Safety monitoring method and system for VR flight device

Abstract

The embodiment of the invention provides a safety monitoring method and system for a VR flight device, and the method comprises the steps: firstly obtaining a real-time operation data set of the VR flight device, and enabling the real-time operation data set to cover environment perception data and equipment state data; performing safety analysis processing on the real-time operation data set to generate a safety state evaluation result, performing dynamic safety strategy matching processing based on the safety state evaluation result to obtain a safety regulation and control instruction, finally adjusting operation parameters of the VR flight device according to the safety regulation and control instruction, and triggering early warning feedback operation. Therefore, the operation safety of the VR flight device can be comprehensively and dynamically monitored, potential risks can be found in time, effective regulation and control are carried out, and stable operation of equipment and user safety experience are guaranteed.

Description

Technical Field

[0001] The present application relates to the field of virtual reality technology, and in particular, to a safety monitoring method and system for a VR flight device. Background Art

[0002] With the continuous development of virtual reality (VR) technology, VR flight devices, as equipment that can provide users with immersive flight experiences, have been widely used in many fields such as entertainment, education, and training. However, there are many deficiencies in the safety monitoring of current VR flight devices.

[0003] Existing safety monitoring methods for VR flight devices are often static, using pre-set fixed standards to determine whether the device is safe. This method cannot be dynamically adjusted according to the real-time operating conditions of the VR flight device, resulting in an inability to make effective safety assessments in the face of sudden situations or subtle changes in the device's operating state.

[0004] In addition, when detecting possible safety problems, existing safety regulation means are relatively limited and lack pertinence. Usually, only an alarm is simply issued or the device operation is stopped, without formulating personalized regulation strategies according to the specific safety conditions. This not only may affect the user experience but also cannot guarantee the safety of the device and users to the greatest extent. Summary of the Invention

[0005] In view of this, the purpose of the present application is to provide a safety monitoring method and system for a VR flight device.

[0006] According to the first aspect of the present application, a safety monitoring method for a VR flight device is provided, which is applied to a VR service system. The method includes: Obtaining a real-time operation data set of the VR flight device, where the real-time operation data set includes environmental perception data and device status data; Performing safety analysis and processing on the real-time operation data set to generate a safety status evaluation result; Performing dynamic safety policy matching processing based on the safety status evaluation result to obtain a safety regulation instruction; Adjusting the operation parameters of the VR flight device according to the safety regulation instruction and triggering a warning feedback operation.

[0007] According to the second aspect of the present application, a VR service system is provided. The VR service system includes a processor and a readable storage medium. The readable storage medium stores a program, and when the program is executed by the processor, the foregoing safety monitoring method for a VR flight device is implemented.

[0008] According to a third aspect of the present application, there is provided a computer-readable storage medium storing computer-executable instructions, which, when monitored to be executed, implement the aforementioned safety monitoring method for a VR flight device.

[0009] According to any of the above aspects, in the embodiments of the present application, by obtaining a real-time operation data set including environmental perception data and device status data, performing safety analysis and processing on the real-time operation data set, and generating a safety status evaluation result, it is possible to deeply mine potential safety risks behind the data, realize accurate evaluation of the operation safety of the VR flight device. Based on this safety status evaluation result, dynamic safety policy matching processing is performed to obtain a safety control instruction, and the most suitable response strategy can be flexibly matched according to different safety states, making safety management more targeted and dynamically adaptable. Adjusting the operation parameters of the VR flight device according to the safety control instruction and triggering an early warning feedback operation can timely respond to potential dangers, effectively reduce safety hazards during the operation of the VR flight device, and ensure the stable operation of the device and the safety experience of users. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0011] Figure 1 The flowchart of the safety monitoring method for a VR flight device provided by the embodiments of the present application is shown; Figure 2 The component structure diagram of the VR service system for implementing the aforementioned safety monitoring method for a VR flight device provided by the embodiments of the present application is shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0012] In order to enable those skilled in the art to better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0013] It should be noted that the terms "first", "second", etc. in the description, claims and the above-mentioned drawings of this application are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of this application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0014] Figure 1 The flowchart of the safety monitoring method for the VR flight device provided by the embodiment of this application is shown. It should be understood that in other embodiments, the order of some steps of the safety monitoring method for the VR flight device can be interchanged according to actual needs, or some of the steps can also be omitted or deleted. The detailed steps of the safety monitoring method for the VR flight device are introduced as follows.

[0015] Embodiment part: Step S110: Obtain the real-time operation data set of the VR flight device, where the real-time operation data set includes environmental perception data and device status data.

[0016] In this embodiment, for the VR flight device to ensure safety monitoring during the flight process, it is necessary to collect comprehensive and accurate operation data in real time. For the acquisition of environmental perception data, the VR flight device is equipped with various types of environmental sensors. When the lidar sensor works, it emits laser beams into the surrounding environment. According to the time and angle information of the reflected laser beams, the distance and azimuth between the surrounding objects and the flight device can be accurately calculated. Its scanning frequency is relatively high, and it can scan the surrounding environment multiple times in a short time to capture the dynamic changes of the environment. For example, in a relatively complex flight scenario, the lidar sensor continuously emits laser beams and may perform dozens or even hundreds of scans per second, forming a series of point cloud data, which reflects the spatial distribution of the surrounding objects.

[0017] The camera sensor obtains visual information by taking images of the flight scene. It has different resolution and frame rate settings to adapt to different flight environments and monitoring requirements. In a well-lit environment, a high-resolution camera can capture clear images, facilitating subsequent identification and analysis of objects in the environment. At the same time, the camera can also adopt a multi-angle shooting method to expand the field of view and reduce visual blind spots. For example, multiple cameras can be installed at different positions of the flight device to capture the surrounding environment from different angles, and then these images are stitched and fused to obtain a more comprehensive environmental image.

[0018] The meteorological sensor is used to measure the meteorological parameters of the environment, including temperature, humidity, wind speed, air pressure, etc. These meteorological parameters are crucial for the flight safety of the VR flight device. The temperature sensor measures the temperature by sensing the thermal effect of the surrounding air, and the humidity sensor measures the humidity by utilizing the adsorption and desorption characteristics of specific materials to water vapor. The wind speed sensor can measure the wind speed and direction using different working principles such as mechanical and ultrasonic, and the air pressure sensor obtains the air pressure information by measuring the change in atmospheric pressure. These meteorological sensors will transmit the measured data to the control system of the flight device in real time.

[0019] For the acquisition of device status data, various monitoring systems inside the flight device play a key role. The accelerometer utilizes Newton's second law to determine the acceleration change of the flight device by measuring the force generated by the mass block under the action of acceleration. It can accurately measure the acceleration components of the flight device in the three coordinate axis directions and has high sensitivity and response speed. During flight, the accelerometer continuously collects acceleration data, and these data can reflect the motion states of the flight device such as acceleration, deceleration, and turning.

