CN108304524B - The lightweight webpage method for visualizing and system of extensive fire dynamic smog field - Google Patents

Abstract

The present invention relates to the lightweight webpage method for visualizing and system of a kind of extensive fire dynamic smog field; the described method comprises the following steps: 1) raw flue gas data are carried out light-weight technologg using voxelization mode by server end, obtain lightweight flue gas data；2) browser end receives the lightweight flue gas data and is rendered, and realizes real-time webpage visualization.Compared with prior art, the present invention has many advantages, such as automation, actualization, precision, lightweight, it can be achieved that the extensive smog field real-time visual of webpage grade.

Description

Lightweight webpage visualization method and system for large-scale fire dynamic smoke field

technical field

The invention relates to a fire scene simulation technology, in particular to a lightweight web page visualization method and system for a large-scale fire dynamic smoke field.

Background technique

The data volume of dynamic smoke generated by fire is huge, and the shape changes are complicated when it diffuses, which makes the amount of calculation and storage of smoke visualization rendering very large, so the real-time visualization of dynamic smoke field has always been a problem. Even if it is a stand-alone version, it still requires huge hardware resources to visualize smoke in large-scale scenes. However, large-scale smoke visualization on the web is more limited in terms of hardware memory, and the rendering capability of the webpage is far lower than that of a single computer. As a result, the visualization of lightweight large-scale fire smoke scenes has not been solved for a long time.

At present, there are various forms of smoke expression. From the beginning of Stam J et al., the research on the spread of smoke has been carried out. Zhu B et al. in the United States have used adaptive grids to highlight the details of smoke spread. Tang Yong of Yanshan University has used the Euler method and GPU acceleration realizes a real-time and effective smoke fusion dynamic spread model on the PC machine, which can accurately simulate the multi-fire source smoke spread process in the real world. The above research focuses on the calculation process of smoke spread, but the method for lightweight visualization of smoke in the scene has not yet appeared.

In the visualization of flue gas, due to the irregular dynamic changes of flue gas, scholars at home and abroad have successively proposed models based on process texture functions, models based on fractal geometry, models based on cellular automata, models based on physics, and models based on particle systems.

Based on the process texture function model, it is not convenient to simulate the external force. Based on the fractal geometry model, several rules can be defined, and the self-similarity of infinite regression is used to simulate the smoke spreading process. The disadvantage is that the fidelity and accuracy are low. The cellular automata model is a model proposed by von Neumann and Ulan in 1950. It uses the state of grid cells at a certain moment and the state of neighboring grid cells to fill the smoke. The structure is simple, but the combination effect complex. Based on physics models, for example, Jos Stam proposed a method to describe the gas phenomenon and its propagation by the diffusion process based on the laws of thermodynamics. Although the calculation accuracy of smoke spread is high, the algorithm of this model is very complicated. Particle-based and voxel-based visualization methods are mainstream. This is because the particle system is recognized as one of the most successful methods for simulating irregular objects, using primitives to define the volume of objects instead of polygons. However, in the particle-based method, it is necessary to determine the position and visualization of each smoke particle, which occupies a large amount of computing resources.

Contents of the invention

The object of the present invention is to provide a lightweight webpage visualization method and system for large-scale fire dynamic smoke fields in order to overcome the above-mentioned defects in the prior art.

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

A lightweight webpage visualization method for large-scale fire dynamic smoke field, including the following steps:

1) The server uses voxelization to lighten the original smoke data to obtain lightweight smoke data;

2) The browser receives the light-weight flue gas data for rendering to realize real-time web page visualization.

Preferably, the lightweight treatment specifically includes:

101) Obtain the packaged original smoke data calculated by the fire dynamics simulation tool;

102) dividing the fire scene space into three-dimensional matrix voxelized scenes, and then converting the original smoke data into original voxelized smoke data;

103) Perform de-redundancy, data normalization and data de-duplication processing on the original voxelized smoke data in sequence to obtain lightweight smoke data.

