Position correction method of main bracket of solar thermal power generation collector based on laser profilometer

Abstract

The present invention discloses a method for correcting the position of a main bracket of a solar thermal power generation collector based on a laser profiler, which relates to the field of data processing technology, including: obtaining a set of position correction scanning configuration parameters; executing the layout of the profiler and the ground rail module, connecting the mobile gantry module with the ground rail module, and installing the laser profiler on the mobile gantry module; collecting the scanning data of the laser profiler to generate a scanning point cloud map; performing 3D point cloud simulation to generate a simulated point cloud map; performing single-point deviation calculation on multiple scanning points and simulation points to obtain multiple single-point deviation values; performing nearest neighbor same-direction fusion to obtain multiple deviation areas; identifying multiple area deviation values ​​and multiple area deviation directions to obtain a position correction scheme. The present invention solves the technical problem of low accuracy of position correction of the main bracket of a solar thermal power generation collector when the prior art faces a complex working environment and high-precision correction requirements, and achieves the technical effect of accurately correcting the position of the main bracket of the solar thermal power generation collector.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and in particular to a method for correcting the position of a main bracket of a photothermal power generation collector based on a laser profilometer. Background Art

[0002] With the development of renewable energy, CSP technology has been widely used in many countries. CSP converts solar energy into thermal energy and then generates electricity through thermal energy, providing an important way to achieve green energy. However, with the continuous expansion of the scale of CSP projects and the increasing technical requirements, the position correction of the main bracket of the CSP collector is also facing increasing challenges.

[0003] Although the existing methods for correcting the position of the main bracket of the CSP collector have achieved a certain degree of automation, there are still some problems. Traditional automated correction methods mainly rely on conventional sensors and control systems, which are often unable to cope with complex working environments and high-precision correction requirements. Specifically, the existing methods still have a lot of room for improvement in correction accuracy. With the expansion of the scale of CSP systems and the improvement of technical requirements, the limitations of traditional automated correction methods have become increasingly obvious, making it difficult to meet the requirements of modern CSP systems for efficient, precise and stable operation. Summary of the invention

[0004] The present application provides a method for calibrating the position of a main bracket of a solar thermal power generation collector based on a laser profilometer, which is used to solve the technical problem of low accuracy of position calibration of a main bracket of a solar thermal power generation collector when facing complex working environments and high-precision calibration requirements in the prior art.

[0005] In view of the above problems, the present application provides a method for correcting the position of a main bracket of a solar thermal power generation collector based on a laser profilometer.

[0006] The present application provides a method for correcting the position of a main bracket of a solar thermal power generation collector based on a laser profilometer, the method comprising:

[0007] Perform position correction scanning parameter configuration optimization to generate a position correction scanning configuration parameter set, wherein the position correction scanning configuration parameter set includes the number of laser profilers and the moving speed of the mobile gantry module; perform profiler layout based on the number of laser profilers, by respectively arranging a group of ground rail modules on the left and right sides of the target solar thermal power generation collector main bracket, connecting the mobile gantry module to the two ground rail modules by bolts, and installing K laser profilers on the mobile gantry module through a visual frame at a preset interval, K is the number of laser profilers, and K is an integer greater than or equal to 1; control the mobile gantry module to move from one axial end of the target solar thermal power generation collector main bracket to the other axial end according to the moving speed of the mobile gantry module, collect scanning data of the K laser profilers, and generate a scanning point cloud map, wherein the scanning point cloud map includes multiple scanning points and multiple scanning points reference position; performing 3D point cloud simulation according to the moving speed of the mobile gantry module and the design information of the target solar thermal power generation collector main bracket to generate a simulated point cloud map, wherein the simulated point cloud map includes multiple simulated points and multiple simulated point reference positions; based on the multiple scanning point reference positions and the multiple simulation point reference positions, performing single-point deviation calculations on the multiple scanning points and the multiple simulation points to obtain multiple single-point deviation values, wherein the multiple single-point deviation values ​​include multiple deviation direction identifiers, and the deviation directions include positive identifiers and negative identifiers; performing neighbor same-direction fusion according to the multiple single-point deviation values ​​and the multiple deviation direction identifiers to obtain multiple deviation areas, wherein the multiple deviation areas include multiple regional deviation values ​​and multiple regional deviation directions; using a position correction unit to identify the multiple regional deviation values ​​and the multiple regional deviation directions to obtain a position correction scheme.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] The present application performs position correction scanning parameter configuration optimization to generate a position correction scanning configuration parameter set, wherein the position correction scanning configuration parameter set includes the number of laser profilers and the moving speed of the mobile gantry module; the profiler is arranged based on the number of laser profilers, by arranging a group of ground rail modules on the left and right sides of the target solar thermal power generation collector main bracket respectively, connecting the mobile gantry module to the two ground rail modules by bolts, and installing K laser profilers on the mobile gantry module through a visual frame at a preset interval, K is the number of laser profilers, and K is an integer greater than or equal to 1; the mobile gantry module is controlled to move from one axial end of the target solar thermal power generation collector main bracket to the other axial end according to the moving speed of the mobile gantry module, and the scanning data of the K laser profilers are collected to generate a scanning point cloud map, wherein the scanning point cloud map includes multiple scanning points and multiple scanning point reference positions; perform 3D point cloud simulation according to the moving speed of the mobile gantry module and the design information of the main bracket of the target solar thermal power generation collector, and generate a simulated point cloud map, wherein the simulated point cloud map includes multiple simulated points and multiple simulated point reference positions; based on the multiple scanning point reference positions and the multiple simulation point reference positions, perform single-point deviation calculations on the multiple scanning points and the multiple simulation points, and obtain multiple single-point deviation values, wherein the multiple single-point deviation values ​​include multiple deviation direction identifiers, and the deviation directions include positive identifiers and negative identifiers; perform nearest neighbor same-direction fusion according to the multiple single-point deviation values ​​and the multiple deviation direction identifiers, and obtain multiple deviation areas, wherein the multiple deviation areas include multiple regional deviation values ​​and multiple regional deviation directions; use a position correction unit to identify the multiple regional deviation values ​​and the multiple regional deviation directions, and obtain a position correction solution. The present invention solves the technical problem of low accuracy in the correction of the position of the main bracket of a solar thermal power generation collector when the prior art faces a complex working environment and high-precision correction requirements. By configuring the optimized number of laser profilers and the speed of the mobile gantry module, combining 3D point cloud simulation with single-point deviation calculation, and nearest neighbor same-direction fusion technology, the technical effect of accurately correcting the position of the main bracket of the solar thermal power generation collector is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 A schematic flow chart of a method for correcting the position of a main bracket of a solar thermal power generation collector based on a laser profilometer provided in an embodiment of the present application;

[0012] Figure 2A schematic diagram of a flow chart of generating a set of position correction scanning configuration parameters in a method for correcting the position of a main bracket of a solar thermal power generation collector based on a laser profilometer provided in an embodiment of the present application. DETAILED DESCRIPTION

[0013] The present application provides a method for correcting the position of a main bracket of a solar thermal power generation collector based on a laser profiler, which is used to solve the technical problem of low accuracy of the position correction of the main bracket of a solar thermal power generation collector when facing a complex working environment and high-precision correction requirements in the prior art. By configuring the optimized number of laser profilers and the speed of the mobile gantry module, combining 3D point cloud simulation and single-point deviation calculation, as well as the nearest neighbor same-direction fusion technology, the technical effect of accurately correcting the position of the main bracket of the solar thermal power generation collector is achieved.

[0014] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0015] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0016] Embodiment 1

[0017] like Figure 1 As shown, the present application provides a method for correcting the position of a main bracket of a solar thermal power generation collector based on a laser profilometer, the method comprising:

[0018] Step S100: Execute position correction scanning parameter configuration optimization to generate a position correction scanning configuration parameter set, wherein the position correction scanning configuration parameter set includes the number of laser profilers and the moving speed of the mobile gantry module.

[0019] In the embodiment of the present application, when performing the position correction scanning parameter configuration optimization, the initial configuration parameter set is first determined. The initial configuration parameter set includes the initial number of laser profilers and the initial moving gantry module moving speed. The initial value of the number of laser profilers is set according to the size of the target collector main bracket, and the moving speed is set according to the mechanical characteristics and the required scanning accuracy.

[0020] During the parameter optimization process, the number of laser profilers and the moving speed are fine-tuned multiple times by using preset adjustment steps. During fine-tuning, different numbers of laser profilers are used and the speed of the moving gantry module is adjusted. The scanning effect and efficiency are recorded after each adjustment. In this way, different simulated point clouds are generated using multiple scanning simulations. The simulated point cloud includes multiple simulated points and reference positions.

[0021] In order to evaluate the effect of each set of configuration parameters, the generated simulated point cloud map is analyzed using the fitness analysis network layer, and the fitness score is based on indicators such as scanning coverage and overlap rate. According to the fitness score, the optimal configuration parameter set is screened out. In this process, the random combination method is used to combine different numbers of laser profilers and moving speeds to generate multiple fine-tuning configuration parameter sets. Two-dimensional optimization is performed to optimize the scanning overlap rate and scanning coverage rate to obtain the best configuration parameters. In the final optimization stage, the parameter set with the highest fitness is selected as the main particle, and the rest are used as shadow particles. The shadow particles are fine-tuned by the preset step size, and the optimization is continuously iterated in the direction of the main particle until the preset number of optimizations or satisfactory fitness is reached. The final set of configuration parameters is the position correction scanning configuration parameter set, including the number of laser profilers and the moving speed of the mobile gantry module.

[0022] Step S200: Execute profiler deployment based on the number of laser profilers, by respectively deploying a group of ground rail modules on the left and right sides of the target solar thermal power generation collector main bracket, connecting the mobile gantry module to the two ground rail modules by bolts, and installing K laser profilers on the mobile gantry module through a visual frame at a preset spacing distance, where K is the number of laser profilers and is an integer greater than or equal to 1.

[0023] In the embodiment of the present application, when the profiler is deployed, the number K of laser profilers to be used is determined according to the optimization result of the position correction scanning parameter configuration performed previously, where K is an integer greater than or equal to 1. Next, on the left and right sides of the target solar thermal power generation collector main bracket, a flat, stable and obstacle-free position is selected to deploy the ground rail module. The ground rail module should be parallel to the main bracket, and the ground rail module should be firmly mounted on the ground with bolts or other fixing devices to ensure that the ground rail modules are parallel and the spacing is consistent to provide a stable moving track.

[0024] Then, the bottom of the mobile gantry module is tightly connected to the two ground rail modules by bolts or other connectors to ensure that the mobile gantry module can slide smoothly on the ground rails and will not derail due to vibration or external force. Prepare the corresponding vision frame according to the number and size of the laser profilers. The vision frame is designed with appropriate spacing and mounting holes so that the laser profilers can be installed on it at the preset spacing distance. Install K laser profilers on the vision frame by bolts or other fixing devices at the preset spacing distance. Make sure that each laser profiler is firmly fixed to the frame and the scanning range can seamlessly cover the entire bracket surface.

[0025] Step S300: Control the mobile gantry module to move from one axial end to the other axial end of the target solar thermal power generation collector main bracket according to the moving speed of the mobile gantry module, collect scanning data of the K laser profilers, and generate a scanning point cloud map, wherein the scanning point cloud map includes multiple scanning points and multiple scanning point reference positions.

[0026] In the embodiment of the present application, the relevant parameters of the control system are set according to the previously determined moving speed of the mobile gantry module. Next, the mobile gantry module is started by the control system so that it starts to move along the ground rail module at the set speed. While the mobile gantry module is moving, K laser profilers are started for synchronous scanning. The laser profiler emits a laser beam to the surface of the main bracket based on the principle of laser triangulation, receives the reflected light signal, and converts it into scanning data through internal processing to obtain the scanning data of the K laser profilers.

[0027] Next, the point cloud processing technology is used to pre-process the scanning data of K laser profilers, filter out the noise data, and smooth the valid data to generate a scanning point cloud map. The scanning point cloud map includes multiple scanning points and multiple scanning point reference positions, where each scanning point represents a position in three-dimensional space, and multiple scanning points together constitute the three-dimensional shape of the main bracket. The scanning point reference position is pre-set according to the design drawing or CAD model of the main bracket, and is used for subsequent deviation calculation and other steps.

[0028] Step S400: Perform 3D point cloud simulation according to the moving speed of the mobile gantry module and the design information of the target solar thermal power generation collector main bracket to generate a simulated point cloud map, wherein the simulated point cloud map includes multiple simulated points and multiple simulated point reference positions.

[0029] In the embodiment of the present application, CAD software or a design document viewer is first used to access the design information of the target solar thermal power generation collector main bracket, including CAD models, dimensions, material properties, etc.

[0030] Next, in the 3D simulation software, set the corresponding scanning speed and scanning path parameters according to the moving speed of the mobile gantry module. At the same time, according to the technical specifications of the laser profiler, adjust the parameters such as scanning frequency, scanning range and accuracy in the simulation to ensure that the simulation process is as consistent as possible with the actual scanning process. In this process, use computer-aided design tools and simulation software such as SolidWorks, AutoCAD, MATLAB or Simulink to set these parameters.

[0031] Then, a virtual model of the target solar thermal power generation collector main bracket is constructed based on the collected design information using the modeling function in the 3D modeling tool or simulation software. Subsequently, the simulation scanning engine is started in the simulation software to allow the virtual mobile gantry module to carry the virtual laser profiler along the set path and speed. During the movement, the simulation engine simulates the interaction between the laser beam and the surface of the virtual main bracket according to the set parameters and the characteristics of the virtual model, and calculates and generates simulated scanning data. Finally, the simulated scanning data is processed by the point cloud generation algorithm to obtain a simulated point cloud map. The simulated point cloud map includes multiple simulation points and multiple simulation point reference positions.

[0032] Step S500: Based on the multiple scanning point reference positions and the multiple simulation point reference positions, single-point deviation calculations are performed on the multiple scanning points and the multiple simulation points to obtain multiple single-point deviation values, wherein the multiple single-point deviation values ​​include multiple deviation direction identifiers, and the deviation directions include positive identifiers and negative identifiers.

[0033] In the embodiment of the present application, the coordinates of each scan point are first extracted from the scan point cloud data, and the coordinates of each simulation point are extracted from the simulation point cloud data. The scan point reference position and the simulation point reference position are aligned and matched by pre-set coordinates.

[0034] Next, for each scan point, use KD tree or other fast neighborhood search algorithm to find the corresponding simulation point to obtain a matching point pair. Then, calculate the single point deviation value, that is, calculate the height difference of each matching point pair. When the scan point is higher than the simulation point, it is a positive deviation, and when the scan point is lower than the simulation point, it is a negative deviation. Through the above process, multiple single point deviation values ​​are obtained, and the multiple single point deviation values ​​include multiple deviation direction identifiers, and the deviation direction includes a positive identifier and a negative identifier.

[0035] Step S600: performing neighbor same-direction fusion according to the multiple single-point deviation values ​​and the multiple deviation direction identifiers to obtain multiple deviation regions, wherein the multiple deviation regions include multiple regional deviation values ​​and multiple regional deviation directions.

[0036] In an embodiment of the present application, when performing neighbor fusion in the same direction, a fast neighborhood search algorithm is used, such as a KD tree to identify the neighboring points of each single point deviation. After identifying the neighboring points, these points are fused in the same direction according to the deviation direction, that is, when the deviation direction of the neighboring points is consistent with the target point, these points are classified as the same deviation area. In this way, points with consistent deviation directions and adjacent in space are aggregated together to form multiple deviation areas. During the fusion process, the regional deviation value and regional deviation direction of each deviation area are calculated. The regional deviation value is the average value of all single-point deviation values ​​in the area, and the regional deviation direction is the consistency identifier of the direction of all single-point deviation values ​​in the area. Through the above process, multiple deviation areas are obtained.

[0037] Step S700: using a position correction unit to identify the multiple regional deviation values ​​and the multiple regional deviation directions to obtain a position correction solution.

[0038] In an embodiment of the present application, the position correction unit is pre-trained, and the obtained regional deviation value and multiple regional deviation directions are input into the position correction unit for analysis to obtain a position correction solution.

[0039] Further, such as Figure 2 As shown, the method provided in the application embodiment also includes:

[0040] Data extraction is performed based on the design information of the target solar thermal power generation collector main bracket to obtain the main bracket length information and the main bracket upper surface area; the detection range of the laser profiler is collected; the main bracket length information, the main bracket upper surface area and the detection range are identified by a parameter configuration unit to generate an initial configuration parameter set, wherein the initial configuration parameter set includes an initial number of laser profilers and an initial moving gantry module moving speed; position correction scanning parameter configuration optimization is performed on the initial configuration parameter set to obtain the position correction scanning configuration parameter set.

[0041] In the embodiment of the present application, when extracting data based on the design information of the main bracket of the target solar thermal power generation collector, a CAD tool such as AutoCAD or SolidWorks is used to directly extract the main bracket length information and the main bracket upper surface area from the CAD model. Next, the detection range of the laser profiler is obtained by querying the specification table of the laser profiler.

[0042] Next, the parameter configuration unit is used to identify and integrate the extracted main bracket length information, upper surface area, and detection range of the laser profiler. The parameter configuration unit is an integrated data processing and analysis system, which consists of a data input module, a data processing module, and a result output module. First, the length information and upper surface area of ​​the main bracket are input into the parameter configuration unit; secondly, the detection range of the laser profiler is input. The parameter configuration unit uses these input data to generate an initial configuration parameter set, including the initial number of laser profilers and the initial moving gantry module moving speed, through internal algorithms such as geometric calculation and coverage analysis.

[0043] The initial configuration parameter set is optimized for position correction scanning parameter configuration, and an optimization algorithm such as genetic algorithm, particle swarm optimization or gradient descent algorithm is used to adjust the parameters to minimize the scanning overlap rate and maximize the scanning coverage rate. Through the above process, a position correction scanning configuration parameter set is obtained.

[0044] Furthermore, the method provided in the application embodiment also includes:

[0045] The initial number of laser profilers and the initial moving gantry module moving speed are fine-tuned multiple times according to a preset adjustment method to obtain multiple numbers of fine-tuned laser profilers and multiple speeds of fine-tuned moving gantry modules, wherein the preset adjustment method is to increase or decrease the initial number of laser profilers according to a preset number adjustment step and to increase or decrease the initial moving gantry module moving speed according to a preset moving speed adjustment step; the multiple numbers of fine-tuned laser profilers and the multiple speeds of fine-tuned moving gantry modules are randomly combined to obtain multiple sets of fine-tuning configuration parameters; the multiple sets of fine-tuning configuration parameters are subjected to two-dimensional optimization to obtain the position correction scanning configuration parameter set, wherein the two-dimensional optimization is to reduce the laser profiler scanning overlap rate and to improve the scanning coverage rate when scanning the upper surface of the main bracket of the target solar thermal power generation collector.

[0046] In the embodiment of the present application, the initial number of laser profilers and the initial moving gantry module moving speed are first fine-tuned multiple times according to a preset adjustment method. The preset adjustment method is to obtain multiple fine-tuning numbers of laser profilers and multiple fine-tuning moving gantry module moving speeds by gradually increasing or decreasing the number of laser profilers and increasing or decreasing the initial moving gantry module moving speed according to a preset adjustment step.

[0047] Next, a combination algorithm is used to randomly combine multiple fine-tuning laser profiler numbers and multiple fine-tuning moving gantry module moving speeds to obtain multiple fine-tuning configuration parameter sets, and the obtained multiple fine-tuning configuration parameter sets cover all adjusted parameter values. For example, if the number of laser profilers is adjusted between 4 and 10, and the moving speed is adjusted between 1.0 and 2.0 m / s, all possible combinations within these parameter ranges are generated by the combination algorithm.

[0048] Then, a two-dimensional optimization is performed on multiple sets of fine-tuning configuration parameters. The two-dimensional optimization aims to reduce the scanning overlap rate of the laser profiler and improve the coverage rate when scanning the upper surface of the main bracket of the target solar thermal power generation collector. The performance of each parameter combination is evaluated by a two-dimensional optimization algorithm, such as particle swarm optimization or genetic algorithm. Specifically, an initial population is generated based on the fine-tuning parameter combination, and each population member represents a parameter combination. The fitness of each parameter combination is calculated, and the optimal parameter combination is selected based on the fitness. A new parameter combination is generated through crossover and mutation, and the fitness evaluation and parameter optimization are repeated until the preset optimization criteria are reached. Through the above process, a set of position correction scanning configuration parameters is obtained.

[0049] Furthermore, the method provided in the application embodiment also includes:

[0050] A scanning simulation is performed based on the multiple fine-tuning configuration parameter sets to generate multiple scanning simulation images; the multiple scanning simulation images are analyzed using a fitness analysis network layer to obtain multiple fine-tuning fitnesses; the fine-tuning configuration parameter set corresponding to the maximum value of the multiple fine-tuning fitnesses is used as the main particle, and the remaining multiple fine-tuning configuration parameter sets are used as multiple shadow particles; with the main particle as the adjustment direction, the multiple shadow particles are fine-tuned according to the preset number adjustment step and the preset moving speed adjustment step to obtain multiple fine-tuning shadow particles, wherein the multiple fine-tuning shadow particles have multiple fine-tuning shadow fitnesses, and the main particle has a main particle fitness; based on the multiple fine-tuning shadow particles, the main particle is iteratively updated until the preset number of iterations is met to obtain the position correction scanning configuration parameter set.

[0051] In an embodiment of the present application, a scanning simulation is first performed based on multiple sets of fine-tuning configuration parameters. For each set of fine-tuning configuration parameters, a point cloud simulation tool, such as a PCL library, is used to simulate the scanning process of a laser profiler. Corresponding scanning point cloud data are generated by simulating the working state of the laser profiler under different configuration parameters. These point cloud data are then converted into scanning simulation images to generate multiple scanning simulation images. Next, the multiple scanning simulation images are analyzed using a fitness analysis network layer, wherein the fitness analysis network layer is pre-built. When building, a large number of annotated scanning image data sets are first collected from a historical database. These data sets contain scanning images under different configuration parameters, as well as corresponding fitness scores. Next, the convolutional neural network is trained using this evil data to obtain a fitness analysis network layer. Multiple scanning simulation images are analyzed through the fitness analysis network layer to obtain multiple fine-tuning fitnesses.

[0052] Then, the fine-tuning configuration parameter set corresponding to the maximum value of multiple fine-tuning fitness is used as the main particle, and the remaining multiple fine-tuning configuration parameter sets are used as shadow particles. The main particle is used as the adjustment direction, and the shadow particles are fine-tuned according to the preset number adjustment step and the preset movement speed adjustment step. During the adjustment process, the direction of the main particle is kept unchanged, and new configuration parameter combinations are explored by changing the parameter values ​​of the shadow particles. After each shadow particle is fine-tuned, a new fine-tuned shadow particle is generated. The newly generated fine-tuned shadow particles are evaluated using the fitness analysis network layer to calculate the fitness value of each shadow particle.

[0053] Compare the fitness values ​​of the main particle and the shadow particle. If the fitness of a shadow particle exceeds that of the main particle, use the shadow particle as the new main particle. Repeat this process until the preset number of iterations is reached. Through the above process, the configuration parameters are gradually optimized, and finally a set of position correction scanning configuration parameters is obtained.

[0054] Furthermore, the method provided in the application embodiment also includes:

[0055] Determine whether the maximum value among the multiple fine-tuning shadow fitnesses is greater than or equal to the fitness of the main particle. If so, update the fine-tuning shadow particle corresponding to the maximum value among the multiple fine-tuning shadow fitnesses as the main particle, and perform optimization as a new adjustment direction until a preset number of optimization times is met, and use the main particle corresponding to the maximum value of the fitness in the optimization process as the target particle; and use the configuration parameter set corresponding to the target particle as the position correction scanning configuration parameter set.

[0056] In an embodiment of the present application, firstly, the maximum value among a plurality of fine-tuning shadow fitnesses is compared with the fitness of the main particle. If the maximum value among the fine-tuning shadow fitnesses is greater than or equal to the fitness of the main particle, the main particle is updated, and the fine-tuning shadow particle with the maximum fitness value is used as the new main particle, and its fitness value is used as the new main particle fitness. Next, with the new main particle as the direction, the shadow particles are fine-tuned according to the preset number of adjustment steps and the preset moving speed adjustment step length to generate new fine-tuning shadow particles. The fitness of these new fine-tuning shadow particles is then evaluated, and the fitness comparison and update process is repeated. The above process is repeated until the preset number of optimization searches is reached, and the main particle corresponding to the maximum fitness value in the entire optimization process is selected as the target particle. The target particle represents the optimal configuration parameter set found in the optimization process. Finally, the configuration parameter set corresponding to the target particle is used as the position correction scanning configuration parameter set.

[0057] Furthermore, the method provided in the application embodiment also includes:

[0058] When the maximum value among the multiple fine-tuning shadow fitnesses is less than the fitness of the main particle, a first random amount and a second random amount are obtained based on a random number generation function; when the first random amount is less than or equal to the second random amount, the fine-tuning shadow particle corresponding to the maximum value among the multiple fine-tuning shadow fitnesses is accepted as the main particle, and it is used as a new adjustment direction for optimization; when the first random amount is greater than the second random amount, the fine-tuning shadow particle corresponding to the maximum value among the multiple fine-tuning shadow fitnesses is refused to be accepted to update the main particle, and the adjustment scheme for obtaining the multiple fine-tuning shadow particles is disabled, and the main particle is still used as the adjustment direction for optimization until the preset number of optimization times is met, and the main particle corresponding to the maximum value of the fitness in the optimization process is used as the target particle, and the configuration parameter set corresponding to the target particle is used as the position correction scanning configuration parameter set.

[0059] In an embodiment of the present application, when the maximum value among multiple fine-tuned shadow fitnesses is less than the main particle fitness, a random number generation function is used to obtain a first random amount and a second random amount, which are used to determine whether to accept the current shadow particle as a new main particle.

[0060] If the first random quantity is less than or equal to the second random quantity, the fine-tuned shadow particle corresponding to the maximum fitness value in the shadow particle is accepted as the new main particle, and it is used as the new adjustment direction to continue optimization, avoiding falling into the local optimum and increasing the diversity and exploration of the optimization process.

[0061] If the first random value is greater than the second random value, the fine-tuned shadow particle corresponding to the maximum value of the multiple fine-tuned shadow fitnesses is rejected as the new main particle, and the adjustment scheme of the fine-tuned shadow particle is disabled. At this time, the original main particle is still used as the adjustment direction to continue the optimization until the preset optimization times are met. In this way, some unfavorable adjustment directions can be avoided.

[0062] Repeat the above randomized decision-making process, perform fine-tuning and fitness evaluation until the preset number of optimizations is reached. After the preset number of optimizations is reached, the main particle corresponding to the maximum fitness in the entire optimization process is selected as the target particle. Finally, the configuration parameter set corresponding to the target particle is used as the position correction scanning configuration parameter set.

[0063] Furthermore, the method provided in the application embodiment also includes:

[0064] A first single-point deviation value is randomly extracted from the multiple single-point deviation values, wherein the first single-point deviation value has a first deviation direction identifier; a first neighborhood set of the first single-point deviation value is obtained, wherein the first neighborhood set includes multiple first neighborhood single-point deviation values ​​and multiple first neighborhood deviation direction identifiers; a plurality of scanning points corresponding to multiple first neighborhood single-point deviation values ​​whose first neighborhood deviation direction identifier is in the same direction as the first deviation direction identifier and whose difference between the first neighborhood single-point deviation value and the first single-point deviation value is within a preset difference range are aggregated with the first scanning point corresponding to the first single-point deviation value to generate a first deviation area; an average of the single-point deviation values ​​in the first deviation area is calculated to generate a first area deviation value, and the direction corresponding to the first deviation direction identifier is used as the first area deviation direction; and a plurality of deviation areas are obtained by performing nearest neighbor same-direction fusion according to the multiple single-point deviation values ​​and the multiple deviation direction identifiers.

[0065] In an embodiment of the present application, firstly, a plurality of single-point deviation values ​​are randomly sampled to obtain a first single-point deviation value, and the first single-point deviation value has a first deviation direction identifier. Next, a KD tree or other neighborhood search algorithm is used to find neighborhood points around the first single-point deviation value, and a first neighborhood set of the first single-point deviation value is obtained. The first neighborhood set includes a plurality of first neighborhood single-point deviation values ​​and a plurality of first neighborhood deviation direction identifiers.

[0066] Then, the neighborhood points in the first neighborhood set whose first neighborhood deviation direction identifier is in the same direction as the first deviation direction identifier are screened out. By screening out those points whose deviation direction identifier is the same as the deviation direction identifier of the first single-point deviation value, and whose difference between the neighborhood single-point deviation value and the first single-point deviation value is within the preset difference range, it is ensured that the selected neighborhood points are similar to the first single-point deviation value in spatial position and deviation direction. The scanning points corresponding to the multiple first neighborhood single-point deviation values ​​after screening are aggregated with the scanning points corresponding to the first single-point deviation value to generate a first deviation area. Finally, the mean of the single-point deviation values ​​in the first deviation area is calculated to generate the first area deviation value, and the direction corresponding to the first deviation direction identifier is used as the first area deviation direction.

[0067] Repeat the above steps to perform nearest neighbor same-direction fusion on multiple single-point deviation values ​​and multiple deviation direction identifiers to obtain multiple deviation areas.

[0068] Furthermore, the method provided in the application embodiment also includes:

[0069] A plurality of sample area deviation values, a plurality of sample area deviation directions and a plurality of sample position correction schemes are obtained as construction data; the encoder and the decoder are supervisedly trained using the construction data until convergence, and the trained position correction unit is obtained.

[0070] In an embodiment of the present application, a plurality of sample area deviation values, sample area deviation directions and sample position correction schemes are first extracted from a historical database. The historical database stores a large amount of scanning data and corresponding correction schemes, which are collected and saved by the laser profiler during multiple scanning processes in the past. The sample area deviation value represents the error of the scanning point relative to the design model, and the sample area deviation direction represents the direction of these errors. The sample position correction scheme is a correction parameter generated based on these deviation data, including the position adjustment and movement speed of the laser profiler. The sample data extracted from the historical database is used as construction data, that is, a training data set. Each construction data includes two parts, an input and an output, the input is the sample area deviation value and the deviation direction, and the output is the corresponding correction scheme.

[0071] Next, use a deep learning framework such as TensorFlow or PyTorch to build an encoder-decoder model. The encoder encodes the input data into a vector of fixed length, and the decoder generates a correction scheme based on the encoded vector. Before training, the construction data is preprocessed, including standardization and normalization. Then, the supervised learning method is used to train the model using the preprocessed construction data. The back-propagation algorithm is used to adjust the model parameters so that the output correction scheme gradually approaches the real correction scheme. Continue training until the model loss function converges, stop training, and obtain a trained position correction unit.

[0072] In the embodiments of the present application, in summary, the embodiments of the present application have at least the following technical effects:

[0073] The present application performs position correction scanning parameter configuration optimization to generate a position correction scanning configuration parameter set, wherein the position correction scanning configuration parameter set includes the number of laser profilers and the moving speed of the mobile gantry module; the profiler is arranged based on the number of laser profilers, by arranging a group of ground rail modules on the left and right sides of the target solar thermal power generation collector main bracket respectively, connecting the mobile gantry module to the two ground rail modules by bolts, and installing K laser profilers on the mobile gantry module through a visual frame at a preset interval, K is the number of laser profilers, and K is an integer greater than or equal to 1; the mobile gantry module is controlled to move from one axial end of the target solar thermal power generation collector main bracket to the other axial end according to the moving speed of the mobile gantry module, and the scanning data of the K laser profilers are collected to generate a scanning point cloud map, wherein the scanning point cloud map includes multiple scanning points and multiple scanning point reference positions; perform 3D point cloud simulation according to the moving speed of the mobile gantry module and the design information of the main bracket of the target solar thermal power generation collector, and generate a simulated point cloud map, wherein the simulated point cloud map includes multiple simulated points and multiple simulated point reference positions; based on the multiple scanning point reference positions and the multiple simulation point reference positions, perform single-point deviation calculations on the multiple scanning points and the multiple simulation points, and obtain multiple single-point deviation values, wherein the multiple single-point deviation values ​​include multiple deviation direction identifiers, and the deviation directions include positive identifiers and negative identifiers; perform nearest neighbor same-direction fusion according to the multiple single-point deviation values ​​and the multiple deviation direction identifiers, and obtain multiple deviation areas, wherein the multiple deviation areas include multiple regional deviation values ​​and multiple regional deviation directions; use a position correction unit to identify the multiple regional deviation values ​​and the multiple regional deviation directions, and obtain a position correction solution. The present invention solves the technical problem of low accuracy in the correction of the position of the main bracket of a solar thermal power generation collector when the prior art faces a complex working environment and high-precision correction requirements. By configuring the optimized number of laser profilers and the speed of the mobile gantry module, combining 3D point cloud simulation with single-point deviation calculation, and nearest neighbor same-direction fusion technology, the technical effect of accurately correcting the position of the main bracket of the solar thermal power generation collector is achieved.

[0074] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0075] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0076] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A method for correcting the position of a main bracket of a solar thermal power generation collector based on a laser profiler, characterized in that: include: Execute position correction scanning parameter configuration optimization to generate a position correction scanning configuration parameter set, wherein the position correction scanning configuration parameter set includes the number of laser profilers and the moving speed of the mobile gantry module; Performing profiler deployment based on the number of laser profilers, by respectively deploying a set of ground rail modules on the left and right sides of the target solar thermal power generation collector main bracket, connecting the mobile gantry module to the two ground rail modules by bolts, and installing K laser profilers on the mobile gantry module through a visual frame at a preset interval, where K is the number of laser profilers and K is an integer greater than or equal to 1; Controlling the mobile gantry module to move from one axial end to the other axial end of the target solar thermal power generation collector main bracket according to the moving speed of the mobile gantry module, collecting the scanning data of the K laser profilers, and generating a scanning point cloud map, wherein the scanning point cloud map includes a plurality of scanning points and a plurality of scanning point reference positions; Performing 3D point cloud simulation according to the moving speed of the mobile gantry module and the design information of the target solar thermal power generation collector main bracket to generate a simulated point cloud map, wherein the simulated point cloud map includes a plurality of simulated points and a plurality of simulated point reference positions; Based on the multiple scanning point reference positions and the multiple simulation point reference positions, single-point deviation calculations are performed on the multiple scanning points and the multiple simulation points to obtain multiple single-point deviation values, wherein the multiple single-point deviation values ​​include multiple deviation direction identifiers, and the deviation directions include positive identifiers and negative identifiers; Performing neighbor same-direction fusion according to the multiple single-point deviation values ​​and the multiple deviation direction identifiers to obtain multiple deviation regions, wherein the multiple deviation regions include multiple regional deviation values ​​and multiple regional deviation directions; The position correction unit is used to identify the multiple regional deviation values ​​and the multiple regional deviation directions to obtain a position correction solution.

2. The method for correcting the position of the main bracket of a solar thermal power generation collector based on a laser profiler according to claim 1, characterized in that: include: Extracting data based on the design information of the target solar thermal power generation collector main bracket to obtain the main bracket length information and the main bracket upper surface area; Collect the detection range of the laser profiler; Using a parameter configuration unit to identify the main bracket length information, the main bracket upper surface area and the detection range, and generate an initial configuration parameter set, wherein the initial configuration parameter set includes an initial number of laser profilers and an initial moving gantry module moving speed; Performing position correction scanning parameter configuration optimization on the initial configuration parameter set to obtain the position correction scanning configuration parameter set.

3. The method for correcting the position of the main bracket of a solar thermal power generation collector based on a laser profiler according to claim 2, characterized in that: include: The initial number of laser profilers and the initial moving gantry module moving speed are fine-tuned multiple times according to a preset adjustment method to obtain multiple fine-tuned numbers of laser profilers and multiple fine-tuned moving gantry module moving speeds, wherein the preset adjustment method is to increase or decrease the initial number of laser profilers according to a preset number adjustment step and to increase or decrease the initial moving gantry module moving speed according to a preset moving speed adjustment step; Randomly combining the number of the plurality of fine-tuning laser profilers and the moving speeds of the plurality of fine-tuning mobile gantry modules to obtain a plurality of fine-tuning configuration parameter sets; The multiple fine-tuning configuration parameter sets are subjected to two-dimensional optimization to obtain the position correction scanning configuration parameter set, wherein the two-dimensional optimization is to reduce the scanning overlap rate of the laser profiler and to improve the scanning coverage rate when scanning the upper surface of the main bracket of the target solar thermal power generation collector.

4. The method for correcting the position of the main bracket of a solar thermal power generation collector based on a laser profiler according to claim 3, characterized in that: include: Performing scanning simulation based on the multiple fine-tuning configuration parameter sets to generate multiple scanning simulation images; Analyzing the plurality of scanned simulation images using a fitness analysis network layer to obtain a plurality of fine-tuning fitnesses; The fine-tuning configuration parameter set corresponding to the maximum value among the multiple fine-tuning fitness values ​​is used as the main particle, and the remaining multiple fine-tuning configuration parameter sets are used as multiple shadow particles; Taking the main particle as the adjustment direction, fine-tuning the plurality of shadow particles according to the preset number adjustment step and the preset moving speed adjustment step to obtain a plurality of fine-tuned shadow particles, wherein the plurality of fine-tuned shadow particles have a plurality of fine-tuned shadow fitnesses, and the main particle has a main particle fitness; The main particle is iteratively updated based on the multiple fine-tuning shadow particles until a preset number of iterations is met, thereby obtaining the position correction scanning configuration parameter set.

5. The method for correcting the position of the main bracket of a solar thermal power generation collector based on a laser profiler according to claim 4, characterized in that: include: Determine whether the maximum value among the multiple fine-tuning shadow fitness values ​​is greater than or equal to the fitness value of the main particle. If so, update the fine-tuning shadow particle corresponding to the maximum value among the multiple fine-tuning shadow fitness values ​​as the main particle, and perform optimization as a new adjustment direction until the preset optimization times are met, and use the main particle corresponding to the maximum value of the fitness value in the optimization process as the target particle; The configuration parameter set corresponding to the target particle is used as the position correction scanning configuration parameter set.

6. The method for correcting the position of the main bracket of a solar thermal power generation collector based on a laser profilometer according to claim 5, characterized in that: include: When the maximum value among the plurality of fine-tuning shadow fitnesses is less than the main particle fitness, obtaining a first random amount and a second random amount based on a random number generation function; When the first random amount is less than or equal to the second random amount, accepting the fine-tuning shadow particle corresponding to the maximum value among the multiple fine-tuning shadow fitnesses as the main particle, and taking it as a new adjustment direction for optimization; When the first random amount is greater than the second random amount, the fine-tuning shadow particle corresponding to the maximum value of the multiple fine-tuning shadow fitnesses is rejected to update the main particle, and the adjustment scheme for obtaining the multiple fine-tuning shadow particles is disabled, and the optimization is still performed with the main particle as the adjustment direction until the preset optimization times are met, and the main particle corresponding to the maximum value of the fitness in the optimization process is used as the target particle, and the configuration parameter set corresponding to the target particle is used as the position correction scanning configuration parameter set.

7. The method for correcting the position of a main bracket of a solar thermal power generation collector based on a laser profiler according to claim 1, characterized in that: include: Randomly extracting a first single-point deviation value from the multiple single-point deviation values, wherein the first single-point deviation value has a first deviation direction identifier; Acquire a first neighborhood set of the first single point deviation value, wherein the first neighborhood set includes a plurality of first neighborhood single point deviation values ​​and a plurality of first neighborhood deviation direction identifiers; Aggregate a plurality of scanning points corresponding to a plurality of first neighborhood single point deviation values ​​whose first neighborhood deviation direction identifier is in the same direction as the first deviation direction identifier and whose difference between the first neighborhood single point deviation value and the first single point deviation value is within a preset difference range and a first scanning point corresponding to the first single point deviation value, to generate a first deviation area; Calculate the average of the single-point deviation values ​​in the first deviation area to generate a first area deviation value, and use the direction corresponding to the first deviation direction identifier as the first area deviation direction; The multiple single-point deviation values ​​and the multiple deviation direction identifiers are fused in the same direction as the neighbors to obtain multiple deviation areas.

8. The method for calibrating the position of a main bracket of a solar thermal power generation collector based on a laser profiler according to claim 1, characterized in that: include: obtaining a plurality of sample area deviation values, a plurality of sample area deviation directions, and a plurality of sample position correction schemes as construction data; The encoder and decoder are supervisedly trained using the constructed data until convergence, thereby obtaining the trained position correction unit.

Images