first neural network model for estimating the wind speed ratio is trained based on wind speed ratio training data by:
presetting a first prediction time step, a first sliding step and a first sliding window length according to practical experience, converting a wind speed proportion training sequence of each training fan into a plurality of first training samples by using a sliding window method, taking each group of first training samples as input of a first neural network model, taking a wind speed proportion training sequence of a future first prediction time step as output of the first neural network model, taking a wind speed proportion training sequence of a first prediction time step in a follow-up first prediction time step of each first training sample as a prediction target, and training the first neural network model; generating a first neural network model of a predicted wind speed ratio; the first neural network model is an RNN neural network model.
5. The composite cylindrical foundation offshore construction cloud monitoring system of claim 4, wherein the first machine learning model for predicting blade life is trained based on life prediction training data by:
taking the life characteristics of each training fan blade as input of a first machine learning model, wherein the first machine learning model takes the predicted blade life of each life characteristic as output, takes the blade life corresponding to the life characteristic in a life prediction training data set as a prediction target, and takes the sum of prediction errors of all life characteristics as a training target; training the first machine learning model until the sum of the prediction errors reaches convergence, and stopping training to train the first machine learning model for evaluating the service life of the blade; the first machine learning model is any one of a polynomial regression model or an SVM model.
6. The composite tubular foundation marine construction cloud monitoring system of claim 5, wherein the second machine learning model for training out the predictive support tag based on the support force training data is:
taking a support feature set of each training fan blade as input of a second machine learning model, wherein the second machine learning model takes a predicted support label of each support feature set as output, takes a support label corresponding to the support feature set in a support force training data set as a predicted target, and takes the sum of prediction errors of all the support feature sets as a training target; training the second machine learning model until the sum of the prediction errors reaches convergence, and stopping training to train the second machine learning model of the prediction support label; the second machine learning model is any one of a polynomial regression model or an SVM model.
7. The composite tubular foundation marine construction cloud monitoring system of claim 6, wherein selecting the inclination of the tower installation comprises the steps of:
step one: the length of the tower barrel is marked as L, the acquisition basic height is marked as B, and the acquisition height step length is marked as c;
calculating the wind speed average value of the wind speed sequence of each height in the construction information, and sequencing all the wind speed average values from low to high to obtain a construction average wind speed sequence;
step two:taking the last sequence of the first sliding window length of the construction average wind speed sequence as the input of a first neural network model to obtain the prediction of the average wind speed of the subsequent first prediction time step; adding the predicted average wind speed of the subsequent first predicted time step to the construction average wind speed sequence, and repeatedly executing the second step until the number of elements in the construction average wind speed sequence reaches；
Step three: randomly selecting the inclination angle of a tower barrel to be installed; marking the inclination angle of the tower to be installed as x, and reading the first wind speed sequence from the construction average wind speed sequence when the installed height L1 of the tower is L x cosxThe average wind speed is divided by the first average wind speed in the construction average wind speed sequence to be used as the construction wind speed proportion;
step four: multiplying each element in the wind speed sequence corresponding to the acquired foundation height by the construction wind speed proportion to obtain a wind speed sequence born by the blade after the tower barrel is installed, and expressing the wind speed sequence as a construction wind speed sequence;
step five: calculating a flow angle corresponding to each wind direction in the wind direction sequence based on the inclination angle x of the tower and the wind direction sequence in the construction information and based on geometric common knowledge, so as to obtain a flow angle sequence; the flow angle is the angle between the wind direction and the blade;
step six: calculating the average wind speed of a construction wind speed sequence, taking the average wind speed as the construction average wind speed, calculating the average pressure of the blades corresponding to the tower barrel to be installed based on the blade area and the construction average wind speed in the construction information, and marking the average pressure as the construction average pressure;
step seven: inputting the average pressure of the tower to be installed, the weight of the tower, the weight of the fan and the inclination angle x of the tower into a second machine learning model to obtain a predicted support label; if the support label is 1, re-executing the third to seventh steps; if the support label is 0, executing the step eight;
step eight: inputting construction average pressure into a first machine learning model to obtain predicted blade life, and marking the predicted life as M;
step nine: calculating the flow F in unit time which can be obtained by the blade corresponding to the tower to be installed; calculating the comprehensive weight Wx of the dip angle x, wherein Wx=F×M; circularly executing the third step to the ninth step until the cycle times reach a preset cycle times threshold value;
step ten: and selecting the inclination angle with the largest comprehensive weight Wx from all the circulating inclination angles x as the installation inclination angle of the tower to be installed.
