A method and device for rendering a cloud pattern of an instrument of an aircraft simulator

Abstract

The application discloses a kind of cloud stripe dynamic rendering method and device of aircraft simulator instrument, belong to aircraft simulator, data processing and artificial intelligence technical field.The method includes: based on the real-time data of aircraft simulator, utilize self-attention mechanism to generate the dynamic weight of each physical quantity for cloud stripe rendering;Utilize the dynamic weight of each physical quantity, calculate the characteristic parameter of cloud stripe;Utilize characteristic parameter, carry out cloud stripe rendering and display on the instrument of aircraft simulator.The application can carry out cloud stripe rendering in instrument, utilize cloud stripe to characterize flight scene state, utilize intuitive visual feature to eliminate pilot to carry out the multi-step abstract conversion cost of numerical value mapping into physical state, physical state mapping into flight situation, pilot can perceive the change of flight environment from instrument instantaneously, when facing emergency flight state, pilot can utilize intuitive feeling to make flight decision quickly and accurately, so as to improve user experience.

Description

Technical Field

[0001] This invention relates to the fields of aircraft simulators, data processing, and artificial intelligence, and particularly to a method and apparatus for dynamic rendering of cloud patterns on aircraft simulator instruments. Background Technology

[0002] In aircraft simulator training, pilots perceive the flight scenario status through values ​​from backup high-altitude, high-speed instruments. However, in emergency flight situations, this method cannot provide quick and accurate flight decisions, impacting the user experience. Summary of the Invention

[0003] This invention provides a method and apparatus for dynamically rendering cloud-like patterns on aircraft simulator instruments. The technical solution is as follows:

[0004] On the one hand, a method for dynamically rendering cloud patterns on aircraft simulator instruments is provided, the method comprising:

[0005] Based on real-time data from an aircraft simulator, a self-attention mechanism is used to generate dynamic weights for each physical quantity when rendering cloud patterns.

[0006] The characteristic parameters of the cloud pattern are calculated by using the dynamic weights of each physical quantity; the characteristic parameters include at least one of density, fragmentation, edge sharpness and color offset.

[0007] Using the aforementioned feature parameters, cloud-like patterns are rendered and displayed on the instruments of the aircraft simulator.

[0008] On the other hand, a dynamic rendering device for cloud patterns on an aircraft simulator instrument is provided, the device comprising:

[0009] The generation unit is used to generate dynamic weights of various physical quantities for cloud pattern rendering based on real-time data from an aircraft simulator using a self-attention mechanism.

[0010] The calculation unit is used to calculate the characteristic parameters of the cloud pattern using the dynamic weights of various physical quantities; the characteristic parameters include at least one of density, fragmentation, edge sharpness and color offset.

[0011] The rendering unit is used to render and display cloud patterns on the instruments of the aircraft simulator using the aforementioned feature parameters.

[0012] On the other hand, a computer device is provided, the computer device including a memory and a processor, the memory for storing computer programs, and the processor for executing the computer programs stored in the memory to implement the steps of the above-described method for dynamic rendering of cloud patterns on aircraft simulator instruments.

[0013] On the other hand, a computer-readable storage medium is provided, wherein a computer program is stored therein, and when the computer program is executed by a processor, the steps of the above-described method for dynamic rendering of cloud patterns in aircraft simulator instruments are implemented.

[0014] On the other hand, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the above-described method for dynamically rendering cloud patterns on aircraft simulator instruments.

[0015] The technical solution provided by this invention can bring at least the following beneficial effects:

[0016] When the aircraft simulator is in the process of simulating flight, real-time data from the simulator is acquired, and a self-attention mechanism is used to generate dynamic weights for each physical quantity used in cloud pattern rendering. The dynamic weights are then used to calculate the feature parameters of the cloud pattern, which are then used to render and display the cloud pattern on the instrument panel. This solution uses cloud pattern rendering in the instrument panel to represent the flight scene state. The intuitive visual features eliminate the multi-step abstraction and transformation costs for pilots, which involve mapping numerical values ​​to physical states and then to flight situations. Pilots can instantly perceive changes in the flight environment from the instrument panel. When facing emergency flight situations, pilots can make quick and accurate flight decisions based on their intuitive perception, thereby improving the user experience. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a cloud-pattern dynamic rendering method for an aircraft simulator instrument provided by an embodiment of the present invention;