[0020] The gyroscope is based on the principle of conservation of angular momentum and determines its attitude change by measuring the angular velocity of the flight device. It can real-time monitor the attitude information of the flight device such as pitch, roll, and yaw, and has high accuracy and stability. During flight, the gyroscope continuously outputs attitude data, and these data are crucial for the attitude control and navigation of the flight device.

[0021] The energy management system is responsible for monitoring the energy consumption of the flight device, including battery power, energy output power, etc. It calculates the remaining battery power and energy output power in real time by measuring parameters such as the voltage and current of the battery. At the same time, the energy management system can also optimize the distribution and use of energy to improve energy utilization efficiency. For example, when the flight device is in different flight modes, the energy management system will adjust the energy distribution ratio according to the actual demand to ensure that the flight device makes the most of energy under the premise of safety.

[0022] Integrating the above environmental perception data and device status data forms a real-time operation data set of the VR flight device.

[0023] Step S120: Perform security analysis and processing on the real-time operation data set to generate a security status evaluation result.

[0024] In this embodiment, in order to effectively perform security analysis and processing on the real-time operation data set of the VR flight device, it is necessary to further mine the key information in the data, extract representative features, and conduct comprehensive evaluation in combination with a pre-trained model.

[0025] Step S121: Extract environmental dynamic features from the environmental perception data and extract device operation features from the device status data.

[0026] In this embodiment, the environmental dynamic features can reflect the real-time changes in the flight environment, and the device operation features can reflect the operation status of the flight device itself. Accurately extracting these two types of features is a key step in security analysis.

[0027] Step S1211: Perform dynamic scene segmentation processing on the environmental perception data to generate multiple scene regions.

[0028] In this embodiment, performing dynamic scene segmentation on the environmental perception data is to divide the complex flight environment into relatively independent regions with different characteristics, so as to analyze the environmental situation more carefully. Scene segmentation can be based on spatial information and object distribution.

[0029] For the point cloud data obtained by lidar, a density-based clustering algorithm can be used for scene segmentation. This algorithm first defines a neighborhood radius and a minimum number of points threshold, and then starts from a point, searches for the points in its neighborhood. If the number of points in the neighborhood exceeds the minimum number of points threshold, these points are divided into a cluster. Then, continue to search and cluster the unprocessed points until all points are processed. In this way, the point cloud data can be divided into multiple clusters, and each cluster corresponds to a scene region.

[0030] For the image data captured by a camera, an image segmentation algorithm can be used for scene segmentation. For example, a semantic segmentation algorithm based on deep learning, which classifies each pixel in the image and divides the image into different semantic regions. These semantic regions can represent different scene elements, such as the ground, buildings, obstacles, etc. In practical applications, appropriate semantic categories can be selected according to specific requirements, and then the image is segmented into multiple scene regions.

[0031] When performing scene segmentation, the dynamic changes of the environment also need to be considered. For example, when the flying device is in flight, the surrounding objects may move or change, so the results of scene segmentation need to be updated in real time. The time window method can be adopted to perform scene segmentation on the environmental perception data within each time window, and then compare the segmentation results of adjacent time windows to re-segment and update the changed areas.

[0032] Step S1212: Perform obstacle recognition processing for each scene area to determine the obstacle distribution characteristics.

[0033] In this embodiment, after generating multiple scene areas, it is necessary to accurately identify the obstacles in each scene area to determine the obstacle distribution characteristics. For obstacle recognition in lidar point cloud data, a method based on feature extraction and classification can be adopted. First, extract the geometric features of each point from the point cloud data, such as curvature, normal vector, etc. Then, use these features to construct feature vectors, and through machine learning classification algorithms, such as support vector machines, random forests, etc., classify the points in the point cloud data into obstacle points and non-obstacle points. Next, cluster the points classified as obstacle points to obtain the point cloud set of each obstacle.

[0034] For obstacle recognition in camera image data, object detection algorithms can be adopted. Object detection algorithms can locate and identify different types of obstacles in the image and give their bounding box information. Common object detection algorithms include FasterR-CNN, YOLO, etc. based on deep learning. These algorithms learn the characteristics and patterns of different types of obstacles through training on a large amount of image data, so as to accurately detect obstacles in new images.

[0035] When determining the obstacle distribution characteristics, information such as the position, size, and shape of the obstacles needs to be considered. For the position of the obstacle, its center coordinates can be used to represent it. In lidar point cloud data, the centroid coordinates of the obstacle point cloud set can be calculated as the center coordinates of the obstacle; in camera image data, the center coordinates can be calculated according to the bounding box information of the obstacle. For the size of the obstacle, the length and width of its circumscribed rectangle can be used to represent it. In lidar point cloud data, the maximum and minimum coordinate values of the obstacle point cloud set in different coordinate axis directions can be calculated to obtain the length and width of its circumscribed rectangle; in camera image data, the length and width of its circumscribed rectangle can be directly obtained according to the bounding box information of the obstacle. For the shape of the obstacle, its contour information can be used to describe it. In lidar point cloud data, the contour information can be obtained by performing boundary extraction on the obstacle point cloud set; in camera image data, the contour information can be obtained by performing edge detection on the image within the bounding box of the obstacle.

[0036] By identifying and analyzing obstacles in each scene area, the distribution characteristics of obstacles in the entire flight environment can be obtained.

[0037] Step S1213: Perform feature calculation based on the obstacle distribution characteristics to generate the environmental dynamic characteristics; wherein, the environmental dynamic characteristics include obstacle density characteristics, dynamic change rate characteristics, and spatial distribution uniformity characteristics; the obstacle density characteristics are determined by statistically calculating the ratio of the number of obstacles in each scene area to the area of the scene area; the dynamic change rate characteristics are determined by calculating the change amplitude of the obstacle distribution within adjacent time windows; the spatial distribution uniformity characteristics are determined by calculating the position dispersion of obstacles in each scene area, and the dispersion is analyzed using the statistical variance method.

[0038] In this embodiment, after obtaining the obstacle distribution characteristics, it is necessary to further calculate the environmental dynamic characteristics to more comprehensively describe the dynamic changes in the flight environment.

[0039] For the calculation of the obstacle density characteristics, first, count the number of obstacles in each scene area. In the lidar point cloud data, the number of obstacles in each scene area can be obtained by counting the obstacle point cloud set; in the camera image data, the number of obstacles in each scene area can be obtained by counting the detected obstacles. Then, calculate the area of each scene area. For the scene area segmented from the lidar point cloud data, its area can be calculated according to its spatial coordinate information; for the scene area segmented from the camera image data, its actual area can be calculated according to the number of pixels in the image and the resolution of the image. Finally, divide the number of obstacles in each scene area by its area to obtain the obstacle density characteristics of the scene area.