Preferably, the acquisition process of the original smoke data is:

The fire dynamics simulation tool sets different fire source points that are easy to catch fire, and performs fire simulation at the fire source point according to the combustible material in the actual situation, and uses the fire fluid dynamics algorithm to calculate the dynamic spread of fire smoke, so as to obtain the smoke distribution. Spread process data to form encapsulated raw smoke data.

Preferably, in step 103), the redundancy processing is specifically:

Remove the data whose smoke concentration is 0 and smoke concentration data less than 0.00001 in the original voxelized smoke data.

Preferably, in step 103), the data normalization process is specifically:

Step301: For the flue gas data after deduplication, compare the size of the flue gas data to obtain the largest flue gas data max and the smallest flue gas data min;

Step302: Calculate Δ=max-min to obtain 10 levels of smoke data segment datasets, these 10 smoke data segments are Δ/10, 2Δ/10, 3Δ/10, 4Δ/10, 5Δ/10, 6Δ /10, 7Δ/10, 8Δ/10, 9Δ/10 and Δ;

step303: Normalize all smoke data to the above 10 levels.

Preferably, in step 103), the data deduplication processing is specifically:

Step311: According to the position of a certain normalized smoke data, compare whether the smoke data is at the same level as the surrounding smoke data, and record the smoke at the same level at the position adjacent to this position gas data, and record the new gas position;

Step312: Take the newly added smoke position as an object, continue to traverse the smoke data of the same level around it, and record the position of the smoke data if there is smoke data of the same level;

Step313: Repeat step312 until there is no smoke data of the same level around, so as to obtain a position data set of smoke data of the same level;

Step314: Take the location data set as a whole.

Preferably, in step 2), the webpage visualization includes smoke visualization and poisonous gas visualization.

Preferably, step 2) also includes:

Perform texture particle visualization simulation during rendering to realize image texture visualization.

The present invention also provides a lightweight webpage visualization system for a large-scale fire dynamic smoke field using the lightweight webpage visualization method.

Compared with the prior art, the present invention provides an automatic, realistic, accurate and lightweight method for real-time visualization of large-scale smoke fields at the web page level, which saves hardware resources and smoke rendering resources, and has the following beneficial effects:

①Automation: The present invention can automatically carry out lightweight work on the heavyweight original flue gas data, and can carry out lightweight processing on the heavyweight flue gas data, so as to automatically realize the data processing of flue gas automation that needs to be solved urgently at present. Among them, in the data deduplication process, the position data group of the same level of smoke data is used as the data matrix when the smoke is visualized, which can reduce the number of voxel smoke drawing.

② Realization: No matter the visualization effect of smoke data or the need for other calculations, it needs to have very high requirements for the authenticity of smoke, so that real fire and smoke can be simulated in the scene The present invention can guarantee the authenticity of the smoke data.

③Precise: The present invention not only needs to be able to ensure the authenticity of the smoke data, but also to make the smoke have precise attributes, so that in the virtual reality fire escape, accurate data can be obtained, so as to carry out accurate path planning and complete effective escape navigation.

④ Lightweight: At present, there is no method for reducing the weight of flue gas. The present invention can present lightweight flue gas data visualization on the webpage, and effectively reduce the weight of the current heavyweight flue gas data, thereby realizing flue gas visualization core needs. The present invention can further reduce the cache load and calculation load of smoke scene rendering by means of smoke texture pictures.

Description of drawings

Fig. 1 is a schematic flow sheet of the present invention;

Fig. 2 is a schematic diagram of a scene model of a double-deck subway station in an embodiment;

Fig. 3 is the concentration value of the flue gas in the double-deck subway station after lightweighting in the embodiment;

Figure 4 is a visual rendering of lightweight voxelized smoke;

Figure 5 is a visual effect diagram of lightweight voxelized poisonous gas;

Fig. 6 is a visual scene effect diagram based on texture particle smoke;

Figure 7 is an effect diagram of the poisonous gas visualization scene based on texture particles.

Detailed ways

The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is carried out on the premise of the technical solution of the present invention, and detailed implementation and specific operation process are given, but the protection scope of the present invention is not limited to the following embodiments.