[0019] Figure 2 This is a flowchart of a method for generating dynamic weights according to an embodiment of the present invention;

[0020] Figure 3 This is a structural diagram of a cloud-pattern dynamic rendering device for an aircraft simulator instrument provided in an embodiment of the present invention;

[0021] Figure 4 This is a hardware architecture diagram of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0023] As mentioned earlier, aircraft simulators are equipped with a backup high-speed altitude instrument, which provides numerical information about the current flight altitude and speed. During simulator training, pilots need to map the instrument readings to the current flight conditions to understand the flight scenario. However, this mapping process increases the user's perception time, leading to reaction delays. In emergency situations, this can reduce the pilot's decision-making efficiency and negatively impact the user experience.

[0024] Based on this, the inventive concept of the present invention is to render cloud patterns in the instrument to represent the flight scene status. During flight, the cloud patterns are constantly changing, so that users can intuitively feel the flight scene status based on the changes in cloud patterns displayed on the instrument. When facing an emergency flight situation, the pilot can make a quick and accurate flight decision based on the intuitive feeling, thereby improving the user experience.

[0025] The specific implementation of the above concept is described below.

[0026] Please refer to Figure 1 The present invention provides a method for dynamic rendering of cloud patterns on aircraft simulator instruments, the method comprising:

[0027] Step 100: Based on real-time data from the aircraft simulator, a self-attention mechanism is used to generate dynamic weights for each physical quantity during cloud pattern rendering.

[0028] Step 102: Calculate the characteristic parameters of the cloud pattern using the dynamic weights of each physical quantity; the characteristic parameters include at least one of density, fragmentation, edge sharpness, and color offset.

[0029] Step 104: Using the aforementioned feature parameters, cloud-like patterns are rendered and displayed on the instruments of the aircraft simulator.

[0030] In this embodiment of the invention, when the aircraft simulator is in the process of simulated flight, real-time data from the aircraft simulator is acquired, and a self-attention mechanism is used to generate dynamic weights for each physical quantity used in cloud pattern rendering. The dynamic weights are then used to calculate the feature parameters of the cloud pattern, which are then used to render and display the cloud pattern on the instrument panel. This solution renders cloud patterns on the instrument panel, using cloud patterns to represent the flight scene state. The intuitive visual features eliminate the multi-step abstraction and transformation costs for pilots, which involve mapping numerical values ​​to physical states and then to flight situations. Pilots can instantly perceive changes in the flight environment from the instrument panel. When facing emergency flight situations, pilots can make quick and accurate flight decisions based on their intuitive perception, thereby improving the user experience.

[0031] The following description Figure 1 The execution method of each step is shown.

[0032] First, for step 100, based on real-time data from the aircraft simulator, a self-attention mechanism is used to generate dynamic weights for each physical quantity when rendering cloud patterns.

[0033] To provide pilots with a more realistic experience during flight simulation training and improve scene perception efficiency, an instrument for real-time display of cloud patterns can be added to the aircraft simulator, or the backup altimeter in the aircraft simulator can be upgraded. In one embodiment, to ensure the instrument clearly displays dynamic cloud patterns, it can be an electronic display screen.

[0034] In this embodiment of the invention, the real-time data from the aircraft simulator is calculated using the model. In a real aircraft, this data can be collected by sensors. The physical quantities include at least altitude, pressure gradient, and temperature. These physical quantities are used to calculate the cloud pattern feature parameters. These cloud pattern feature parameters include at least one of density, fragmentation, edge sharpness, and color offset.

[0035] In one embodiment of the present invention, the physical quantities have fixed weights, which are pre-set based on atmospheric physical laws and flight scenario characteristics. Specifically, the fixed weights of each physical quantity are as follows:

[0036] 1. Height

[0037] According to the International Standard Atmospheric Model (ISA), cloud vertical density and height have the following exponential relationship:

[0038]

[0039] in, The density of clouds at altitude h; Cloud density at sea level, =7.5km is the altitude scale; for every 1km increase in altitude, cloud density decreases by about 37%.

[0040] Since the cruising altitude of commercial airliners is generally around 10,000 meters, the cloud pattern density at this altitude varies much more with altitude than with factors such as air pressure and temperature. Therefore, altitude becomes the dominant factor determining cloud pattern characteristics and should be given a high weight. In one implementation, altitude has a fixed weight W during the cruising phase. H It is 70%.