[0040] For the calculation of the dynamic change rate characteristics, it is necessary to compare the obstacle distribution within adjacent time windows. First, determine the size of the time window, which can be set to 1 second or 2 seconds, for example. Then, calculate the distribution characteristics of obstacles within each time window, such as the position and number of obstacles. Next, compare the changes in the obstacle distribution characteristics within adjacent time windows and calculate their change amplitudes. For example, the average moving distance of the obstacle positions and the change rate of the number of obstacles within adjacent time windows can be calculated. These change amplitudes are used as components of the dynamic change rate characteristics.

[0041] For the calculation of the spatial distribution uniformity feature, it is necessary to calculate the position dispersion of obstacles in each scene area. First, take the central coordinates of the obstacles in each scene area as sample points. Then, calculate the mean coordinates of these sample points. Next, calculate the square of the distance between each sample point and the mean coordinates, and sum up the squares of the distances of all sample points. Finally, divide the sum result by the number of sample points to obtain the variance of the obstacle positions, that is, the spatial distribution uniformity feature.

[0042] Through the above calculations, the environmental dynamic features can be obtained, including the obstacle density feature, the dynamic change rate feature, and the spatial distribution uniformity feature. These features can be represented in the form of a vector, and each feature component corresponds to a specific feature value.

[0043] Step S1214: Perform time-series segmentation processing on the device status data to generate multiple device operation segments.

[0044] In this embodiment, in order to better analyze the device status data, it is necessary to perform time-series segmentation processing to divide the continuous device status data into multiple device operation segments with different features. The time-series segmentation can be performed according to the change trend and features of the device status data.

[0045] For the acceleration data collected by the accelerometer, a threshold-based segmentation method can be used. First, set an acceleration change threshold. Then, starting from the starting point of the acceleration data, compare the acceleration change values between adjacent data points. If the acceleration change value exceeds the threshold, it is considered that this point is a segmentation point, and the data is divided into two segments. Then, continue to process the subsequent data until all the data is segmented.

[0046] For the attitude data collected by the gyroscope, a clustering-based segmentation method can be used. Consider the attitude data as points in a high-dimensional space, and use a clustering algorithm to divide these points into different clusters. Each cluster corresponds to a device operation segment. For example, the K-Means clustering algorithm can be used to divide the attitude data into K clusters, where the value of K can be adjusted according to the actual situation.

[0047] When performing time-series segmentation, the actual situation of the device operation also needs to be considered. For example, when the flying device is in different flight phases such as takeoff, cruise, and landing, the device status data will change significantly. Therefore, the information of the flight phase can be combined to further adjust and optimize the results of the time-series segmentation.

[0048] Step S1215: Perform operation mode recognition processing for each device operation segment to determine the device operation mode feature.

[0049] In this embodiment, after generating multiple device operation segments, it is necessary to perform operation mode recognition processing on each segment to determine the operation mode characteristics of the device. The operation mode characteristics can reflect the operation states of the flying device at different stages.

[0050] For acceleration data, the operation mode can be recognized by analyzing the shape and characteristics of the acceleration curve. For example, when the acceleration curve shows a large positive value, it may indicate that the flying device is accelerating; when the acceleration curve shows a large negative value, it may indicate that the flying device is decelerating; when the acceleration curve approaches zero, it may indicate that the flying device is flying at a constant speed. Rules for operation mode recognition can be constructed based on characteristics such as the slope and peak value of the acceleration curve.

[0051] For attitude data, the operation mode can be recognized by analyzing the changes in the attitude angles. For example, when there are large changes in the pitch angle, it may indicate that the flying device is climbing or descending; when there are large changes in the roll angle, it may indicate that the flying device is turning. Rules for operation mode recognition can be constructed based on characteristics such as the change rate and change range of the attitude angles.

[0052] At the same time, the energy consumption data can also be combined to further confirm the operation mode. For example, when the flying device is accelerating, the energy consumption rate usually increases; when the flying device is flying at a constant speed, the energy consumption rate is relatively stable. By comprehensively analyzing the acceleration data, attitude data, and energy consumption data, the operation mode of the device can be more accurately recognized, and the corresponding operation mode characteristics can be determined.

[0053] Step S1216: Calculate the device load fluctuation index and the device energy consumption trend index according to the device operation mode characteristics.

[0054] In this embodiment, after determining the device operation mode characteristics, it is necessary to further calculate the device load fluctuation index and the device energy consumption trend index to evaluate the operation state and energy utilization of the device.

[0055] For the calculation of the device load fluctuation index, the load situation of the device can be reflected based on the acceleration data and the attitude data. First, calculate the standard deviation of the acceleration and the standard deviation of the attitude angles within each device operation segment. The standard deviation of the acceleration can reflect the amplitude of the acceleration change during the movement of the device, and the standard deviation of the attitude angles can reflect the amplitude of the change during the attitude adjustment of the device. Then, perform a weighted sum of the standard deviation of the acceleration and the standard deviation of the attitude angles to obtain the device load fluctuation index. The value of the weight can be adjusted according to the actual situation. For example, the weight can be determined based on the influence degree of the acceleration and the attitude on the device load.

[0056] For the calculation of the device energy consumption trend index, the energy consumption situation of the device can be analyzed based on the energy consumption data. First, calculate the average energy consumption rate within each device operation segment. Then, analyze the change in the average energy consumption rate between adjacent device operation segments. If the average energy consumption rate shows an upward trend, it indicates that the energy consumption of the device is increasing; if the average energy consumption rate shows a downward trend, it indicates that the energy consumption of the device is decreasing. The device energy consumption trend index can be obtained by calculating the difference in the average energy consumption rate between adjacent device operation segments.

[0057] Step S1217: Aggregate the device load fluctuation index and the device energy consumption trend index to generate the device operation characteristics.

[0058] In this embodiment, after calculating the device load fluctuation index and the device energy consumption trend index, these two indexes need to be aggregated to generate the device operation characteristics. The weighted splicing method can be used for aggregation processing.

[0059] First, determine the weights of the device load fluctuation index and the device energy consumption trend index. The value of the weight can be adjusted according to the actual situation. For example, the weight can be determined according to the influence degree of the device load fluctuation and the device energy consumption trend on the device operation state. Then, multiply the device load fluctuation index and the device energy consumption trend index by their corresponding weights respectively to obtain the weighted indexes. Finally, splice the weighted device load fluctuation index and the device energy consumption trend index to obtain the device operation characteristics. The device operation characteristics can be represented in the form of a vector, where one component corresponds to the device load fluctuation index and the other component corresponds to the device energy consumption trend index.

[0060] Step S122: Input the environmental dynamic characteristics and the device operation characteristics into a pre-trained safety assessment model to generate a comprehensive risk assessment index.