The present invention provides a lightweight webpage visualization method for large-scale fire dynamic smoke field, comprising the following steps:

1) The server uses voxelization to lighten the original smoke data to obtain lightweight smoke data. The lightweight treatment is specifically:

101) Obtain the packaged original smoke data calculated by the fire dynamics simulation tool, the acquisition process of the original smoke data is:

The fire dynamics simulation tool sets different fire source points that are easy to catch fire, and performs fire simulation at the fire source point according to the combustible material in the actual situation, and uses the fire fluid dynamics algorithm to calculate the dynamic spread of fire smoke, so as to obtain the smoke distribution. Spread process data to form encapsulated raw smoke data.

102) Divide the fire scene space into three-dimensional matrix voxelized scenes, and then transform the original smoke data into original voxelized smoke data.

This voxelization method can divide the space into 3D matrix voxelized scenes. The more detailed the division, the more detailed the voxel blocks obtained, but the computer resources and time it takes more. Considering the needs of the visualization effect of the human eye, this embodiment can set the smoke visualization voxel at 0.25m, which not only greatly reduces the amount of rendering calculation, but also reduces the burden on the memory.

103) Perform de-redundancy, data normalization and data de-duplication processing on the original voxelized smoke data in sequence to obtain lightweight smoke data.

Flue gas data deduplication steps:

step1: Input the original flue gas data;

Step2: Remove the data whose smoke concentration is 0, and remove the smoke data whose smoke concentration data is less than 0.00001;

Step3: Prepare the obtained flue gas data for normalization processing of the flue gas data.

Smoke data normalization steps:

Step1: For the flue gas data after deduplication, compare the size of the flue gas data, obtain the largest flue gas data max, and obtain the smallest flue gas data (usually selected as min=0.00001);

Step2: Calculate Δ=max-min, and then obtain 10 levels of smoke data segment datasets, these 10 smoke data segments are Δ/10, 2Δ/10, 3Δ/10, 4Δ/10, 5Δ/10 ,6Δ/10,7Δ/10,8Δ/10,9Δ/10,Δ.

Step3: Complete the normalization process of the above 10 paragraphs for all the flue gas data.

Smoke data deduplication steps:

Step1: According to the position of the normalized smoke data, compare whether the smoke data is at the same level as the surrounding smoke data, and record the smoke data at the same level at the position adjacent to this position , and record the new smoke position;

Step2: Take the newly added smoke position as an object, continue to traverse the smoke data of the same level around it, and record the position of the smoke data if there is smoke data of the same level;

Step3: Repeat step2 until there is no smoke data of the same level around, so as to obtain a position data set of smoke data of the same level;

Step4: The position data group is used as a whole smoke data, and used as a data matrix when the smoke is visualized, which can reduce the number of voxel smoke drawing.

2) The browser receives the light-weight smoke data for rendering to realize real-time webpage visualization, including smoke visualization and poisonous gas visualization.

In some embodiments, step 2) further includes: performing texture particle visualization simulation during rendering to realize image texture visualization.

The above method breaks through three difficulties in the light weight of flue gas: first, it performs light weight processing on the flue gas data completely automatically, and does not need to manually light weight the flue gas data every time, and the data volume is large, and manual light weight processing is not necessary. Quantification is also unrealistic; at the same time, after obtaining reliable and real smoke data, the method undergoes a series of lightweight processing, and finally obtains lightweight smoke data, which is accurate; on this basis, further The light-weight smoke data is applied to the actual crowd escape in the subway station, and real-time smoke visualization can be realized.

The present invention also provides a lightweight webpage visualization system for a large-scale fire dynamic smoke field using the lightweight webpage visualization method.