[0041] 2. Pressure gradient

[0042] According to NACA (National Advisory Committee for Aeronautics) TN-3457 experimental data, the pressure gradient and cloud fragmentation exhibit the following linear relationship:

[0043]

[0044] Where F represents cloud fragmentation. This represents the air pressure gradient.

[0045] For every 1 hPa / km increase in the pressure gradient, the fragmentation increases by 85%. Due to the drastic changes in low-altitude air pressure during takeoff / landing, in one implementation, the pressure gradient has a fixed weight W during takeoff / landing. P The weighting is 25%. Takeoff / landing is a high-risk period for flight accidents (accounting for about 70%). Pilots need to quickly assess low-altitude weather conditions. Therefore, a fixed weight of 25% makes cloud fragmentation more sensitive to changes in air pressure. For example, when the air pressure gradient increases sharply, cloud patterns will show a more fragmented shape in real time, providing a direct indication of turbulence risk.

[0046] 3. Temperature

[0047] Based on the WMO Cloud Atlas ice crystal formation theory, the sharpness of cloud edges at low temperatures has the following functional relationship with temperature:

[0048]

[0049] Where S is the edge sharpness of the cloud, which determines the visual clarity of the cloud pattern, and its value is 0-1; T is the ambient temperature. Temperature is the activation threshold for sharpness. This is the coefficient for the maximum sharpness amplitude; This is the temperature response intensity coefficient; Humidity modulation index; Relative humidity; index term The humidity term is used to characterize the temperature-driven sharpness growth function. A factor used to characterize the effect of dryness on sharpness.

[0050] The physical mechanisms of edge sharpness and temperature are analyzed below.

[0051] The temperature core functions as follows:

[0052] When T < Supercooled water droplets freeze, and ice crystals grow rapidly through sublimation.

[0053] When T = At this point, the edge sharpness S is approximately 0, and there is no sharpening.

[0054] When T = At -10℃, edge sharpness S≈ ,like =0.4, then S≈0.98 At this point, the edge sharpness is close to saturation.

[0055] The humidity modulation effect is as follows:

[0056] The ambient humidity (RH) determines the deposition and sublimation rates of ice crystals.

[0057] If the RH is low, it is net sublimation, and the ice crystals have sharper edges;

[0058] If the RH is high, then it is net sublimation, and the ice crystal boundaries are passivated.

[0059] For humidity item If RH = 100%, then the humidity value is 0, the edge sharpness is 0, and there is no edge; if RH = 30%, then... =0.3, then the humidity value is approximately 1.25, and the edge sharpness is improved by 25%.

[0060] Based on the above analysis, the fixed weights of each physical quantity under different flight stages can be obtained:

[0061] Table 1: Fixed weights of various physical quantities at different flight stages

[0062]

[0063] Considering that fixed weights suffer from rigidity and lag in highly complex and rapidly changing flight environments, in order to improve the realism, immersion, and situational awareness efficiency of simulation training, in this embodiment of the invention,

[0064] Fixed weights are based on idealized and averaged physical models of the atmosphere, such as the International Standard Atmosphere, to describe the relationship between cloud patterns and altitude, gradient pressure, and temperature under normal circumstances. However, considering the real flight environment, which is prone to sudden turbulence, thunderstorms, and flight over complex terrain, atmospheric conditions change drastically and nonlinearly. In such cases, relying solely on fixed weights based on average patterns to calculate cloud pattern characteristic parameters can lead to distortion or slow response.

[0065] Based on this, in this embodiment of the invention, it is necessary to dynamically adjust the weights corresponding to the physical quantities used to calculate the feature parameters of cloud patterns.

[0066] Specifically, in one implementation, please refer to [link / reference]. Figure 2 The dynamic weights of physical quantities are determined as follows:

[0067] Step 200: Extract scene demand features from the real-time data, normalize the scene demand features, and use a fully connected layer to map the normalized scene demand features into a query vector; the scene demand features include turbulence intensity, weather warning level, and flight phase status;

[0068] Turbulence intensity can be calculated using the rate of change of air pressure and the amplitude of altitude fluctuations. Weather warning levels can be derived from the weather model built into the aircraft simulator; for example, weather warning levels can be divided into 1-4 levels. Flight phase states can include takeoff, climb, cruise, and landing, which can be expressed using one-hot encoding; for example, the cruise phase can be [0,0,1,0].