[0061] In this embodiment, after obtaining the environmental dynamic characteristics and the device operation characteristics, they need to be input into a pre-trained safety assessment model to generate a comprehensive risk assessment index. The comprehensive risk assessment index can comprehensively reflect the safety risk degree of the VR flight device in the current environment and operation state.

[0062] Step S1221: Perform the first normalization process on the environmental dynamic characteristics to obtain the normalized environmental characteristics.

[0063] In this embodiment, in order to make the environmental dynamic characteristics better adapt to the safety assessment model, it is necessary to perform the first normalization process on them. The normalization process can eliminate the dimensional difference between different characteristics and make the characteristics comparable.

[0064] The Z-score normalization method can be used to process the environmental dynamic features. First, calculate the mean and standard deviation of each feature component in the environmental dynamic feature vector. Then, for each feature component, subtract its mean and divide by the standard deviation to obtain the normalized feature component. By normalizing all the feature components in the environmental dynamic feature vector, the normalized environmental feature vector can be obtained.

[0065] Step S1222: Perform a second normalization process on the device operation features to obtain normalized device features.

[0066] In this embodiment, similarly, in order to enable the device operation features to be better input into the security assessment model, a second normalization process needs to be performed on them.

[0067] The same Z-score normalization method as the first normalization process is adopted. Calculate the mean and standard deviation of each feature component in the device operation feature vector, then subtract the mean from each feature component and divide by the standard deviation to obtain the normalized feature component. After normalizing all the feature components in the device operation feature vector, the normalized device feature vector can be obtained.

[0068] Step S1223: Input the normalized environmental features and the normalized device features into the feature interaction layer of the pre-trained security assessment model to generate an interaction feature vector.

[0069] In this embodiment, the feature interaction layer in the pre-trained security assessment model is used to fuse the normalized environmental features and the normalized device features to generate an interaction feature vector. The feature interaction layer adopts a cross-attention mechanism to achieve feature fusion.

[0070] The cross-attention mechanism first calculates the attention scores between the normalized environmental features and the normalized device features. For each feature component in the normalized environmental feature vector, calculate its similarity with each feature component in the normalized device feature vector. The similarity can be calculated by methods such as dot product and cosine similarity. Then, normalize these similarity scores to obtain the attention weights. Next, perform a weighted sum on the normalized device feature vector according to the attention weights to obtain the interaction features corresponding to each feature component in the normalized environmental feature vector. Finally, combine all the interaction features to form an interaction feature vector.

[0071] Step S1224: Invoke the risk assessment layer in the security assessment model to perform non-linear mapping processing on the interaction feature vector to generate the comprehensive risk assessment index; wherein, the feature interaction layer fuses the standardized environment feature and the standardized device feature through a cross-attention mechanism; the risk assessment layer performs weighted calculation on the interaction feature vector through a multi-layer perceptron network and outputs the comprehensive risk assessment index; the comprehensive risk assessment index includes the weighted sum of the environmental risk contribution degree and the device risk contribution degree.

[0072] In this embodiment, the risk assessment layer in the security assessment model uses a multi-layer perceptron network to perform non-linear mapping processing on the interaction feature vector to generate a comprehensive risk assessment index.

[0073] The multi-layer perceptron network consists of an input layer, a hidden layer, and an output layer. The interaction feature vector serves as the input to the input layer, undergoes non-linear transformation and weighted calculation in the hidden layer, and finally outputs the comprehensive risk assessment index in the output layer. The neurons in the hidden layer perform non-linear transformation on the input through an activation function, such as the ReLU function, to enhance the expression ability of the model. During the training process of the multi-layer perceptron network, the connection weights between neurons are continuously adjusted through the backpropagation algorithm, enabling the model to learn the mapping relationship between the interaction feature vector and the comprehensive risk assessment index.

[0074] The comprehensive risk assessment index includes the weighted sum of the environmental risk contribution degree and the device risk contribution degree. The environmental risk contribution degree is calculated based on the output result of the standardized environment feature in the multi-layer perceptron network, and the device risk contribution degree is calculated based on the output result of the standardized device feature in the multi-layer perceptron network. First, the multi-layer perceptron network processes the corresponding partial interaction features of the standardized environment feature and the standardized device feature respectively to obtain the intermediate outputs related to the environment and the device. Then, these intermediate outputs are weighted and summed to obtain the environmental risk contribution degree and the device risk contribution degree. Specifically, for the calculation of the environmental risk contribution degree, a weight is assigned to each component of the intermediate output related to the environment in the multi-layer perceptron network, and these weights are learned during the model training process. After multiplying each component by the corresponding weight and summing them up, the environmental risk contribution degree is obtained. Similarly, for the device risk contribution degree, weights are assigned to each component of the intermediate output related to the device in the multi-layer perceptron network, and then weighted summation is performed to obtain the device risk contribution degree. Finally, the environmental risk contribution degree and the device risk contribution degree are multiplied by their respective total weights, and the two results are added together to obtain the comprehensive risk assessment index. The total weights here are also adjusted and determined according to the influence degree of environmental factors and device factors on the overall security risk during the model training process.

[0075] Step S123: Perform dynamic comparison processing based on the comprehensive risk assessment index and a preset safety threshold to determine the safety status assessment result.

[0076] In this embodiment, in order to accurately evaluate the safety status of the VR flight device, it is necessary to perform dynamic comparison processing on the comprehensive risk assessment index and a preset safety threshold. Since the flight environment and the device operating status are constantly changing, the preset safety threshold also needs to be corrected in real time according to the actual situation.

[0077] Step S1231: Obtain a dynamic safety threshold adjustment coefficient for the current environmental scenario.

[0078] In this embodiment, obtaining the dynamic safety threshold adjustment coefficient requires comprehensive consideration of environmental complexity and device stability. Calculate the environmental complexity parameter based on the obstacle density index and the dynamic change rate index in the environmental perception data. First, for the obstacle density index, it is obtained by statistically calculating the ratio of the number of obstacles in each scene area to the area of the scene area. The obstacle density indexes of each scene area are summarized and analyzed to obtain a comprehensive obstacle density index. For the dynamic change rate index, it is determined by calculating the change amplitude of the obstacle distribution within adjacent time windows. Similarly, the dynamic change rate indexes of each scene area are summarized and analyzed to obtain a comprehensive dynamic change rate index. Then, assign weights to the obstacle density index and the dynamic change rate index respectively. These two weights are determined according to their influence on environmental complexity. Multiply the obstacle density index by the corresponding weight, multiply the dynamic change rate index by the corresponding weight, and then add these two results to obtain the environmental complexity parameter.