The above method is illustrated by taking the smoke simulation of a double-deck subway station as an example. As shown in Figure 2, firstly, according to the CAD electronic data map of the scene, the scene model in FDS is built, and the scene model has the characteristics of high precision. After the smoke is lightened, the voxelized smoke formed has different smoke concentrations in different voxel blocks. Figure 3 shows the concentration of the lightened smoke in the subway station. In Figure 3, the first three numbers in each row of data represent the coordinates corresponding to x, y, and z voxels in the scene, and the latter values represent the smoke concentration at the position at each time point. The intervals are of the same length.

1) First of all, it can be seen from the data that the change of the concentration data in the scene is between 0-3. In addition, most of the data are below 0.1, and during the simulation process, the data below 0.1 is almost invisible of. Therefore, the data can be discretely stratified accordingly. In this set of data, the data is divided into 16 stages: 0-0.1, 0.1-0.3, 0.3-0.5, 0.5-0.7, 0.7-0.9..., 2.7-3. The first stage is the data that is not rendered, that is, when the data is less than 0.1, it is treated as 0. Only after the data exceeds 0.1 does the concentration time change start to occur.

2) When creating a block, if the concentration of the current block is 0 (when the smoke concentration ρ<0.1), do not create a block at this position, and create a block at the corresponding position when the concentration is greater than 0.1 Model. In order to ensure that the number of models in the scene is as small as possible. Through the above optimization algorithm, the voxel-based fire scene can run more smoothly.

After the binary conversion of the smoke concentration, it can be quickly drawn and rendered in the scene. By stripping the smoke data in the smoke data, separate smoke data can be obtained, and then the smoke can be visualized and rendered in the scene. The rendered scene is shown in Figure 4.

The visualization of smoke is based on the actual scene, the smoke concentration, and the visualization effect completed by computer. The tobacco poison is invisible but is also spreading in a concrete way. Tobacco data obtained according to FDS can obtain voxel-based lightweight tobacco gas through the lightweight method in Figure 1.

Since poisonous smoke gas is often invisible, it is more harmful and poisonous gas is more lethal. Therefore, by visually presenting the toxic gas, the position where the smoke spreads to and the concentration of the smoke at that position can be seen at any time in space and time, and the path planning of the escaped personnel can be carried out according to these conditions. Guide the escape behavior of the escaped personnel.

According to the smoke data obtained by FDS, the toxic gas can be stripped out, and according to the original heavyweight data of the toxic gas, it can be light-weighted. The gas concentration can be visualized and rendered as shown in Figure 5.

Before calculating the smoke, FDS gives the type of toxic gas released by the specific combustibles. For example, there will be co,so ₂ in a fire, so the specific smoke and poison stripping algorithm is as follows:

step1: Input the original FDS flue gas data;

Step2: Decompress the original FDS flue gas data, complete the format conversion, and obtain the semantic file;

Step3: Capture all the smoke data whose attribute is poisonous gas in the semantic file, build a new semantic file, and save the semantic file;

Step4: According to specific needs, specific toxic gases (such as: CO, SO ₂ ) can be separated from it;

step5: Output the required smoke gas.

Based on the visualization goal of the light weight of the smoke, the present invention further carries out the visualization work of the smoke based on the texture particle, that is, the position of the smoke calculated by FDS, and the concentration data of the smoke, and the smoke texture image is used In this way, the visual simulation of texture particles is carried out, so that the drawing method of voxelized smoke concentration is transformed into the rendering method of image texture, and the effect is shown in Figure 6 and Figure 7. Using this method can further reduce the smoke scene Rendered cache load as well as compute load.

For the first time, the present invention performs voxel-based lightweight processing on heavyweight smoke data, forming a lightweight effect that reduces the amount of data by 300 times. It also ensures the accuracy of the numerical range of the smoke concentration and the location of the smoke, so that the authenticity of the smoke based on FDS is also very well maintained.

The present invention applies light-weight smoke data to the webpage for the first time, and realizes the effect of visualization and precise pathfinding on the webpage, providing a scientific basis and a feasible visualization scheme for crowd escape under the fire smoke scene .

For the first time, the present invention realizes a series of lightweight processing methods for heavyweight smoke data on the server through coding, and completes lightweight processing of heavyweight smoke data on web pages (including mobile Internet pages) through lightweight technology. The rendering and presentation technology of lightweight smoke is realized on the web page of the client.