[0069] To avoid attention bias caused by different levels of scenario requirements, one implementation method can use the Min-Max Normalization method to normalize the scenario requirements.

[0070] Assuming the original scenario requirement characteristics are x, the normalization formula is:

[0071]

[0072] in, The normalized scenario requirements characteristics, and These represent the minimum and maximum values ​​in the corresponding scenario's requirement characteristics, respectively.

[0073] It should be noted that the flight phase status can be directly encoded using one-hot encoding without the need for normalization.

[0074] When mapping the normalized scenario requirement features to the query vector Q, the mapping can be achieved using the following formula:

[0075]

[0076] in, The weight matrix is ​​a learnable matrix; This is the bias vector, used to adjust the feature distribution.

[0077] Step 202: Standardize the physical quantities and extract physical features from the standardized physical quantities using a one-dimensional convolutional neural network, and generate a key vector based on the physical features; the physical quantities include at least altitude, air pressure gradient and temperature;

[0078] In one embodiment of the present invention, each physical quantity can be processed according to the following standardized processing formula:

[0079] Highly standardized formula:

[0080]

[0081] in, The height is the standardized value, where h is the height in feet.

[0082] Pressure gradient normalization formula:

[0083]

[0084] in, This is the standardized pressure gradient. This represents the rate of change of air pressure over time, expressed in hPa / min.

[0085] Temperature standardization formula:

[0086]

[0087] in, T represents the standardized temperature.

[0088] In one implementation, the one-dimensional convolutional neural network (1D-CNN) is used to extract local correlation features from the standardized physical quantities. The CNN structure can be:

[0089] Kernel size: You can choose 3 kernels with sizes of 3, 5, and 7 respectively. Smaller kernels can capture local details, while larger kernels can capture a wider range of contextual information.

[0090] Number of convolutional layers: Two convolutional layers can be set. The first convolutional layer has 32 output channels, and the second convolutional layer has 64 output channels, gradually increasing the feature dimension and improving the model's ability to extract complex features.

[0091] After two convolutional operations, the feature maps output by different convolutional kernels can be concatenated along the channel dimension, and then the feature maps can be compressed into a one-dimensional vector through global average pooling, finally obtaining a 64-dimensional key vector K.

[0092] Step 204: Determine the fixed weight matrix based on the flight phase state, and initialize the fixed weight matrix as a value vector;

[0093] In this embodiment of the invention, a fixed weight matrix can be used as the initial value vector V.

[0094] Step 206: Substitute the query vector, the key vector, and the value vector into the self-attention formula to calculate the dynamic weights of each physical quantity.

[0095] In one implementation, the self-attention formula can be:

[0096]

[0097] in, The dimension of the key vector, for example, 64 dimensions.

[0098] In this embodiment of the invention, the generation process of dynamic weights is the system's intelligent decision-making process. In this intelligent decision-making process, a query vector is constructed to determine which factors need to be considered to render the cloud patterns that best reflect the current scene; a key vector is constructed to represent the current physical environment; and a value vector is initialized to represent what operation needs to be performed.

[0099] The following example illustrates the process of generating the dynamic weights described above.

[0100] The aircraft simulator was in the cruise phase when it suddenly entered a clear-air turbulent zone. At this moment:

[0101] Physical quantities: Altitude 30,000 feet, air pressure change rate 0.8 hPa / min, temperature -45℃; it should be noted that the air pressure change rate is the pressure gradient perceived by the aircraft during flight. P / The time representation produced by the combined effect of h and aircraft flight.

[0102] The scenario requirements are as follows: turbulence intensity, calculated as strong based on altitude fluctuations, with a normalized value of 0.8; weather warning level, currently no thunderstorms, warning level is low, with a normalized value of 0.1; flight phase status, cruise phase, one-hot encoding is [0,0,1,0].

[0103] Based on Table 1 above, the fixed weight matrix W during the cruise phase can be obtained. fixed for:

[0104] W fixed = [W H W P W T ] = [0.7, 0.2, 0.1]

[0105] Among them, W H W P W T These are weighted by altitude, pressure gradient, and temperature, respectively.