[0079] Calculate the device stability parameter based on the load fluctuation index and the energy consumption trend index in the device status data. The load fluctuation index is obtained by weighted summation of the standard deviation of acceleration and the standard deviation of the attitude angle within the device operation segment. The energy consumption trend index is obtained by analyzing the change in the average energy consumption rate between adjacent device operation segments. Similarly, assign weights to the load fluctuation index and the energy consumption trend index respectively. Multiply the load fluctuation index by the corresponding weight, multiply the energy consumption trend index by the corresponding weight, and then add these two results to obtain the device stability parameter.

[0080] Finally, perform weighted summation on the environmental complexity parameter and the device stability parameter to obtain the dynamic safety threshold adjustment coefficient. The weights assigned to the environmental complexity parameter and the device stability parameter here are determined according to the influence degrees of environmental factors and device factors on the safety threshold adjustment.

[0081] Step S1232: Perform real-time correction processing on the preset safety threshold according to the dynamic safety threshold adjustment coefficient to generate a dynamic safety threshold.

[0082] In this embodiment, after obtaining the dynamic safety threshold adjustment coefficient, it is necessary to correct the preset safety threshold in real time. The preset safety threshold is a fixed value determined during model training or system initialization, which represents the safety risk limit of the VR flight device under normal circumstances. Multiplying the preset safety threshold by the dynamic safety threshold adjustment coefficient yields the dynamic safety threshold. In this way, the dynamic safety threshold can be adjusted in real time according to the current environmental complexity and device stability, more accurately reflecting the safety requirements in the current flight state.

[0083] Step S1233: Compare the comprehensive risk assessment index with the dynamic safety threshold.

[0084] In this embodiment, the comprehensive risk assessment index calculated previously is directly compared with the dynamic safety threshold. This comparison is based on the comparison of numerical values, aiming to determine whether the safety risk of the current VR flight device exceeds the range allowed under the current environment and device state.

[0085] Step S1234: When the comprehensive risk assessment index is greater than the dynamic safety threshold, generate a first risk assessment result.

[0086] In this embodiment, if the comprehensive risk assessment index is greater than the dynamic safety threshold, it indicates that the current VR flight device faces a relatively high safety risk and a dangerous situation may occur. At this time, generate a first risk assessment result, which represents a high-risk state.

[0087] Step S1235: When the comprehensive risk assessment index is less than or equal to the dynamic safety threshold, generate a second risk assessment result.

[0088] In this embodiment, if the comprehensive risk assessment index is less than or equal to the dynamic safety threshold, it indicates that the safety risk of the current VR flight device is within an acceptable range and the flight state is relatively safe. At this time, generate a second risk assessment result, which represents a low-risk state.

[0089] Step S1236: Determine the safety state assessment result according to the first risk assessment result or the second risk assessment result.

[0090] In this embodiment, the first risk assessment result and the second risk assessment result are used as the judgment basis to finally determine the safety state assessment result of the VR flight device. If the first risk assessment result is obtained, the safety state assessment result is a high-risk state; if the second risk assessment result is obtained, the safety state assessment result is a low-risk state.

[0091] Step S130: Perform dynamic security policy matching processing based on the security status evaluation result to obtain a security regulation instruction.

[0092] In this embodiment, according to the previously determined security status evaluation result, it is necessary to perform dynamic security policy matching processing to obtain the corresponding security regulation instruction, so as to adjust the operation of the VR flight device and ensure flight safety.

[0093] Step S131: When the security status evaluation result is the first risk assessment result, call the pre-stored emergency regulation policy library to match the first security regulation instruction corresponding to the current environmental risk level and equipment operation risk level.

[0094] In this embodiment, if the security status evaluation result is the first risk assessment result, that is, the high-risk state, it is necessary to call the pre-stored emergency regulation policy library. The emergency regulation policy library contains various regulation policies for high-risk situations, such as the equipment emergency deceleration policy, the environmental obstacle avoidance path adjustment policy, and the energy supply strengthening policy, etc.

[0095] First, determine the current environmental risk level and equipment operation risk level. The environmental risk level can be determined according to the environmental complexity parameter. The higher the environmental complexity parameter, the higher the environmental risk level. The equipment operation risk level can be determined according to the equipment stability parameter. The lower the equipment stability parameter, the higher the equipment operation risk level. Then, search for the policy combination that matches the current environmental risk level and equipment operation risk level in the emergency regulation policy library. For example, if the environmental risk level is high and the equipment operation risk level is also high, it may be possible to select the equipment emergency deceleration policy, the environmental obstacle avoidance path adjustment policy, and the energy supply strengthening policy to be executed simultaneously. Combining these matching policies gives the first security regulation instruction. The first security regulation instruction contains the specific parameters required to execute these policies, such as the priority parameter of the equipment emergency deceleration policy, the path offset parameter of the environmental obstacle avoidance path adjustment policy, and the energy allocation ratio parameter of the energy supply strengthening policy, etc.

[0096] Step S132: When the security status evaluation result is the second risk assessment result, call the pre-stored regular regulation policy library to match the second security regulation instruction corresponding to the current environmental risk level and equipment operation risk level.

[0097] In this embodiment, if the security status evaluation result is the second risk assessment result, that is, the low-risk state, it is necessary to call the pre-stored regular regulation policy library. The regular regulation policy library contains various regulation policies for low-risk situations, such as the equipment operation parameter optimization policy, the environmental monitoring frequency adjustment policy, and the energy consumption balanced distribution policy, etc.

[0098] Similarly, first determine the current environmental risk level and the equipment operation risk level. Then, search for a policy combination in the conventional regulation strategy library that matches the current environmental risk level and the equipment operation risk level. For example, if the environmental risk level is low and the equipment operation risk level is also low, strategies such as equipment operation parameter optimization strategy and energy consumption balanced distribution strategy may be selected for execution. Combine these matching strategies to obtain the second safety regulation instruction. The second safety regulation instruction also includes the specific parameters required to execute these strategies, such as the parameter adjustment value of the equipment operation parameter optimization strategy, the monitoring time interval of the environmental monitoring frequency adjustment strategy, etc.

[0099] Step S133: Generate the safety regulation instruction according to the first safety regulation instruction or the second safety regulation instruction.

[0100] In this embodiment, according to the first safety regulation instruction or the second safety regulation instruction obtained previously, the safety regulation instruction is finally generated. If the safety status evaluation result is the first risk assessment result, then the safety regulation instruction is the first safety regulation instruction; if the safety status evaluation result is the second risk assessment result, then the safety regulation instruction is the second safety regulation instruction. The safety regulation instruction contains a series of parameters, and these parameters will be used for subsequent adjustment of the operation parameters of the VR flight device.

[0101] Step S140: Adjust the operation parameters of the VR flight device according to the safety regulation instruction and trigger the warning feedback operation.

[0102] In this embodiment, after obtaining the safety regulation instruction, it is necessary to adjust the operation parameters of the VR flight device according to these instructions and trigger the corresponding warning feedback operation to ensure flight safety.