The present invention realizes the lightweight processing and visual presentation of smoke in a large-scale scene for the first time. Combined with the visualization of large-scale scenes, accurate smoke visualization is carried out, so that in the large-scale scene where the smoke is on fire, the smoke can be presented without error. The dynamic spreading process of air.

The present invention realizes for the first time the visualization of smoke data on the Web based on texture particles in a large-scale scene after lightweight processing, and also combines the visualization of large-scale scenes to perform accurate smoke visualization, which can present the spread of smoke in a more lightweight manner the process of.

The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make many modifications and changes according to the concept of the present invention without creative efforts. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning or limited experiments on the basis of the prior art shall be within the scope of protection defined by the claims.

Claims (6)

1. a lightweight webpage visualization method of large-scale fire dynamic smoke field, is characterized in that, comprises the following steps: 1) The server uses voxelization to lighten the original smoke data to obtain lightweight smoke data; 2) The browser receives the lightweight flue gas data for rendering to realize real-time web page visualization; The lightweight treatment is specifically: 101) Obtain the packaged original smoke data calculated by the fire dynamics simulation tool; 102) dividing the fire scene space into three-dimensional matrix voxelized scenes, and then converting the original smoke data into original voxelized smoke data; 103) Perform de-redundancy, data normalization, and data deduplication processing on the original voxelized smoke data in sequence to obtain light-weight smoke data, perform binary conversion on the smoke concentration, and then perform step 2); The de-redundancy processing is specifically as follows: Remove the data whose smoke concentration data is less than 0.00001 in the original voxelized smoke data; After the smoke is lightweight, voxelized smoke is formed. Different voxel blocks have different smoke concentrations. When rendering, the texture particle visualization simulation is performed to realize the visualization of the image texture. When creating a voxel block, when When the smoke concentration is greater than 0.1, create a voxel block model at the corresponding position. 2. the lightweight webpage visualization method of large-scale fire dynamic smoke field according to claim 1, is characterized in that, the acquisition process of described original smoke data is: The fire dynamics simulation tool sets different fire source points that are prone to fire, performs fire simulation at the fire source point according to combustible materials, and uses the fire fluid mechanics algorithm to calculate the dynamic spread of fire smoke, so as to obtain the smoke spread process data, Form the packaged raw flue gas data. 3. the lightweight webpage visualization method of large-scale fire dynamic smoke field according to claim 1, is characterized in that, in step 103), data normalization processing is specifically: Step301: For the flue gas data after deduplication, compare the size of the flue gas data to obtain the largest flue gas data max and the smallest flue gas data min; Step302: Calculate Δ=max-min to obtain 10 levels of smoke data segment datasets, these 10 smoke data segments are Δ/10, 2Δ/10, 3Δ/10, 4Δ/10, 5Δ/10, 6Δ /10, 7Δ/10, 8Δ/10, 9Δ/10 and Δ; step303: Normalize all smoke data to the above 10 levels. 4. The lightweight webpage visualization method of the large-scale fire dynamic smoke field according to claim 3, characterized in that, in step 103), the data deduplication process is specifically: Step311: According to the position of a certain normalized smoke data, compare whether the smoke data is at the same level as the surrounding smoke data, and record the smoke at the same level at the position adjacent to this position data, and record the corresponding position of the smoke data as the new smoke position; Step312: Take the newly added smoke position as an object, continue to traverse the smoke data of the same level around it, and record the position of the smoke data if there is smoke data of the same level; Step313: Repeat step312 until there is no smoke data of the same level around, so as to obtain a position data set of smoke data of the same level; Step314: Take the location data set as a whole. 5. The lightweight webpage visualization method of large-scale fire dynamic smoke field according to claim 1, characterized in that, in step 2), the webpage visualization includes smoke visualization and poisonous gas visualization. 6. A lightweight webpage visualization system for a large-scale fire dynamic smoke field applying the lightweight webpage visualization method according to claim 1.