[0106] When generating dynamic weights, the normalized scene requirement features are first concatenated to obtain a multi-dimensional vector, and then mapped using a trained fully connected layer to obtain the query vector Q.

[0107] Next, the physical quantities (altitude = 30,000 feet, air pressure change rate = 0.8 hPa / min, temperature -45℃) are standardized to obtain the standardized physical quantities [3.0, 0.8, 0133]. The standardized physical quantities are then input into a one-dimensional convolutional neural network. The one-dimensional convolutional neural network will learn local and combined features such as "the air pressure gradient of 0.8 is a very high peak" and "the altitude is very high and the temperature is extremely low", and thus output a key vector K.

[0108] Then, the fixed weights of the cruise phase can be directly used as the initial value vector, i.e., V. initial = [0.7, 0.2, 0.1].

[0109] Finally, self-attention calculations are used to adjust the weights based on current needs and physical state.

[0110] In the calculation process using the above self-attention formula The matching score is used to calculate the match score between a query (e.g., focusing on turbulence) and (the current high pressure gradient). Since Q focuses on turbulence, and the most prominent feature in K is the high pressure gradient, the matching score for the pressure gradient part will be very high. This involves transforming the scores into a probability distribution. Weights corresponding to higher air pressure gradients receive a greater probability of attention, while altitude and temperature receive lower probabilities. Based on this, multiplying by V enables the weighted initial value vector V of the attention probability distribution, so that the main components of the output are based on the pressure gradient weight (0.2) of the initial value vector and are significantly adjusted.

[0111] In this embodiment of the invention, after weighting and adjustment using a self-attention mechanism, a new weight vector is output, for example, W. dynamic= [0.5, 0.4, 0.1]. This indicates that under the current strong turbulence, the influence of the pressure gradient on cloud patterns should be doubled, changing from 0.2 to 0.4, because drastic pressure fluctuations are a direct indicator of turbulence and cloud breakup. Simultaneously, the dominance of altitude is appropriately weakened, changing from 0.7 to 0.5, because altitude is now stable and not the primary factor causing change. The temperature weight remains unchanged.

[0112] In this embodiment of the invention, through the Transformer attention mechanism, the model can analyze the current specific and unconventional meteorological data combination in real time and dynamically adjust the weight of each physical quantity on the cloud pattern, so that the rendered cloud pattern can accurately match the current special and non-standard weather phenomena, such as the clear structure of the edges of rapidly generated broken cumulus clouds and thunderstorm clouds.

[0113] Then, for step 102, the characteristic parameters of the cloud pattern are calculated using the dynamic weights of each physical quantity.

[0114] In this embodiment of the invention, the characteristic parameters of the cloud pattern include at least one of density, fragmentation, edge sharpness, and color offset.

[0115] In one embodiment of the present invention, dynamic weights can be directly used as the final weights to calculate cloud pattern feature parameters. By using dynamic weights to represent cloud patterns, it is possible to adjust the weights in real time through a self-attention mechanism when facing sudden situations such as turbulence or severe weather, rendering extreme cloud patterns with a sense of oppression, fragmentation, and dynamism. This can greatly highlight and amplify the current abnormal weather conditions, attracting the pilot's attention.

[0116] In another embodiment of the present invention, considering that using only dynamic weights may be subject to noise interference, causing cloud patterns to flicker and jitter meaninglessly and frequently, which does not conform to the inertia of real atmospheric changes and may also cause visual fatigue and distract the pilot, the dynamic weights and fixed weights of each physical quantity can be fused together before calculating the feature parameters of the cloud patterns to obtain the final weights; the final weights can then be used to perform the calculation of the feature parameters of the cloud patterns.

[0117] In one implementation, fusion can be achieved in the following way:

[0118]

[0119] in, For the final weight, For dynamic weights, For fixed weights, This is a scene adaptation factor.

[0120] It should be noted that the scene adaptation factor can be dynamically adjusted based on radar echo intensity. This is because radar echo intensity is a key indicator characterizing atmospheric convection intensity, cloud thickness, and turbulence risk. Higher radar echo intensity indicates more intense convection, a higher risk of turbulence or icing, and a greater potential flight safety hazard. The value of the scene adaptation factor directly determines the fusion ratio of dynamic and fixed weights.

[0121] Specifically, under clear-sky conditions where the radar echo intensity is less than 20 dBZ, The value ranges from 0.2 to 0.4;

[0122] Under cumulus cloud conditions with radar echo intensity of 20-40 dBZ, The value ranges from 0.5 to 0.7;

[0123] Under conditions of strong convection or icing where radar echo intensity is greater than 40 dBZ The value ranges from 0.8 to 1.0.

[0124] The radar echo intensity is standard data provided by the meteorological simulation model built into the aircraft simulator.

[0125] In this embodiment of the invention, fixed weights and dynamic weights are fused to calculate the feature parameters of cloud patterns using the final fused weights. Thus, both fixed and dynamic weights can influence the final representation of the cloud patterns. Fixed weights ensure that during standard, stable flight phases, cloud pattern rendering strictly adheres to physical laws, providing pilots with stable and predictable visual feedback, thereby reinforcing their basic understanding of standard procedures and the environment. Dynamic weights endow the system with a keen responsiveness in complex and sudden special scenarios, amplifying the visual representation of key physical quantities (such as drastic changes in air pressure) in real time, thereby highlighting abnormal states and forcing the pilot's attention to focus.

[0126] By dynamically merging the two weights using scene adaptation factors, the system can achieve the following adaptive switching: under normal conditions, fixed weights dominate to optimize training efficiency and cognitive habits; under special scenarios, dynamic weights dominate to emphasize threat perception and emergency response training. This fusion mechanism ensures that the overall evolution of cloud patterns is smooth, coherent, and consistent with the macroscopic trends of atmospheric changes.

[0127] After calculating the final weights, in one embodiment of the present invention, the feature parameters of the cloud pattern can be calculated using the following formula:

[0128]

[0129]

[0130]

[0131]

[0132] Where D is density; F is fragmentation degree; The height change rate is expressed in units of 1000 ft / min and is obtained from real-time data; S represents edge sharpness. This is the color offset. This refers to the change in temperature. is the change in air pressure; L is the ambient light intensity, obtained from real-time data.

[0133] In this embodiment of the invention, the characteristic parameters of cloud patterns are calculated using the above calculation formula, and then the characteristic parameters are used to render and display dynamic cloud patterns on the instrument.

[0134] Furthermore, in order to ensure a natural and smooth transition of cloud patterns during stable flight, and to quickly adjust the cloud pattern shape when the environment changes abruptly, so as to reflect the changes in flight status in a timely manner, in one embodiment of the present invention, before rendering cloud patterns, it may also include: performing time-dimensional smoothing transition processing on the feature parameters of cloud patterns.

[0135] In one implementation, a smooth transition can be achieved using the following method:

[0136]

[0137] in, This is the matrix of feature parameters ultimately used for rendering in frame t (the current frame) after smooth transition processing. This is a matrix formed by calculating the feature parameters using dynamic weights; It is a matrix formed by the smoothed output of the feature parameters in the (t-1)th frame (the previous frame); This is the mixing ratio coefficient.

[0138] The mixing ratio coefficient can be dynamically adjusted according to the urgency of the flight scenario.

[0139] In one implementation, when the turbulence intensity is below the first threshold, there is no weather warning, and the flight is in a stable cruise / climb phase, the flight scenario is determined to be in a normal transition, and the mixing ratio coefficient can be 0.3. When the turbulence intensity is above the first threshold, the weather warning level is greater than the specified level, the air pressure change rate is greater than the second threshold, and the flight is in a critical takeoff / landing phase, the flight scenario is determined to be in an emergency transition, and the mixing ratio coefficient can be 0.7.

[0140] This avoids flickering or abrupt changes in cloud patterns caused by parameter jumps, resulting in a smooth transition during normal flight and a rapid response in emergency situations. It ensures the physical realism of cloud pattern rendering and further improves visual comfort and contextual awareness efficiency.

[0141] In one embodiment of the present invention, the cloud pattern rendering can be used as an instrument background. While retaining the abstract digital / pointer readings, it adds an intuitive visual form that conforms to the laws of atmospheric physics. This allows the pilot to perceive the overall state of the flight environment, such as smoothness, turbulence, and icing risk, from the overall visual texture of the instrument panel without having to complete the multi-step logical conversion of "numerical value → weather condition → flight status" in their mind.

[0142] Furthermore, during training using aircraft simulators, pilots can be provided with a high-fidelity, immersive instrument environment, whereby the instruments themselves become an important source of scene information and an enhancer of immersion.

[0143] This invention, through upgrading a simple backup instrument into a miniature atmospheric condition visualization system based on a physical model and driven by artificial intelligence, fundamentally changes the information interaction mode between pilots and instruments, transforming it from passive reading and calculation to active perception and understanding, effectively improving training efficiency, flight safety, and situational awareness.

[0144] Please refer to Figure 3 This invention provides a dynamic rendering device for cloud patterns on aircraft simulator instruments, the device comprising:

[0145] The generation unit 300 is used to generate dynamic weights of various physical quantities for cloud pattern rendering based on real-time data from an aircraft simulator and using a self-attention mechanism.

[0146] The calculation unit 302 is used to calculate the characteristic parameters of the cloud pattern using the dynamic weights of each physical quantity; the characteristic parameters include at least one of density, fragmentation, edge sharpness and color offset.

[0147] The rendering unit 304 is used to render and display cloud patterns on the instruments of the aircraft simulator using the aforementioned feature parameters.

[0148] In one embodiment of the present invention, the generation unit is specifically used to perform the following operations:

[0149] Scene demand features are extracted from the real-time data and normalized. The normalized scene demand features are then mapped to query vectors using a fully connected layer. The scene demand features include turbulence intensity, weather warning level, and flight phase status.

[0150] The physical quantities are standardized, and physical features are extracted from the standardized physical quantities using a one-dimensional convolutional neural network. A key vector is generated based on the physical features. The physical quantities include at least altitude, air pressure gradient, and temperature.

[0151] Determine the fixed weight matrix based on the flight phase status, and initialize the fixed weight matrix as a value vector;

[0152] Substituting the query vector, the key vector, and the value vector into the self-attention formula, the dynamic weights of each physical quantity are calculated.

[0153] In one embodiment of the present invention, the device may further include: a fusion unit, configured to fuse the dynamic weights of each physical quantity with fixed weights to obtain a final weight; and to use the final weight to perform the calculation of the feature parameters of the cloud pattern.

[0154] In one embodiment of the present invention, the dynamic weights and fixed weights of each physical quantity are fused using the following formula:

[0155]

[0156] in, For the final weight, For dynamic weights, For fixed weights, This is a scene adaptation factor.

[0157] In one embodiment of the present invention, when the feature parameters include density, fragmentation, edge sharpness, and color offset, the feature parameters are calculated using the following formula:

[0158]

[0159]

[0160]

[0161]

[0162] Where D is density; W H W P W T These are weighted by altitude, pressure gradient, and temperature, respectively. , , These represent the standardized altitude, pressure gradient, and temperature; F represents the degree of fragmentation. S represents the rate of change of height; S represents the edge sharpness. This is the color offset. This refers to the change in temperature. L represents the change in air pressure; L represents the ambient light intensity.

[0163] In one embodiment of the present invention, the device may further include: a smoothing processing unit, used to perform time-dimensional smoothing transition processing on the feature parameters of the cloud pattern, so as to trigger the rendering unit to perform cloud pattern rendering using the feature parameters obtained after the smoothing transition processing.

[0164] It should be noted that the cloud-pattern dynamic rendering device for aircraft simulator instruments provided in the above embodiments is only an example illustrating the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. Furthermore, the cloud-pattern dynamic rendering device for aircraft simulator instruments provided in the above embodiments and the cloud-pattern dynamic rendering method embodiments for aircraft simulator instruments belong to the same concept, and their specific implementation process is detailed in the method embodiments, which will not be repeated here.

[0165] Embodiments of this application also provide a computer device, please refer to... Figure 4 The computer device includes a processor and a memory, the memory storing at least one instruction, at least one program, code set or instruction set, the at least one instruction, at least one program, code set or instruction set being loaded and executed by the processor to implement the cloud pattern dynamic rendering method for aircraft simulator instruments provided in the above method embodiments.

[0166] The embodiments of this application also provide a computer-readable storage medium storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, at least one program, code set, or instruction set is loaded and executed by a processor to implement the cloud pattern dynamic rendering method for aircraft simulator instruments provided in the above-described method embodiments.

[0167] Embodiments of this application also provide a computer program product, which includes a computer program. A processor of a computer device reads the computer program from a computer-readable storage medium and executes the computer program, causing the computer device to perform the cloud pattern dynamic rendering method for aircraft simulator instruments as described in any of the above embodiments.

[0168] For ease of description, the above systems or devices are described separately as various modules or units based on their functions. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware components.

[0169] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0170] Finally, it should be noted that in this document, relational terms such as first, second, third, and fourth are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0171] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for dynamically rendering cloud-like patterns on aircraft simulator instruments, characterized in that, The method includes: Based on real-time data from an aircraft simulator, a self-attention mechanism is used to generate dynamic weights for each physical quantity when rendering cloud patterns. The characteristic parameters of the cloud pattern are calculated by using the dynamic weights of each physical quantity; the characteristic parameters include at least one of density, fragmentation, edge sharpness and color offset. Using the aforementioned feature parameters, cloud-like patterns are rendered and displayed on the instruments of the aircraft simulator; The real-time data from the aircraft simulator utilizes a self-attention mechanism to generate dynamic weights for each physical quantity during cloud rendering, including: Scene demand features are extracted from the real-time data and normalized. The normalized scene demand features are then mapped to query vectors using a fully connected layer. The scene demand features include turbulence intensity, weather warning level, and flight phase status. The physical quantities are standardized, and physical features are extracted from the standardized physical quantities using a one-dimensional convolutional neural network. A key vector is generated based on the physical features. The physical quantities include at least altitude, air pressure gradient, and temperature. Determine the fixed weight matrix based on the flight phase status, and initialize the fixed weight matrix as a value vector; Substituting the query vector, the key vector, and the value vector into the self-attention formula, the dynamic weights of each physical quantity are calculated.

2. The method according to claim 1, characterized in that, Before calculating the feature parameters of the moiré pattern, the following is also included: The dynamic weights of each physical quantity are fused with the fixed weights to obtain the final weights; the final weights are then used to perform the calculation of the feature parameters of the cloud pattern.

3. The method according to claim 2, characterized in that, The dynamic weights and fixed weights of each physical quantity are combined using the following formula: in, For the final weight, For dynamic weights, For fixed weights, This is a scene adaptation factor.

4. The method according to claim 1, characterized in that, When the feature parameters include density, fragmentation, edge sharpness, and color offset, the feature parameters are calculated using the following formula: Where D is density; W H W P W T These are weighted by altitude, pressure gradient, and temperature, respectively. , , These represent the standardized altitude, pressure gradient, and temperature; F represents the degree of fragmentation. S represents the rate of change of height; S represents the edge sharpness. This is the color offset. This refers to the change in temperature. L represents the change in air pressure; L represents the ambient light intensity.

5. The method according to any one of claims 1-4, characterized in that, Before performing cloud pattern rendering, the method further includes: performing a time-dimensional smoothing transition processing on the feature parameters of the cloud pattern, so as to use the feature parameters obtained after the smoothing transition processing to perform the cloud pattern rendering.

6. A dynamic rendering device for cloud patterns on an aircraft simulator instrument, characterized in that, The device includes: The generation unit is used to generate dynamic weights of various physical quantities for cloud pattern rendering based on real-time data from an aircraft simulator using a self-attention mechanism. The calculation unit is used to calculate the characteristic parameters of the cloud pattern using the dynamic weights of various physical quantities; the characteristic parameters include at least one of density, fragmentation, edge sharpness and color offset. The rendering unit is used to render and display cloud patterns on the instruments of the aircraft simulator using the aforementioned feature parameters. The generation unit is specifically used to perform the following operations: Scene demand features are extracted from the real-time data and normalized. The normalized scene demand features are then mapped to query vectors using a fully connected layer. The scene demand features include turbulence intensity, weather warning level, and flight phase status. The physical quantities are standardized, and physical features are extracted from the standardized physical quantities using a one-dimensional convolutional neural network. A key vector is generated based on the physical features. The physical quantities include at least altitude, air pressure gradient, and temperature. Determine the fixed weight matrix based on the flight phase status, and initialize the fixed weight matrix as a value vector; Substituting the query vector, the key vector, and the value vector into the self-attention formula, the dynamic weights of each physical quantity are calculated.

7. A computer device, characterized in that, The computer device includes a memory and a processor. The memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory to implement the steps of the method according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the method described in any one of claims 1-5.

9. A computer program product, characterized in that, Includes a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1-5.

Images