[0103] Step S141: Analyze the priority parameters in the safety regulation instruction to determine the execution order of the equipment control instructions.

[0104] In this embodiment, the safety regulation instruction contains priority parameters, and these parameters are used to determine the execution order of the equipment control instructions. The priority parameters are assigned according to the importance and urgency of different strategies during the policy matching process. First, analyze the priority parameters in the safety regulation instruction and sort the various equipment control instructions from high to low according to the priority. For example, if the safety regulation instruction includes an equipment emergency deceleration strategy, an environmental obstacle avoidance path adjustment strategy, and an energy supply enhancement strategy, and the equipment emergency deceleration strategy has the highest priority, then when executing the equipment control instructions, the equipment emergency deceleration strategy will be executed first, and then other strategies will be executed.

[0105] Step S142: Adjust the navigation path planning of the VR flight device according to the path offset parameter to generate an updated navigation path.

[0106] In this embodiment, the safety control instruction includes the path offset parameter of the environmental obstacle avoidance path adjustment strategy. The navigation path planning of the VR flight device is adjusted according to these parameters. The navigation path planning was originally determined based on the initial environmental information and flight target. When an obstacle is detected in the environment or the environment changes, the navigation path needs to be corrected according to the path offset parameter.

[0107] The path offset parameter includes information such as the direction and distance of obstacle avoidance. First, the avoidance direction is determined according to the position of the obstacle and the path offset parameter, and then the navigation path is offset by the specified distance. For example, if an obstacle is detected ahead and the path offset parameter indicates an offset to the right by a certain distance, then the navigation path is offset to the right accordingly. During the offset process, the flight performance of the flight device and other limiting factors of the environment also need to be considered to ensure that the offset path is feasible. By adjusting the navigation path, an updated navigation path is generated, and the flight device will fly according to the updated navigation path.

[0108] Step S143: Reallocate the energy supply mode of the device according to the energy distribution ratio parameter.

[0109] In this embodiment, the safety control instruction includes the energy distribution ratio parameter of the energy supply enhancement strategy. The energy supply mode of the device is reallocated according to these parameters. The energy supply mode includes the switching ratio control of the main energy module and the standby energy module and the smooth transition processing of the energy output power.

[0110] First, the switching ratio of the main energy module and the standby energy module is determined according to the energy distribution ratio parameter. If the energy distribution ratio parameter indicates an increase in the usage ratio of the standby energy module, then the energy supply system is adjusted accordingly to output more energy from the standby energy module. During the switching process, smooth transition processing of the energy output power is required to avoid the impact of sudden changes in energy output on the flight device. For example, the energy output power can be gradually increased or decreased to make the change in energy supply more stable. By reallocating the energy supply mode, it is ensured that the flight device can obtain sufficient energy support in different operating states.

[0111] Step S144: Synchronize the execution order, the updated navigation path, and the reallocated energy supply mode to the control system of the VR flight device to complete the adjustment of the operating parameters.

[0112] In this embodiment, the execution order of the device control instructions determined previously, the updated navigation path, and the reallocated energy supply mode are synchronized to the control system of the VR flight device. The control system will control and adjust each component of the flight device according to this information, thereby completing the adjustment of the operating parameters.

[0113] During the synchronization process, it is necessary to ensure the accurate transmission and timely processing of information. The execution order, the updated navigation path, and the reallocated energy supply mode can be sent to the control system through the data communication interface. After receiving this information, the control system will execute the corresponding control instructions in sequence according to the execution order, control the flight direction and speed of the flight device according to the updated navigation path, and adjust the energy output according to the reallocated energy supply mode. Through such a synchronization process, the flight device can operate according to the requirements of the safety control instructions.

[0114] Step S145: Generate a warning level identifier according to the safety status assessment result.

[0115] In this embodiment, a warning level identifier is generated according to the safety status assessment result determined previously. If the safety status assessment result is the first risk assessment result, that is, a high-risk state, a high-risk warning level identifier is generated; if the safety status assessment result is the second risk assessment result, that is, a low-risk state, a low-risk warning level identifier is generated. The warning level identifier is used to distinguish different degrees of safety risks, so as to perform corresponding warning feedback operations subsequently.

[0116] Step S146: When the warning level identifier is a high-risk warning, start multi-level warning notification processing, and the multi-level warning notification processing includes sending an emergency warning signal to the user terminal, displaying dynamic obstacle avoidance prompt information on the VR interface, and activating the automatic braking mechanism of the device.

[0117] In this embodiment, if the warning level identifier is a high-risk warning, it means that the VR flight device faces relatively high safety risks and multi-level warning notification processing needs to be started.

[0118] Send an emergency warning signal to the user terminal. The emergency warning signal includes a combination of sound warning, vibration feedback, and visual flashing prompts. A sharp sound is emitted through the speaker of the user terminal to remind the user to pay attention to flight safety; vibration is generated through the vibration module of the user terminal to give the user a tactile warning; at the same time, a flashing warning icon or text is displayed on the screen of the user terminal to give the user a visual warning.

[0119] Display dynamic obstacle avoidance prompt information on the VR interface. The dynamic obstacle avoidance prompt information marks the positions of obstacles and recommends avoidance directions on the VR interface through augmented reality technology. Utilize the obstacle information in the environmental perception data to mark the positions of the obstacles on the VR interface. At the same time, according to the updated navigation path and path offset parameters, display the recommended avoidance directions on the VR interface to help users better perform obstacle avoidance operations.

[0120] Activate the automatic braking mechanism of the device. The automatic braking mechanism calculates the braking distance parameter based on the current operating speed of the device, the distance to environmental obstacles, and the preset braking response time, and gradually reduces the operating speed of the device to within the safety threshold through segmented deceleration control. First, calculate the traveling distance of the flying device during the braking response time based on the current operating speed of the device and the preset braking response time. Then, combine the distance to environmental obstacles to determine the distance and amplitude of deceleration required. Through segmented deceleration control, divide the deceleration process into multiple stages, and dynamically adjust the braking gradient according to the obstacle distance at each stage to ensure that the speed adjustment is completed within the braking distance parameter range. For example, when the distance to the obstacle is far, use a smaller braking gradient for deceleration; when the distance to the obstacle is close, use a larger braking gradient for deceleration until the operating speed of the device is reduced to within the safety threshold.

[0121] Step S147: When the early warning level identifier is a low-risk early warning, initiate regular early warning notification processing, including sending a status reminder message to the user terminal, displaying parameter optimization suggestion information on the VR interface, and recording the device operation log.

[0122] In this embodiment, if the early warning level identifier is a low-risk early warning, it indicates that the safety risk of the VR flying device is within an acceptable range, but some regular early warning notification processing still needs to be carried out.

[0123] Send a status reminder message to the user terminal. The status reminder message informs the user of the current operating status of the flying device, such as the flying speed, remaining energy, etc. in the form of text or voice, so that the user can clearly understand the status of the flying device.

[0124] Display parameter optimization suggestion information on the VR interface. Analyze whether there is room for optimizing the operating parameters of the current flying device based on the device operation characteristics and safety assessment results. For example, if the device load fluctuation index is high, it may be recommended to adjust the flying speed or attitude; if the device energy consumption trend index shows an increase in energy consumption, it may be recommended to optimize the energy distribution mode. Display these parameter optimization suggestion information on the VR interface to help users adjust the operating parameters of the flying device.

[0125] Record the device operation log. The device operation log records various status information of the flying device during operation, such as acceleration, attitude, energy consumption, etc. By recording the device operation log, the operation situation of the flying device can be analyzed and evaluated subsequently to discover potential problems and make improvements.

[0126] Step S150: After completing the warning feedback operation, continuously monitor the subsequent operation data set of the VR flying device.

[0127] In this embodiment, after completing the warning feedback operation, it is necessary to continuously monitor the subsequent operation data set of the VR flying device. This is to timely understand the state change of the flying device after adjusting the operation parameters and judge whether the safety state has been improved.

[0128] In the same way as obtaining the real-time operation data set, use various sensors and monitoring systems to continuously collect environmental perception data and device status data. The lidar sensor continues to scan the surrounding environment to obtain the position and distribution changes of obstacles; the camera sensor continues to capture images of the flight scene for analyzing the dynamic changes of the environment; the meteorological sensor continuously measures the meteorological parameters of the environment; the accelerometer, gyroscope and energy management system respectively monitor the acceleration, attitude and energy consumption of the flying device. Integrate these collected data together to form the subsequent operation data set.

[0129] Step S151: Perform safety state verification processing on the subsequent operation data set to generate a verification result.

[0130] In this embodiment, perform safety state verification processing on the subsequent operation data set to determine whether the safety state of the flying device has returned to normal.

[0131] Step S1511: Compare the adjusted operation parameters with the safety reference parameters updated according to the dynamic safety threshold adjustment coefficient.

[0132] In this embodiment, first update the safety reference parameters according to the dynamic safety threshold adjustment coefficient. The safety reference parameters are the reference ranges of the operation parameters of the flying device in the normal safety state. Adjust these reference ranges according to the dynamic safety threshold adjustment coefficient to make them more in line with the current environment and device state. Then, compare the adjusted operation parameters with the updated safety reference parameters. For example, compare the current flight speed of the flying device with the updated safe speed range, and compare the energy consumption rate with the updated safe energy consumption rate range. If the adjusted operation parameters are within the updated safety reference parameter range, it means that the operation state of this parameter is normal; if it exceeds the range, it means that the operation state of this parameter has risks.

[0133] Step S1512: Analyze whether the change trend of the obstacle distribution in the environmental perception data meets the preset dynamic obstacle avoidance requirements.

[0134] In this embodiment, analyze the change trend of the obstacle distribution in the environmental perception data. By comparing the information such as the position, quantity, and size of the obstacles within adjacent time windows, determine whether the distribution of the obstacles has changed and the trend of the change. The preset dynamic obstacle avoidance requirements are set according to the flight performance and safety requirements of the flying device. For example, it is required that a certain safety distance be maintained between the flying device and the obstacles, and the flying device can avoid the obstacles in a timely manner when encountering them. If the change trend of the obstacle distribution indicates that the flying device can meet these dynamic obstacle avoidance requirements, such as the obstacles gradually moving away from the flying device or the flying device can avoid the obstacles according to the planned path, it means that the safety state of the environment is good; on the contrary, if the change trend of the obstacle distribution shows that the flying device may collide with the obstacles or cannot avoid them in a timely manner, it means that there are safety risks in the environment.

[0135] Step S1513: Verify whether the load fluctuation stability index in the device status data has recovered to within the safe threshold range.

[0136] In this embodiment, verify the load fluctuation stability index in the device status data. The load fluctuation stability index is obtained by weighted summation of the standard deviation of the acceleration and the standard deviation of the attitude angle within the device operation segment. After the warning feedback operation, calculate this index again and compare it with the safe threshold range. The safe threshold range is determined according to the normal operation state and safety requirements of the device. If the load fluctuation stability index recovers to within the safe threshold range, it means that the operation state of the device tends to be stable and the load fluctuation is within an acceptable range; if this index exceeds the safe threshold range, it means that there is still a large load fluctuation in the device, which may affect the normal operation and safety of the flying device.

[0137] Through the comparison of the adjusted operation parameters and the updated safety reference parameters, the analysis of the change trend of the obstacle distribution in the environmental perception data, and the verification of the load fluctuation stability index in the device status data, generate the verification result comprehensively. The verification result is divided into two situations: the safety state returns to normal and the safety state does not return to normal.

[0138] Step S152: When the verification result indicates that the safety state has not returned to normal, iteratively execute the dynamic safety policy matching process to update the safety control instruction.

[0139] In this embodiment, if the verification result indicates that the safety state has not returned to normal, it means that the current safety control measures taken have not achieved the expected effect, and it is necessary to iteratively execute the dynamic safety policy matching process.

[0140] Re-determine the current environmental risk level and equipment operation risk level. According to the latest subsequent operation data set, determine them according to the method of determining the environmental risk level and equipment operation risk level before, that is, according to the environmental complexity parameter and equipment stability parameter. The environmental complexity parameter is calculated from the obstacle density index and dynamic change rate index in the environmental perception data, and the equipment stability parameter is calculated from the load fluctuation index and energy consumption trend index in the equipment status data.

[0141] According to the re-determined environmental risk level and equipment operation risk level, call the pre-stored emergency control strategy library or conventional control strategy library again. If the safety risk is still relatively high, call the emergency control strategy library; if the safety risk has decreased but has not returned to normal, call the conventional control strategy library. Match the strategy combination corresponding to the current environmental risk level and equipment operation risk level in the corresponding strategy library to generate a new safety control instruction. The new safety control instruction may adjust the previous strategy, such as increasing the amplitude of the emergency speed reduction of the equipment, further optimizing the environmental obstacle avoidance path, or adjusting the energy distribution ratio, etc.

[0142] Step S153: When the verification result indicates that the safety state has returned to normal, terminate the warning feedback operation and restore the default operating parameters.

[0143] In this embodiment, if the verification result indicates that the safety state has returned to normal, it means that through the previous safety control measures and warning feedback operations, the operating state of the VR flight device has returned to the safe range. At this time, terminate the warning feedback operation, stop sending emergency warning signals or status reminder messages to the user terminal, no longer display dynamic obstacle avoidance prompt messages or parameter optimization suggestion messages on the VR interface, and at the same time turn off the automatic braking mechanism of the device.