8. The composite cylindrical foundation offshore construction cloud monitoring system of claim 7, wherein the flow per unit time is calculated by the following steps:
marking the area of the blade corresponding to the tower to be installed with a mark N;
marking the number of each element in the construction wind speed sequence as I, and marking the length of the construction wind speed sequence as I; marking the construction wind speed corresponding to the ith element as vi, and marking the ith flow angle in the flow angle sequence as ri;
flow per unit time。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311711262.4A CN117404258B (en) | 2023-12-13 | 2023-12-13 | Composite cylindrical foundation offshore construction cloud monitoring system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311711262.4A CN117404258B (en) | 2023-12-13 | 2023-12-13 | Composite cylindrical foundation offshore construction cloud monitoring system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117404258A CN117404258A (en) | 2024-01-16 |
| CN117404258B true CN117404258B (en) | 2024-03-29 |
Family
ID=89489372
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311711262.4A Active CN117404258B (en) | 2023-12-13 | 2023-12-13 | Composite cylindrical foundation offshore construction cloud monitoring system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117404258B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107246024A (en) * | 2017-06-19 | 2017-10-13 | 河北工业大学 | The control system that a kind of compound barrel-shaped foundation sinking is surely filled |
| CN112982504A (en) * | 2021-02-05 | 2021-06-18 | 水利部交通运输部国家能源局南京水利科学研究院 | Electroosmosis rectification silt sea area inclined barrel type wind power foundation device and test method |
| CN113984114A (en) * | 2021-10-18 | 2022-01-28 | 大连理工大学 | Method for diagnosing abnormality of underwater structure of ocean floating platform |
| WO2023087601A1 (en) * | 2021-11-22 | 2023-05-25 | 中交天津港湾工程研究院有限公司 | Stability prediction method for deep water thin-walled steel cylinder |
| CN116289960A (en) * | 2023-02-10 | 2023-06-23 | 浙江大学 | Marine wind torch type foundation penetration attitude control device and control method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140195159A1 (en) * | 2013-01-09 | 2014-07-10 | Iteris, Inc. | Application of artificial intelligence techniques and statistical ensembling to forecast power output of a wind energy facility |
| US20210248500A1 (en) * | 2020-02-10 | 2021-08-12 | Schlumberger Technology Corporation | Hybrid modeling process for forecasting physical system parameters |
-
2023
- 2023-12-13 CN CN202311711262.4A patent/CN117404258B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107246024A (en) * | 2017-06-19 | 2017-10-13 | 河北工业大学 | The control system that a kind of compound barrel-shaped foundation sinking is surely filled |
| CN112982504A (en) * | 2021-02-05 | 2021-06-18 | 水利部交通运输部国家能源局南京水利科学研究院 | Electroosmosis rectification silt sea area inclined barrel type wind power foundation device and test method |
| CN113984114A (en) * | 2021-10-18 | 2022-01-28 | 大连理工大学 | Method for diagnosing abnormality of underwater structure of ocean floating platform |
| WO2023087601A1 (en) * | 2021-11-22 | 2023-05-25 | 中交天津港湾工程研究院有限公司 | Stability prediction method for deep water thin-walled steel cylinder |
| CN116289960A (en) * | 2023-02-10 | 2023-06-23 | 浙江大学 | Marine wind torch type foundation penetration attitude control device and control method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117404258A (en) | 2024-01-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112233091B (en) | Wind turbine blade image damage detection and positioning method | |
| Wang et al. | SCADA data based condition monitoring of wind turbines | |
| CN110985310B (en) | Wind driven generator blade fault monitoring method and device based on acoustic sensor array | |
| CN110362045B (en) | Marine doubly-fed wind turbine generator fault discrimination method considering marine meteorological factors | |
| CN112462736B (en) | Wind turbine generator fault diagnosis method based on data analysis | |
| CN103925155B (en) | The self-adapting detecting method that a kind of Wind turbines output is abnormal | |
| CN111415070A (en) | Wind turbine generator gearbox oil temperature over-temperature fault early warning method based on SCADA data | |
| US20220178353A1 (en) | A method for estimating remaining useful life of components of an operational wind turbine | |
| EP3771822A1 (en) | A method for computer-implemented determination of a vertical speed wind profile of a wind field | |
| EP4111054A1 (en) | System and method for monitoring wind turbine rotor blades using infrared imaging and machine learning | |
| CN113374652A (en) | Method for evaluating service life of wind generating set | |
| CN112832960A (en) | Fan blade icing detection method based on deep learning and storage medium | |
| CN114254549A (en) | System and method for operating a power generating asset | |
| CN114021822A (en) | Clean energy power generation power prediction method and system | |
| CN111878322A (en) | Wind power generator device | |
| Dinwoodie et al. | Wave height forecasting to improve off-shore access and maintenance scheduling | |
| CN117404258B (en) | Composite cylindrical foundation offshore construction cloud monitoring system | |
| CN114462305A (en) | Argo buoy long-endurance energy consumption optimization and life estimation method | |
| Visbech et al. | Introducing a data-driven approach to predict site-specific leading-edge erosion from mesoscale weather simulations | |
| CN112986393A (en) | Bridge inhaul cable damage detection method and system | |
| CN110273818B (en) | Fan blade icing fault monitoring method based on shaft transformation thickness classification | |
| JP2005242803A (en) | Performance estimator, performance estimating method, and performance estimating program of machine | |
| EP3855018A1 (en) | A method for computer-implemented monitoring of a wind turbine | |
| CN113902327A (en) | Evaluation method and system for corrosion health state of offshore wind plant foundation structure | |
| CN112348244A (en) | Wind power plant generator set overhauling method based on wind resource assessment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |