A method for calculating and optimizing the performance of a specular reflection type dynamic photovoltaic facade

Abstract

The present invention discloses a method for calculating and optimizing the efficiency of a mirror-reflective dynamic photovoltaic facade. The calculation and optimization method includes building a simulation model in a computer according to a real model, and inputting sunlight, a three-dimensional model, materials, and the model movement mode into the GPU for calculation. Using ray tracing calculation technology, the GPU is utilized to accelerate the calculation of the sunlight ray reflection behavior, and the photovoltaic efficiency of the mirror-reflective type is simulated and calculated. Single result calculation. Based on the calculation methods of steps one, two, and three, according to the set model movement mode, a multi-level subdivision method is used to optimize the movement scheme of real-time variable nodes, and the optimal solution of the dynamic building model at any moment is searched for. The genetic algorithm is used to optimize the building facade parameters, and the comprehensive optimal building facade scheme, movement model, and photovoltaic data are obtained. The data of sampling points at different times throughout the year are concatenated to obtain the optimal movement model and the optimal photovoltaic efficiency. The calculation and optimization method of the present invention solves the problem that the solar radiation calculation scheme cannot calculate the dynamic reflection model.

Description

Technical Field

[0001] The present invention relates to the technical field of solar curtain walls, and specifically to a method for calculating and optimizing the efficiency of a mirror reflection type dynamic photovoltaic facade. Background Art

[0002] The building facade curtain wall is the external wall enclosure of a building, which does not bear weight and can have a certain displacement ability relative to the main structure or has a certain deformation ability itself to achieve specific functions and does not bear the action of the main structure. It is the building's peripheral enclosure structure.

[0003] Combining photovoltaic technology with curtain walls to form photovoltaic curtain walls has become an effective way to solve energy problems. However, for the new type of dynamic photovoltaic energy-saving building facade curtain wall system that uses mirror reflection, there is currently no effective energy efficiency calculation method. At the same time, there are some drawbacks in the traditional building solar photovoltaic energy efficiency calculation scheme: 1. Most solar radiation calculation methods cannot calculate light reflection, and calculating reflection requires a large amount of time; 2. Solar radiation calculation can only calculate the solar radiation of a static model (or certain specific motions, such as always facing the sun); 3. The solar radiation calculation method cannot adaptively adjust the scheme according to the calculation results, and the relationship between calculation and adjustment is fragmented.

[0004] For the dynamic building photovoltaic facade that uses mirror reflection, there is currently no reliable scheme to achieve photovoltaic efficiency calculation, motion model solution, and design scheme optimization. Therefore, a method for calculating the efficiency of a mirror reflection type dynamic photovoltaic facade and an optimization scheme are proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for calculating and optimizing the efficiency of a mirror reflection type dynamic photovoltaic facade, to solve the problem that the traditional solar radiation calculation scheme cannot calculate the dynamic reflection model. This calculation and optimization method has a wide adaptability and a fast calculation speed. Through the combination of calculation and optimization, it provides calculation support for the energy efficiency calculation, motion solution, and scheme optimization of the dynamic reflection type building photovoltaic system.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A method for calculating and optimizing the efficiency of a mirror reflection type dynamic photovoltaic facade, the calculation and optimization method includes the following steps:

[0008] Step 1: Build a simulation model in the computer according to the real model, and input information such as sunlight, 3D model, material, and model motion mode into the GPU for calculation.

[0009] Step 2: Based on the data input into the GPU in Step 1, use ray tracing calculation technology to accelerate the calculation of sunlight ray reflection behavior using the GPU.

[0010] Step 3: Single result calculation: transfer the GPU calculation results in step 2 to the CPU for calculation; according to the relevant calculation formula, the photovoltaic panel radiation efficiency data under the set time and motion position is obtained.

[0011] Step 4: Based on the calculations from Steps 1, 2, and 3, and the pre-defined model motion, a multi-level subdivision method is used to optimize the motion scheme of the real-time variable nodes, finding the optimal solution for the dynamic building model at any given moment. A genetic algorithm is used to optimize the building facade parameters, resulting in the optimal overall building facade scheme, motion model, and photovoltaic data.

[0012] Step 5: Connect the data of sampling points at different times throughout the year to obtain the optimal motion model and optimal photovoltaic efficiency.

[0013] Furthermore, the step 1 establishes a corresponding three-dimensional virtual model in the computer according to the model that needs to be calculated; for components that can move in real time, their movement mode and range are set; different types of material properties are assigned to different models, and information such as solar information, calculation accuracy, and calculation range are specified, and this information is transferred to the GPU for calculation preparation.

[0014] Furthermore, the object material is defined as ordinary material, reflective material and receiving material, and different light interaction behaviors are defined for the three materials: ordinary material will no longer bounce light after light enters, and its data will no longer be counted in the shader; reflective material will reflect light according to the law of light reflection after light enters, and energy will be attenuated according to the law of reflection; receiving material will not reflect light after light enters, but will count the light data into the shader, and at the same time pass the sunlight vector data, calculation accuracy data, and emission range data to the GPU.

[0015] Furthermore, the ray tracing calculation steps in step 2 are as follows:

[0016] S1: The light emitter emits a number of rays from the same direction as the sun toward the target object. After the rays are bounced and absorbed, if they hit the solar panel, the pixel color change data corresponding to the rays changed by the emitter is obtained; that is, from the position of the emission area, the image of the solar panel can be seen, which means that sunlight can pass through this position and hit the target.

[0017] S2: Data transmission, placing the model data in S1 into the GPU, preparing for calculation, generating the corresponding blank texture image according to the sampling accuracy for subsequent data recording; dividing the texture image according to the number of parallel computing cores of the computer graphics card.

[0018] S3: The light emission emits a number of rays parallel to the sunlight direction from a certain square area. The square area is perpendicular to the sunlight direction, and the projection of the object to be irradiated in the sunlight direction is contained within the square area. The ray distribution conforms to the pixel distribution, and the number of rays is determined by the set sampling precision.

[0019] S4: The ray tracing includes two parts of the loop of calculating the ray (Ray cast) and shading. The calculation of the ray solves whether each ray intersects with the object. If it intersects, the next calculation is carried out according to the material type of the object to be irradiated. If the material of the object to be irradiated is a reflective material, the reflection path, energy and other information of the next ray are calculated. If the material of the object to be irradiated is a common material type, the propagation of the ray ends. If the material of the object to be irradiated is a receiving material type, the propagation of the ray ends and it is marked to write the ray information into the texture of the next step.

[0020] The shading writes the calculation result of each time onto a texture image. The depth of the texture image is the number of ray reflections + 1. Each pixel represents each ray emitted. Finally, the rays that can irradiate the target object under different numbers of ray reflections are obtained.

[0021] Further, the ratio of the sum of the values of all pixels in each layer of the statistical result texture map in step three to the total number of pixels can obtain the ratio of the rays that can be received by the solar panel in the rays emitted by the emitter to the total emitted rays. Then, by multiplying the area of the emitter, the equivalent area of the rays that the solar panel can receive is obtained. Multiplying this area by the radiation intensity of the sun at this moment can obtain the radiation amount obtained by the solar panel in this calculation. The equivalent direct irradiation area calculation formula:

[0022]

[0023] Where S D is the equivalent area of the direct rays irradiating the solar panel, V D is the sum of all pixel values in the direct light layer part of the texture map, N is the total number of pixels in the texture map, S E is the area of the ray emitter. The equivalent reflected area calculation formula:

[0024]

[0025] Where S R is the equivalent area of the reflected rays irradiating the solar panel, V R is the sum of all pixel values in the reflected layer part of the texture map, N is the total number of pixels in the texture map, S E is the area of the ray emitter.

[0026] Further, for the optimization of the motion scheme of the real-time variable nodes, for the real-time dynamic model, in order to achieve the optimal solar radiation reception, it is necessary to clarify its all-weather motion model. According to the calculation scheme in Step 3, for the real-time dynamic model, a multi-level step-by-step subdivision method is adopted to calculate the optimal motion model parameters at any time.

[0027] For the same moment, according to the model motion mode and range set in the model in Step 1, the results of the model are calculated. By gradually increasing the accuracy, a better choice is made for higher-precision calculations, and finally the optimal variable at this moment is selected.

[0028] Further, for the optimization of non-real-time variable variables in Step 4, the parts that cannot be changed after completion are adopted, but they can be freely adjusted during the design calculation process. This calculation is to obtain the best values of these variables. On the basis of the optimization of the motion scheme of the real-time variable nodes, based on the all-weather optimal motion model, a genetic algorithm is used to adjust the non-real-time movable variables to find the parameters that can achieve the optimal model.

[0029] Further, in Step 5 during the optimization process in Step 4, for the real-time movable variables, the optimal variable values are recorded. Through full-day or annual interval sampling, the recorded results are concatenated to obtain the change curve of the real-time movable variables under the all-weather optimal condition, the corresponding optimal photovoltaic energy in this state, the corresponding three-dimensional model information, and the total annual solar radiation.

[0030] For non-real-time movable variables, the results of the schemes under different variable combinations are recorded, and different types of schemes are screened for users to choose.

[0031] Advantages of the present invention:

[0032] 1. The calculation method and optimization scheme of the present invention are based on the propagation of real light, can accurately restore the real light behavior to the greatest extent, have no limit on the number of reflections, can adapt to direct light to any number of reflections, and can also adapt to direct light to any number of reflections;

[0033] 2. The calculation method and optimization scheme of the present invention use the graphics card to accelerate the calculation. For this type of parallel calculation model, the calculation speed can be significantly improved compared with the CPU calculation, which provides the possibility for the optimization of the scheme variables and has a fast calculation speed;

[0034] 3. Due to the fast calculation speed of the calculation method and optimization scheme of the present invention, it is possible to continuously adjust the model parameters and calculate to find the optimal solution. Through the combination of calculation and optimization, the building facade scheme can be further adjusted and improved, providing calculation support for the energy efficiency calculation, motion solution, and scheme optimization of the dynamic reflective building photovoltaic system. Description of the Drawings

[0035] The present invention will be further described below with reference to the accompanying drawings.

[0036] Figure 1 is the flowchart of the calculation method and optimization method of the present invention;

[0037] Figure 2 is the definition and composition of the model of the present invention;

[0038] Figure 3 are different types of nodes and optimizations of the present invention;

[0039] Figure 4 is the schematic diagram of the ray tracing process of the present invention;

[0040] Figure 5 is the schematic diagram of the single reflection calculation of the present invention;

[0041] Figure 6 is the schematic diagram of the secondary sampling of the present invention;

[0042] Figure 7 is the variable optimization method of the step-by-step subdivision of the present invention. Specific embodiments

[0043] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0044] A method for calculating and optimizing the efficiency of a specular reflection type dynamic photovoltaic facade, as Figure 1 shown, the calculation and optimization method includes the following steps:

[0045] Step 1: Build a simulation model in the computer according to the real model, and transmit information such as sunlight, 3D model, material, and model movement mode to the GPU for calculation.

[0046] According to the model to be calculated, establish a corresponding 3D virtual model in the computer; for the components that can move in real time, set their movement modes and ranges; assign different types of material properties to different models, and specify information such as solar information, calculation accuracy, and calculation range, and transmit this information to the GPU for calculation preparation, as Figure 2 shown.

[0047] For object models, when calculating, all objects will be converted into triangular meshes composed of several vertices and passed into the GPU; for real-time movable models, the movable nodes of the model and their movement methods need to be set; for models that require optimized parameters, their optimized variables and their ranges need to be set, as Figure 3 shown.

[0048] For object materials, three basic types are defined: normal, reflect, and target. Different light interaction behaviors are defined for the three materials: for normal materials, there is no light bounce after light enters, and its data is no longer included in the shader; for reflect materials, after light enters, it will reflect according to the law of light reflection and undergo energy attenuation according to the law of reflection; for target materials, there is no reflection of light after light enters, but the light data will be counted into the shader, and at the same time, the sunlight vector data, calculation accuracy data, and emission range data will be passed into the GPU.

[0049] Step 2: Based on the data passed into the GPU in Step 1, use ray tracing calculation technology to accelerate the calculation of sunlight reflection behavior using the GPU, as Figure 4 shown. The ray tracing steps are as follows:

[0050] S1: As Figure 5 shown, the ray emitter (Emitter) emits a number of rays from the direction same as the sun direction towards the target object. After the rays bounce and are absorbed, if they hit the solar panel (Target), the color change data of the corresponding pixels of the emitter for the ray is obtained; that is, equivalently, from the position of the emission area, the image of the solar panel can be seen, meaning that sunlight can shine on the target through this position.

[0051] S2: Data transfer, place the model data in S1 into the GPU to prepare for calculation

[0052] First, generate a blank texture map according to the sampling accuracy (such as 128*128, with a depth of 3, a total of 2^14*3 sampling points) for subsequent data recording, and then divide the texture map, as Figure 6 shown, into 8*8*1 threads. (For the 1080Ti graphics card used in the test environment, it has 32 parallel computing cores, and on average each core is responsible for two batches of calculations. For different platforms and different models of graphics cards, the number of threads is different, and an appropriate number of threads can speed up the calculation).

[0053] S3: Ray emission, emit a number of rays parallel to the sunlight direction from a certain square area. The square area is perpendicular to the sunlight direction and the projection of the object to be shot in the sunlight direction is contained within the square area. The ray distribution conforms to the pixel distribution, and the number of rays is determined by the set sampling accuracy.

[0054] S4: Ray tracing consists of a loop with two parts: ray casting and shading. Ray casting calculates whether each ray intersects with an object. If there is an intersection, further calculations are performed based on the material type of the object being hit. If the material of the object being hit is a reflective material, information such as the reflection path and energy of the next ray is calculated. If the material of the object being hit is a common material type, the propagation of the ray ends. If the material of the object being hit is a receiving material type, the propagation of the ray ends and the ray information is marked to be written into the texture in the next step.

[0055] Shading writes the calculation result of each time onto a texture image. The depth of the texture image is the number of ray reflections + 1. Each pixel represents each ray emitted, and finally, the rays that can reach the target object under different numbers of reflections are obtained.

[0056] Step 3: Single - result calculation. Transfer the GPU calculation result in Step 2 to the CPU for calculation; obtain the photovoltaic panel radiation efficiency data at the set time and moving position according to the relevant calculation formula.

[0057] Statistically calculate the ratio of the sum of the values of all pixels in each layer of the result texture map to the total number of pixels, and the ratio of the rays that can be received by the solar panel among the rays emitted by the emitter can be obtained. Then, by multiplying the area of the emitter, the equivalent area of the light that the solar panel can receive is obtained. Multiplying this area by the radiation intensity of the sun at that moment, the radiation amount obtained by the solar panel in this calculation can be obtained. The formula for the equivalent direct - irradiation area is:

[0058]

[0059] Where S D is the equivalent area of the direct rays irradiating the solar panel, V D is the sum of all pixel values in the direct - light layer part of the texture map, N is the total number of pixels in the texture map, and S E is the area of the ray emitter. The formula for the equivalent reflected area is:

[0060]

[0061] Where S R is the equivalent area of the reflected rays irradiating the solar panel, V R is the sum of all pixel values in the reflected - light layer part of the texture map, N is the total number of pixels in the texture map, and S E is the area of the ray emitter.

[0062] Step 4: Based on the calculation method in Step 3, according to the set model movement method, use the multi - level subdivision method to perform iterative calculations repeatedly to find the optimal solution of the dynamic building model at any moment, such asFigure 7 As shown, the genetic algorithm is used to optimize the building facade schemes in different forms, and the comprehensive optimal building facade scheme, motion model and photovoltaic data are obtained. The process is as follows:

[0063] (1) Optimization of the motion scheme of real-time variable nodes

[0064] For the real-time dynamic model, in order to achieve the optimal solar radiation reception, it is necessary to know its all-weather motion model. According to the calculation scheme in Step 3, for the real-time dynamic model, a multi-level and step-by-step subdivision method is used to calculate the optimal motion model parameters at any time.

[0065] For the same moment, according to the model motion mode and range set in the model in Step 1, the results of the model are calculated. By gradually increasing the accuracy, a better calculation with higher accuracy is selected, and finally the optimal variable at this moment is selected.

[0066] (2) Optimization of non-real-time variable variables of the scheme

[0067] For non-real-time variable variables, some parts that cannot be changed after completion are adopted, but they can be freely adjusted during the design calculation process. This calculation is to obtain the best values of these variables. On the basis of optimizing the motion scheme of real-time variable nodes and based on the all-weather optimal motion model, the genetic algorithm is used to adjust non-real-time movable variables to find the parameters that can achieve the optimal model.

[0068] Step 5: Connect the data of sampling points at different times throughout the year to obtain the optimal motion model and the optimal photovoltaic efficiency

[0069] During the optimization process in Step 4, for real-time movable variables, record the optimal variable values. Through all-day or annual interval sampling methods, connect the recorded results to obtain the change curve of real-time movable variables under all-weather optimal conditions, the corresponding optimal photovoltaic energy in its state, the corresponding three-dimensional model information, and the total annual solar radiation.

[0070] For non-real-time movable variables, record the scheme results under different variable combinations, and screen out different types of schemes for users to choose.

[0071] In the description of this specification, the descriptions referring to terms such as "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0072] The foregoing has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and what is described in the above embodiments and the specification is only to illustrate the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will also have various changes and improvements, and these changes and improvements fall within the scope of the present invention claimed.

Claims

1. A method for calculating and optimizing the efficiency of a specular reflection type dynamic photovoltaic facade, characterized in that The described calculation and optimization method includes the following steps: Step 1: Build a simulation model in the computer according to the real model, and input information such as sunlight, 3D model, material, and model movement mode into the GPU for calculation; Step 2: Based on the data input into the GPU in Step 1, use ray tracing calculation technology to accelerate the calculation of sunlight ray reflection behavior using the GPU; Step 3: Single result calculation, transfer the GPU calculation result in Step 2 to the CPU for calculation; obtain the radiation efficiency data of the photovoltaic panel at the set time and movement position according to the relevant calculation formula; Step 4: Based on the calculation methods in Steps 1, 2, and 3, according to the set model movement mode, use the multi-level subdivision method to optimize the movement plan of the real-time variable nodes, and find the optimal solution of the dynamic building model at any time; use the genetic algorithm to optimize the building facade parameters to obtain the comprehensive optimal building facade scheme, movement model, and photovoltaic data; Step 5: Serialize the data of the sampling points at different times throughout the year to obtain the optimal movement model and the optimal photovoltaic efficiency.

2. The method for calculating and optimizing the efficiency of a specular reflection type dynamic photovoltaic facade according to claim 1, wherein In Step 1, a corresponding 3D virtual model is established in the computer according to the model to be calculated; for the components that can move in real time, set their movement mode and range; assign different types of material attributes to different models, and specify sun information, calculation accuracy, and calculation range information, and input these information into the GPU for calculation preparation.

3. A method for calculating and optimizing the performance of a specular reflection type dynamic photovoltaic facade according to claim 2, characterized in that, The material of the model is defined as ordinary material, reflective material, and receiving material; different light interaction behaviors are defined for the three materials: the ordinary material will no longer perform light reflection after the light enters, and its data will no longer be included in the shader; the reflective material will reflect according to the law of light reflection after the light enters, and the energy will decay according to the law of reflection; the receiving material will not reflect the light after the light enters, but will count the light data into the shader, and at the same time input the sunlight vector data, calculation accuracy data, and emission range data into the GPU.

4. A method for calculating and optimizing the efficiency of a specular reflection type dynamic photovoltaic facade according to claim 3, characterized in that The ray tracing calculation steps in Step 2 are as follows: S1: The ray emitter emits a number of rays from the direction same as the sun direction to the target object. After the rays bounce and are absorbed, if they hit the solar panel, the color change data of the corresponding pixel of the emitter for this ray is obtained; that is, equivalently, from the position of the emission area, the image of the solar panel can be seen, which means that the sunlight can irradiate the target through this position; S2: Data transmission, place the model data in S1 into the GPU, prepare for calculation, generate a corresponding blank texture image according to the sampling accuracy for subsequent data recording; divide the texture map according to the number of parallel computing cores of the computer graphics card; S3: Ray emission, emit a number of rays parallel to the sunlight direction from a certain square area. The projection of the object to be irradiated in the sunlight direction perpendicular to the sunlight direction is included in the square area. The ray distribution conforms to the pixel distribution, and the number of rays is determined by the set sampling accuracy; S4: Ray tracing includes two parts of the loop: ray casting and shading. Ray casting calculates whether each ray intersects with an object. If there is an intersection, further calculations are performed according to the material type of the irradiated object. If the material of the irradiated object is a reflective material, the reflection path and energy information of the next ray are calculated. If the material of the irradiated object is a common material type, the propagation of the ray ends. If the material of the irradiated object is a receiving material type, the propagation of the ray ends and the ray information is marked to be written into the texture in the next step. The shading writes the calculation result of each time onto a texture image. The depth of the texture image is the number of ray reflections + 1. Each pixel represents each ray emitted, and finally, the rays that can irradiate the target object under different numbers of ray reflections are obtained.

5. The method for calculating and optimizing the efficiency of a specular reflection type dynamic photovoltaic facade according to claim 4, characterized in that, The ratio of the sum of the values of all pixels in each layer of the statistical result texture map in step three to the total number of pixels can obtain the ratio of the rays that can be received by the solar panel among the rays emitted by the emitter to the total emitted rays. Then, by multiplying the area of the emitter, the equivalent area of the rays that the solar panel can receive can be obtained. Multiplying this area by the solar radiation intensity at that moment can obtain the radiation amount obtained by the solar panel in this calculation. The formula for the equivalent direct irradiation area is: Among which S D is the equivalent area of the direct light rays irradiated on the solar panel, V D is the sum of all pixel values of the direct light layer part in the texture map, N is the total number of pixels in the texture map, S E is the area of the ray emitter; the calculation formula for the equivalent reflection area: where S R is the equivalent area of the reflected light rays incident on the solar panel, V R is the sum of all pixel values of the reflective layer part in the texture map, N is the total number of pixels in the texture map, S E is the area of the ray emitter.

6. The method for calculating and optimizing the efficiency of a specular reflection type dynamic photovoltaic facade according to claim 5, characterized in that, For the optimization of the motion scheme of the real-time variable nodes in step four, for the real-time dynamic model, in order to achieve the optimal solar radiation reception amount, it is necessary to clarify its all-weather motion model. According to the calculation scheme in step three, for the real-time dynamic model, a multi-level hierarchical subdivision method is used to calculate the optimal motion model parameters at any time. For the same moment, according to the model motion mode and range set in the model in step one, the results of the model are calculated. By gradually increasing the accuracy, a higher-precision calculation is preferentially performed, and finally, the optimal variable at that moment is selected.

7. A method for calculating and optimizing the performance of a specular reflection type dynamic photovoltaic facade according to claim 6, characterized in that, For the optimization of non-real-time variable variables in step four, the parts that cannot be changed after completion are used, but they can be freely adjusted during the design calculation process. This calculation is to obtain the best values of these variables. Based on the all-weather optimal motion model and on the basis of the optimization of the motion scheme of the real-time variable nodes, a genetic algorithm is used to adjust the non-real-time movable variables to find the parameters that can achieve the optimal model.

8. A method for calculating and optimizing the performance of a specular reflection type dynamic photovoltaic facade according to claim 7, characterized in that, In step five, during the optimization process in step four, for the real-time movable variables, the optimal variable values are recorded. Through all-day or annual interval sampling methods, the recorded results are concatenated to obtain the change curve of the real-time movable variables under all-weather optimal conditions, the corresponding optimal photovoltaic energy in its state, the corresponding three-dimensional information of the model, and the total annual solar radiation amount. For non-real-time movable variables, the results of different variable combinations are recorded, and different types of solutions are selected for users to choose.

Images