CN113361149B - Surface shape adjusting method, device and equipment for active reflecting surface of astronomical table - Google Patents
Surface shape adjusting method, device and equipment for active reflecting surface of astronomical table Download PDFInfo
- Publication number
- CN113361149B CN113361149B CN202110911032.7A CN202110911032A CN113361149B CN 113361149 B CN113361149 B CN 113361149B CN 202110911032 A CN202110911032 A CN 202110911032A CN 113361149 B CN113361149 B CN 113361149B
- Authority
- CN
- China
- Prior art keywords
- elongation
- data
- node
- piston rod
- surface shape
- 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
- 238000000034 method Methods 0.000 title claims abstract description 65
- 238000012549 training Methods 0.000 claims abstract description 74
- 238000013528 artificial neural network Methods 0.000 claims abstract description 23
- 238000005259 measurement Methods 0.000 claims description 18
- 238000004590 computer program Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 12
- 238000010606 normalization Methods 0.000 claims description 11
- 230000007613 environmental effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 4
- 238000012937 correction Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000007781 pre-processing Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000013100 final test Methods 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Telescopes (AREA)
- Optical Elements Other Than Lenses (AREA)
Abstract
The application discloses a surface shape adjusting method of an active reflecting surface of an astronomical table, which comprises the following steps: acquiring historical data; generating a training sample according to the historical data; training the neural network by using the training sample to obtain an elongation prediction model; inputting the data to be measured of the target node into an elongation prediction model to obtain an elongation prediction value, wherein the target node is any node of an active reflecting surface of the astronomical phenomena table; and controlling the piston rod of the target node to stretch according to the predicted value of the elongation so as to realize surface shape adjustment. Therefore, the method predicts the elongation of the piston rod of the node by using the trained neural network, controls the extension of the piston rod according to the elongation to realize the surface shape adjustment of the active reflecting surface of the file platform, and greatly improves the real-time performance and the accuracy compared with the current surface shape adjustment scheme. In addition, the application also provides a surface shape adjusting device, equipment and a readable storage medium of the active reflecting surface of the astronomical observatory, and the technical effect of the surface shape adjusting device corresponds to that of the method.
Description
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to a method, an apparatus, a device, and a readable storage medium for adjusting a surface shape of an active reflecting surface of an astronomical observatory.
Background
The large-aperture radio telescope adopts an active reflecting surface structure, the active reflecting surface comprises thousands of nodes, and hundreds of nodes are contained in an effective observation aperture of only 300 m. The lower end of each node lower stay cable is connected with a set of mechanical-electrical-hydraulic integrated hydraulic actuator equipment, and the surface shape of the reflecting surface can be effectively controlled through adjustment of the hydraulic actuator equipment, so that the observation purpose is achieved.
In the prior art, the nodes are mainly measured one by an automatic laser total station, control quantity is generated according to a measurement result, and compensation is performed by adopting a finite element model correction method. Specifically, each set of hydraulic actuator is communicated with the upper control system, receives a control instruction of the upper control system, and transmits key information such as the position of a piston rod of the actuator, whether the piston rod is in the caliber and the like to the upper control system. The upper measuring system measures the targets arranged at the nodes through the automatic laser total station to obtain position information. The results of the upper measuring system and the upper control system are used for comprehensively determining the position target value of the piston rod of the actuator, and the position target value is corrected by adopting a finite element model correction method, so that a control command is generated and sent to the corresponding actuator.
Due to the fact that the number of the nodes is large, real-time closed-loop testing cannot be achieved through a mode that the automatic laser total station carries out measurement one by one, and under the existing technical conditions, second-level instantaneity cannot be achieved. The method for compensating by adopting the finite element model correction method increases the content of mechanical model correction, and can accelerate to a certain extent by carrying out finite element modeling prediction on parameters such as various structural element measured values, force/displacement of an intermediate structure system and the like, but the method has a plurality of influencing factors and an unsatisfactory correction effect.
In summary, how to ensure real-time performance and accuracy of the surface shape adjustment process of the active reflection surface is a problem to be solved by those skilled in the art.
Disclosure of Invention
The application aims to provide a method, a device and equipment for adjusting the surface shape of an active reflecting surface of an astronomical table and a readable storage medium, which are used for solving the problem that the real-time performance and the accuracy of the current scheme for adjusting the surface shape of the active reflecting surface of the astronomical table are not ideal. The specific scheme is as follows:
in a first aspect, the present application provides a method for adjusting a surface shape of an active reflecting surface of an astronomical observatory, comprising:
acquiring historical data; generating a training sample according to the historical data;
training a neural network by using the training sample to obtain an elongation prediction model;
inputting the data to be measured of a target node into the elongation prediction model to obtain an elongation prediction value, wherein the target node is any node of the active reflecting surface of the astronomical phenomena table;
and controlling a piston rod of the target node to stretch according to the predicted elongation value so as to realize surface shape adjustment.
Optionally, the generating a training sample according to the historical data includes:
extracting measurement data and environment data from the historical data, wherein the measurement data comprises ground anchor coordinates, node coordinates and piston rod elongation;
fitting the node positions by using the measurement data to obtain node position fitting quantity; taking the node position fitting quantity and the environment data as feature data in a training sample;
and carrying out normalization processing on the elongation of the piston rod, and taking a normalization result as label data in the training sample.
Optionally, the environmental data includes oil pressure data, temperature data, and wind speed data.
Optionally, the normalizing the elongation of the piston rod includes:
and determining the moving range of the piston rod, and performing normalization processing on the elongation of the piston rod according to the moving range.
Optionally, after generating the training sample according to the historical data, the method further includes:
and generating a noise parameter, and performing data enhancement on the training sample by using the noise parameter.
Optionally, the neural network includes a hidden layer, and the training of the neural network by using the training samples to obtain an elongation prediction model includes:
and training the neural network by using the training samples, carrying out minimum solution on the loss of the mean square error through gradient descent in the training process, determining the number of layers and the number of kernels of the hidden layers, and obtaining an elongation prediction model.
Optionally, the controlling, according to the predicted elongation value, a piston rod of the target node to stretch includes:
and generating a control instruction according to the predicted elongation value, and sending the control instruction to an actuator of the target node to control the piston rod to extend and retract.
In a second aspect, the present application provides a device for adjusting a surface shape of an active reflecting surface of an astronomical observatory, comprising:
the sample generation module is used for acquiring historical data; generating a training sample according to the historical data;
the training module is used for training the neural network by using the training sample to obtain an elongation prediction model;
the prediction module is used for inputting the data to be measured of a target node into the elongation prediction model to obtain an elongation prediction value, wherein the target node is any node of the active reflecting surface of the astronomical phenomena table;
and the control module is used for controlling the piston rod of the target node to stretch according to the predicted elongation value so as to realize surface shape adjustment.
In a third aspect, the present application provides a surface shape adjusting apparatus for an active reflecting surface of an astronomical dome, comprising:
a memory: for storing a computer program;
a processor: the computer program is used for executing the computer program to realize the surface shape adjusting method of the active reflecting surface of the astronomical table.
In a fourth aspect, the present application provides a readable storage medium, on which a computer program is stored, the computer program being executed by a processor to implement the method for adjusting the surface shape of the active reflecting surface of the astronomical table as described above.
The application provides a surface shape adjusting method of an active reflecting surface of an astronomical table, which comprises the following steps: acquiring historical data; generating a training sample according to the historical data; training the neural network by using the training sample to obtain an elongation prediction model; inputting the data to be measured of the target node into an elongation prediction model to obtain an elongation prediction value, wherein the target node is any node of an active reflecting surface of the astronomical phenomena table; and controlling the piston rod of the target node to stretch according to the predicted value of the elongation so as to realize surface shape adjustment. Therefore, the method predicts the elongation of the piston rod of the node by using the trained neural network, controls the extension of the piston rod according to the elongation to realize the surface shape adjustment of the active reflecting surface of the astronomical table, greatly improves the real-time performance and accuracy compared with the current surface shape adjustment scheme, and realizes the control optimization of the active reflecting surface of the astronomical table.
In addition, the application also provides a surface shape adjusting device, equipment and a readable storage medium for the active reflecting surface of the astronomical observatory, and the technical effect of the surface shape adjusting device corresponds to that of the method, and the description is omitted here.
Drawings
For a clearer explanation of the embodiments or technical solutions of the prior art of the present application, the drawings needed for the description of the embodiments or prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a schematic diagram of the prior art and the present application of the surface shape adjustment principle;
fig. 2 is a flowchart of a first method of an embodiment of a method for adjusting a surface shape of an active reflecting surface of an astronomical dome provided by the present application;
fig. 3 is a schematic process diagram of a first embodiment of a method for adjusting a surface shape of an active reflecting surface of an astronomical dome provided by the present application;
fig. 4 is a flowchart of a second method of adjusting a surface shape of an active reflective surface of an astronomical dome provided by the present application;
fig. 5 is a schematic network structure diagram of an embodiment of a surface shape adjustment method for an active reflecting surface of an astronomical dome provided by the present application;
fig. 6 is a schematic diagram of a preprocessing process of an embodiment of a surface shape adjustment method for an active reflecting surface of an astronomical dome provided by the present application;
fig. 7 is a functional block diagram of an embodiment of a surface shape adjusting device for an active reflecting surface of an astronomical dome provided by the present application.
Detailed Description
In order that those skilled in the art will better understand the disclosure, the following detailed description will be given with reference to the accompanying drawings. It is to be understood that the embodiments described are only a few embodiments of the present application and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Fig. 1 illustrates the principle of adjusting the surface shape of the active reflecting surface of the astronomical table in the prior art and the present application, wherein "input" in fig. 1 is the common step of the prior art and the present application, except for "input", the dashed frame is the step of the prior art, and the solid frame is the step of the present application.
As shown in fig. 1, the prior art surface shape adjustment principle is: the data input quantity is compared with the quantity fed back by the existing control algorithm such as laser measurement or finite element correction, the deviation is transmitted into the controller, the controller calculates the input parameters of the actuator (namely, the mechanical control device of the piston rod of the active reflecting surface actuator) according to the position deviation at the time t0 and the time t1, then the actuator realizes the output of the actual position of the piston rod at the time t1, and the control feedback adjustment is completed through gradual measurement and adjustment to achieve the final control target. The scheme has the defects of low speed and poor accuracy.
In view of the disadvantages, the core of the present application is to provide a method, an apparatus, a device and a readable storage medium for adjusting the surface shape of an active reflecting surface of an astronomical observatory. As shown in fig. 1, the surface shape adjustment principle of the present application is: the method comprises the steps of firstly training a neural network by utilizing historical data to obtain an elongation prediction model, preprocessing data to be detected in an actual control process, inputting the elongation prediction model to obtain a corresponding elongation prediction value, and directly controlling a piston rod to reach an accurate position by an actuator in one step according to the elongation prediction value. In the training process, output data are sent to the learner in combination with input data, the learner performs preprocessing and label processing on the data to form training samples, then training learning is performed, and the accuracy and robustness of a model algorithm are enhanced.
A first embodiment of a method for adjusting a surface shape of an active reflecting surface of an astronomical dome provided by the present application is described below. Referring to fig. 2, an embodiment includes:
s11, acquiring historical data; generating a training sample according to the historical data;
in practical application, a sample generated according to historical data can be divided into two parts, namely a training sample and a test sample, the training sample is utilized to form an elongation prediction model through deep learning, and the elongation prediction model is optimized through the test sample.
S12, training the neural network by using the training samples to obtain an elongation prediction model;
and presetting a training termination condition, and obtaining a trained elongation prediction model when the training termination condition is reached.
S13, inputting the data to be measured of the target node into an elongation prediction model to obtain an elongation prediction value;
the active reflecting surface comprises a large number of nodes, and the target node is any node of the active reflecting surface of the astronomical phenomena table.
And S14, controlling the piston rod of the target node to stretch according to the predicted elongation value so as to realize surface shape adjustment.
The implementation process of the first embodiment is shown in fig. 3, and includes two parts, namely model training and model reasoning. The training part inputs historical data of all nodes in a preset caliber, a training sample is obtained through preprocessing, then the training sample is input into a deep neural network for training, and after training convergence, an optimal network weight and a corresponding elongation prediction model are obtained. The model is directly used for reasoning, and the reasoning process is that a user inputs data to be tested of nodes in a preset caliber, and the predicted elongation value of each node is output through reasoning by the elongation prediction model.
The embodiment provides a method for adjusting the surface shape of an active reflecting surface of an astronomical table, which comprises the following steps: acquiring historical data; generating a training sample according to the historical data; training the neural network by using the training sample to obtain an elongation prediction model; inputting the data to be measured of the target node into an elongation prediction model to obtain an elongation prediction value, wherein the target node is any node of an active reflecting surface of the astronomical phenomena table; and controlling the piston rod of the target node to stretch according to the predicted value of the elongation so as to realize surface shape adjustment. Therefore, the method predicts the elongation of the piston rod of the node by using the trained neural network, controls the extension of the piston rod according to the elongation to realize the surface shape adjustment of the active reflecting surface of the astronomical table, greatly improves the real-time performance and accuracy compared with the current surface shape adjustment scheme, and realizes the control optimization of the active reflecting surface of the astronomical table.
The second embodiment of the method for adjusting the surface shape of the active reflecting surface of the astronomical dome provided by the present application is described in detail below. Referring to fig. 4, the second embodiment specifically includes:
s21, acquiring historical data;
s22, extracting measurement data and environment data from historical data, wherein the measurement data comprise ground anchor coordinates, node coordinates and piston rod elongation, and the environment data comprise oil pressure data, temperature data and wind speed data;
s23, fitting the node positions by using the measured data to obtain node position fitting quantity; taking the node position fitting quantity and the environment data as feature data in a training sample;
s24, normalizing the elongation of the piston rod, and taking the normalization result as the label data in the training sample;
specifically, the range of motion of the piston rod is determined first, and the extension of the piston rod is normalized according to the range of motion.
S25, generating noise parameters, and performing data enhancement on the training samples by using the noise parameters;
s26, training the neural network by using the training samples, carrying out minimum solution on the loss of the mean square error through gradient descent in the training process, determining the number of layers and the number of kernels of the hidden layers, and obtaining an elongation prediction model;
and S27, generating a control command according to the predicted elongation value, and sending the control command to the actuator of the target node to control the piston rod to extend and retract.
On the basis of the second embodiment, a specific implementation process is described by taking practical application as an example.
Firstly, a hardware prediction platform based on GX4+1080Ti is prepared, and a network, an operating system, deep learning software and the like are configured. And a relatively simple deep learning network is designed in the network model part, relatively simple network parameters are easy to debug, the calculation speed is high, a real-time effect is easy to achieve, and the method is suitable for the application scene. The network design is as shown in fig. 5, and includes 3 large-layer network architectures such as an input layer, a hidden layer, an output layer, etc., where the number of layers and the number of cores of the hidden layer, etc., need to be determined through training tests. The loss of MSE (Mean Square Error) is solved for minimization by gradient descent.
And then, performing data preprocessing, as shown in fig. 6, acquiring historical data, segmenting the historical data according to nodes, and extracting key information, such as measurement data and environment data, wherein the environment data further comprises oil pressure data, temperature data, wind speed data and the like, and the measurement data further comprises ground anchor coordinates, node coordinates, piston rod elongation and the like. And then carrying out inductive sorting and normalization of the data, namely processing the characteristic data and processing the label data. The processing process of the feature data is relatively complex and will be described later; the processing process of the label data is relatively simple, and mainly comprises the steps of determining the motion range of the piston rod and carrying out data normalization processing on the elongation to serve as the label data. And finally, enhancing data, wherein a certain noise parameter is designed to improve the effectiveness of the data and the robustness of the scheme.
The processing of the characteristic data is described in detail below. And fitting the node positions according to the physical characteristics of the actual device. The fitting process mainly uses a ground anchor coordinate Pa, a node coordinate Pn and an elongation Sa, firstly, the change rate Rp between the radial change of a node position and a basic distance caused by the elongation change of a piston rod is calculated, then, the fitting amount Pm of the node position is calculated, and a change rate calculation formula and a fitting amount calculation formula are respectively as follows:
rp = [ [ (Sa-piston rod adjustment) + random disturbance quantity ] + actuator adjustment quantity ]/| Pn-Pa |);
Pm=Pn+Rp*( Pn-Pa)。
it is worth mentioning that in the active reflection surface, the ground anchor is generally arranged on the ground and fixed, the piston rod is controlled by the actuator to extend to drive the node to adjust the surface shape, and the basic distance refers to the difference between the coordinates of the node and the coordinates of the ground anchor. In the surface shape adjusting process, the elongation of the piston rod is changed, and the change of the elongation of the piston rod can cause the radial change of the position of the node; the adjustment amount of the piston rod and the adjustment amount of the actuator are determined by physical characteristics and basically fixed, but different nodes have certain differences. For a fixed node, when the elongation of the piston rod is changed, the adjustment amount of the piston rod and the adjustment amount of the actuator are not changed. The radial direction refers to the direction passing through the axis in the radial plane and can be understood as the direction from the ground anchor to the node, and the difference between the radial direction of the node and the coordinates of the node is one direction and one coordinate value.
The node position fitting amount is used as characteristic data together with oil pressure data, wind speed data, temperature data, and the like, for predicting the elongation. The node position fitting quantity inversely deduced by the elongation is used as an input instead of the node coordinates here because the node position fitting quantity is more in line with the actual physical characteristics of the apparatus than the node coordinates.
After the data are preprocessed, neural network training can be carried out to obtain an elongation prediction model, and the inference module is deployed before. And accessing the data to be tested on the site, and performing model learning and deployment reasoning verification.
The final test showed that: under the hardware platform equipment, the elongation prediction of hundreds of nodes can be completed within 0.2 second, and the second-level real-time test requirement of a closed-loop system can be completely met.
Therefore, the embodiment adopts the neural network in the field of artificial intelligence, and overcomes the limitations of overlong measurement period and complex environmental conditions when the existing scheme is applied to the large-caliber active reflecting surface for adjusting the surface shape. On the basis of GX4+1080Ti hardware, historical data are preprocessed, then a neural network model is designed for training and reasoning, and finally rapid prediction of elongation is achieved, so that the surface shape adjustment process is controlled.
The following describes a surface shape adjusting device for an active reflecting surface of an astronomical dome according to an embodiment of the present application, and the surface shape adjusting device for an active reflecting surface of an astronomical dome described below and the surface shape adjusting method for an active reflecting surface of an astronomical dome described above can be referred to correspondingly.
As shown in fig. 7, the device for adjusting the surface shape of the active reflection surface of the astronomical table of the present embodiment comprises:
a sample generation module 71, configured to obtain historical data; generating a training sample according to the historical data;
the training module 72 is configured to train the neural network by using the training samples to obtain an elongation prediction model;
the prediction module 73 is configured to input data to be measured of a target node into the elongation prediction model to obtain an elongation prediction value, where the target node is any node of the active reflecting surface of the astronomical observatory;
and the control module 74 is configured to control the piston rod of the target node to extend and retract according to the predicted elongation value, so as to realize surface shape adjustment.
The surface shape adjusting device of the active reflecting surface of the astronomical table of the present embodiment is used for realizing the surface shape adjusting method of the active reflecting surface of the astronomical table, and therefore, the specific implementation of the device can be seen in the above embodiment of the surface shape adjusting method of the active reflecting surface of the astronomical table, and will not be described herein.
In addition, this application still provides an astronomical phenomena platform initiative plane shape adjustment device, includes:
a memory: for storing a computer program;
a processor: the computer program is used for executing the computer program to realize the surface shape adjusting method of the active reflecting surface of the astronomical table as described above.
Finally, the present application provides a readable storage medium having stored thereon a computer program for implementing the method for adjusting the surface shape of an active reflecting surface of an astronomical stage as described above when the computer program is executed by a processor.
The embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same or similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
The above detailed descriptions of the solutions provided in the present application, and the specific examples applied herein are set forth to explain the principles and implementations of the present application, and the above descriptions of the examples are only used to help understand the method and its core ideas of the present application; meanwhile, for a person skilled in the art, according to the idea of the present application, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present application.
Claims (9)
1. A surface shape adjusting method of an active reflecting surface of an astronomical table is characterized by comprising the following steps:
acquiring historical data; generating a training sample according to the historical data;
training a neural network by using the training sample to obtain an elongation prediction model;
inputting the data to be measured of a target node into the elongation prediction model to obtain an elongation prediction value, wherein the target node is any node of the active reflecting surface of the astronomical phenomena table;
controlling a piston rod of the target node to stretch according to the predicted elongation value so as to realize surface shape adjustment;
generating a training sample according to the historical data, comprising:
extracting measurement data and environment data from the historical data, wherein the measurement data comprises ground anchor coordinates, node coordinates and piston rod elongation;
fitting the node positions by using the measurement data to obtain node position fitting quantity; taking the node position fitting quantity and the environment data as feature data in a training sample;
and carrying out normalization processing on the elongation of the piston rod, and taking a normalization result as label data in the training sample.
2. The method of claim 1, wherein the environmental data comprises oil pressure data, temperature data, wind speed data.
3. The method of claim 1, wherein normalizing the piston rod elongation comprises:
and determining the moving range of the piston rod, and performing normalization processing on the elongation of the piston rod according to the moving range.
4. The method of claim 1, wherein after generating training samples based on the historical data, further comprising:
and generating a noise parameter, and performing data enhancement on the training sample by using the noise parameter.
5. The method of claim 1, wherein the neural network includes a hidden layer, and wherein training the neural network with the training samples to obtain an elongation prediction model comprises:
and training the neural network by using the training samples, carrying out minimum solution on the loss of the mean square error through gradient descent in the training process, determining the number of layers and the number of kernels of the hidden layers, and obtaining an elongation prediction model.
6. The method according to any one of claims 1 to 5, wherein the controlling of the piston rod of the target node to extend and retract according to the elongation prediction value comprises:
and generating a control instruction according to the predicted elongation value, and sending the control instruction to an actuator of the target node to control the piston rod to extend and retract.
7. A surface shape adjusting device of an active reflecting surface of an astronomical table is characterized by comprising:
the sample generation module is used for acquiring historical data; generating a training sample according to the historical data;
the training module is used for training the neural network by using the training sample to obtain an elongation prediction model;
the prediction module is used for inputting the data to be measured of a target node into the elongation prediction model to obtain an elongation prediction value, wherein the target node is any node of the active reflecting surface of the astronomical phenomena table;
the control module is used for controlling the piston rod of the target node to stretch according to the predicted elongation value so as to realize surface shape adjustment;
the sample generation module is to:
extracting measurement data and environment data from the historical data, wherein the measurement data comprises ground anchor coordinates, node coordinates and piston rod elongation;
fitting the node positions by using the measurement data to obtain node position fitting quantity; taking the node position fitting quantity and the environment data as feature data in a training sample;
and carrying out normalization processing on the elongation of the piston rod, and taking a normalization result as label data in the training sample.
8. A surface shape adjusting device of an active reflecting surface of an astronomical table, comprising:
a memory: for storing a computer program;
a processor: the computer program is executed to implement the surface shape adjustment method of the active reflecting surface of the astronomical table as claimed in any one of claims 1 to 6.
9. A readable storage medium, characterized in that the readable storage medium has stored thereon a computer program, which when executed by a processor, is configured to implement the method for adjusting the surface shape of an active reflecting surface of an astronomical table as claimed in any one of claims 1 to 6.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110911032.7A CN113361149B (en) | 2021-08-10 | 2021-08-10 | Surface shape adjusting method, device and equipment for active reflecting surface of astronomical table |
| PCT/CN2021/142852 WO2023015829A1 (en) | 2021-08-10 | 2021-12-30 | Surface shape adjustment method, apparatus and device for active reflection surface of observatory |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110911032.7A CN113361149B (en) | 2021-08-10 | 2021-08-10 | Surface shape adjusting method, device and equipment for active reflecting surface of astronomical table |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113361149A CN113361149A (en) | 2021-09-07 |
| CN113361149B true CN113361149B (en) | 2021-10-29 |
Family
ID=77540850
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110911032.7A Active CN113361149B (en) | 2021-08-10 | 2021-08-10 | Surface shape adjusting method, device and equipment for active reflecting surface of astronomical table |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN113361149B (en) |
| WO (1) | WO2023015829A1 (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113361149B (en) * | 2021-08-10 | 2021-10-29 | 苏州浪潮智能科技有限公司 | Surface shape adjusting method, device and equipment for active reflecting surface of astronomical table |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109871651A (en) * | 2019-03-14 | 2019-06-11 | 中国科学院国家天文台 | A kind of digital twins' construction method of FAST Active Reflector |
| CN111144609A (en) * | 2019-11-13 | 2020-05-12 | 瀚蓝绿电固废处理(佛山)有限公司 | Boiler exhaust emission prediction model establishing method, prediction method and device |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7343099B2 (en) * | 2004-02-12 | 2008-03-11 | Metrologic Instruments, Inc. | Free space optical (FSO) laser communication system employing fade mitigation measures based on laser beam speckle tracking and locking principles |
| CN109873253B (en) * | 2019-02-25 | 2021-03-16 | 中国科学院紫金山天文台 | Active reflecting surface shape adjusting method based on axial gain measurement |
| CN110597051A (en) * | 2019-09-24 | 2019-12-20 | 南京理工大学 | Stewart stable platform control method based on RBF neural network |
| CN113361149B (en) * | 2021-08-10 | 2021-10-29 | 苏州浪潮智能科技有限公司 | Surface shape adjusting method, device and equipment for active reflecting surface of astronomical table |
-
2021
- 2021-08-10 CN CN202110911032.7A patent/CN113361149B/en active Active
- 2021-12-30 WO PCT/CN2021/142852 patent/WO2023015829A1/en unknown
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109871651A (en) * | 2019-03-14 | 2019-06-11 | 中国科学院国家天文台 | A kind of digital twins' construction method of FAST Active Reflector |
| CN111144609A (en) * | 2019-11-13 | 2020-05-12 | 瀚蓝绿电固废处理(佛山)有限公司 | Boiler exhaust emission prediction model establishing method, prediction method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113361149A (en) | 2021-09-07 |
| WO2023015829A1 (en) | 2023-02-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102052289B1 (en) | Method and apparatus for detecting equivalent load of wind turbine | |
| US8839656B2 (en) | Method of checking a wind turbine in a wind farm for a yaw misalignment, method of monitoring a wind turbine in a wind farm and monitoring apparatus | |
| CN113361149B (en) | Surface shape adjusting method, device and equipment for active reflecting surface of astronomical table | |
| US20220186709A1 (en) | Reinforcement learning-based real time robust variable pitch control of wind turbine systems | |
| EP3608536A1 (en) | Feedforward control method and device for wind turbine set in wind farm | |
| CN110826195A (en) | Wave compensation control algorithm based on ant colony optimization BP neural network | |
| KR101610595B1 (en) | Expedient Storm Surges Prediction System and Method | |
| Tongle et al. | Tailings saturation line prediction based on genetic algorithm and BP neural network | |
| CN115204058B (en) | Ground wave radar flow field calculation method and device based on BP neural network | |
| CN111624522A (en) | Ant colony optimization-based RBF neural network control transformer fault diagnosis method | |
| CN109062040A (en) | Predictive PID method based on the optimization of system nesting | |
| CN117669388B (en) | Fault sample generation method, device and computer medium | |
| CN111680823A (en) | Wind direction information prediction method and system | |
| CN114265368B (en) | Combined state self-adaptive estimation method for aero-engine servo control system | |
| EP3978749A1 (en) | Feedforward control method and apparatus for wind turbine generator set, and control system | |
| EP2849131A2 (en) | Method and system for estimating optimal generation quantity of wind turbine in wind farm | |
| CN111355247A (en) | Power grid low-frequency oscillation prediction method and device | |
| CN114386320A (en) | Steam turbine valve management method, device, equipment and storage medium | |
| CN109507989A (en) | A kind of method of unmanned plane sensor fault diagnosis | |
| CN113868967A (en) | Structure monitoring data prediction method, device and storage medium | |
| CN111290118A (en) | Decoupling control method and device for deformable mirror | |
| CN112327666A (en) | Method for determining target function weight matrix of power cruise system control model | |
| CN109799765A (en) | Steering engine electrohydraulic servo system intelligent control method | |
| CN115809506B (en) | Method and system for forecasting linear wave load in ship structural design | |
| Burghardt et al. | Simplifying Risk Analysis to Determine the Influence of Wind Turbines to the Electric Field of a DVOR Antenna Using Artificial Neural Networks |
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 |