[0144] Restore the default operating parameters. The default operating parameters are the parameter settings used when the flight device is operating normally and there is no safety risk, including flight speed, attitude control parameters, energy distribution mode, etc. Adjust the operating parameters of the flight device back to the default values so that the flight device can continue to operate in the normal mode. Through such operations, ensure that the flight device can work efficiently and stably in a safe state.

[0145] In summary, the safety monitoring method for the VR flight device forms a complete safety monitoring closed-loop by obtaining the real-time operation data set, performing safety analysis and processing to generate a safety state evaluation result, performing dynamic safety policy matching processing based on the evaluation result to obtain a safety control instruction, adjusting the operating parameters according to the instruction and triggering a warning feedback operation, and finally verifying and iteratively processing the safety state, which can effectively guarantee the flight safety of the VR flight device.

[0146] Furthermore, Figure 2The following shows a schematic diagram of the hardware structure of a VR service system 100 for implementing the method provided in the embodiments of the present application. As Figure 2 shown, the VR service system 100 may include at least one processor 102 (the processor 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, a transmission device 106 for communication functions, and a controller 108. Those of ordinary skill in the art can understand that Figure 2 the structure shown is only illustrative and does not limit the structure of the VR service system 100. For example, the VR service system 100 may further include more or fewer components than Figure 2 shown in, or have a different configuration from Figure 2 shown.

[0147] The memory 104 can be used to store software programs and modules of application software, such as the program instructions corresponding to the above method embodiments in the embodiments of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the above-mentioned safety monitoring method for a VR flight device. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely disposed relative to the processor 102, and these remote memories may be connected to the VR service system 100 through a network. Examples of the above networks include, but are not limited to, the Internet, enterprise intranets, local area networks, mobile communication networks, and combinations thereof.

[0148] The transmission device 106 is used to obtain or send data via a network. Specific examples of the above network may include a wireless network provided by the communication provider of the VR service system 100. In one instance, the transmission device 106 includes a network adapter, which can be connected to other network devices through a base station and thus can communicate with the Internet. In one instance, the transmission device 106 may be a radio frequency module, which is used to communicate with the Internet wirelessly.

[0149] It should be noted that: the above sequence of the embodiments of the present application is only for description and does not represent the advantages or disadvantages of the embodiments. And the above specific embodiments of the present application have been described. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be executed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain implementations, multitasking and parallel processing are also possible or may be advantageous.

[0150] The embodiments in the embodiments of the present application are all described in a progressive manner. For the parts that are consistent and similar among the embodiments, reference can be made to each other. The key points of each embodiment are the differences from other embodiments. In particular, for the above different embodiments, since they are basically similar to the method embodiments, the description is relatively simple. For the relevant parts, reference can be made to the partial description of the method embodiments.

[0151] Those of ordinary skill in the art can understand that all or part of the steps to implement the above embodiments can be completed by hardware, or can be completed by instructing relevant hardware through a program. The above program can be stored in a computer-readable storage medium, and the above-mentioned storage medium can be a read-only memory, a magnetic disk or an optical disc, etc.

Claims

1. A safety monitoring method for a VR flight device, characterized in that, The method includes: Obtaining a real-time operation data set of the VR flight device, where the real-time operation data set includes environmental perception data and device status data; Performing security analysis and processing on the real-time operation data set to generate a security status evaluation result; Performing dynamic security policy matching processing based on the security status evaluation result to obtain a security control instruction; Adjusting the operation parameters of the VR flight device according to the security control instruction and triggering a warning feedback operation.

2. The safety monitoring method for a VR flight device according to claim 1, wherein, The performing security analysis and processing on the real-time operation data set to generate a security status evaluation result includes: Extracting environmental dynamic features from the environmental perception data and extracting device operation features from the device status data; Inputting the environmental dynamic features and the device operation features into a pre-trained security evaluation model to generate a comprehensive risk evaluation index; Performing dynamic comparison processing on the comprehensive risk evaluation index and a preset security threshold to determine the security status evaluation result.

3. The safety monitoring method for a VR flight device according to claim 2, characterized in that, The extracting environmental dynamic features from the environmental perception data includes: Performing dynamic scene segmentation processing on the environmental perception data to generate multiple scene regions; Performing obstacle recognition processing on each scene region to determine the obstacle distribution features; Performing feature calculation based on the obstacle distribution features to generate the environmental dynamic features; wherein, the environmental dynamic features include obstacle density features, dynamic change rate features, and spatial distribution uniformity features; the obstacle density features are determined by statistically calculating the ratio of the number of obstacles in each scene region to the area of the scene region; the dynamic change rate features are determined by calculating the change amplitude of the obstacle distribution within adjacent time windows; the spatial distribution uniformity features are determined by calculating the position dispersion of the obstacles in each scene region, and the dispersion is analyzed using the statistical variance method.

4. The safety monitoring method for a VR flight device according to claim 2, wherein, The extracting device operation features from the device status data includes: Performing time series segmentation processing on the device status data to generate multiple device operation segments; Performing operation mode recognition processing on each device operation segment to determine the device operation mode features; Calculating a device load fluctuation index and a device energy consumption trend index according to the device operation mode features; Performing aggregation processing on the device load fluctuation index and the device energy consumption trend index to generate the device operation features; Wherein, the device operation mode features include acceleration change features, attitude adjustment frequency features, and energy consumption rate features; The acceleration change features are determined by analyzing the slope of the acceleration curve of the device operation segment; The attitude adjustment frequency features are determined by statistically calculating the number of executions of the attitude adjustment instructions in the device operation segment; The energy consumption rate features are determined by calculating the change rate of the energy consumption per unit time.

5. The safety monitoring method for a VR flight device according to claim 1, characterized in that, The inputting the environmental dynamic features and the device operation features into a pre-trained security evaluation model to generate a comprehensive risk evaluation index includes: Performing first standardization processing on the environmental dynamic features to obtain standardized environmental features; Performing second standardization processing on the device operation features to obtain standardized device features; Input the standardized environmental features and the standardized device features into the feature interaction layer of the pre-trained security assessment model to generate an interaction feature vector; Call the risk assessment layer in the security assessment model to perform a non-linear mapping process on the interaction feature vector to generate the comprehensive risk assessment index; Among them, the feature interaction layer fuses the standardized environmental features and the standardized device features through a cross-attention mechanism; the risk assessment layer performs weighted calculation on the interaction feature vector through a multi-layer perceptron network and outputs the comprehensive risk assessment index; the comprehensive risk assessment index includes the weighted sum of the environmental risk contribution degree and the device risk contribution degree.

6. The safety monitoring method for a VR flight device according to claim 5, wherein, The dynamic comparison process according to the comprehensive risk assessment index and the preset security threshold to determine the security status assessment result includes: Obtain the dynamic security threshold adjustment coefficient under the current environmental scenario; Perform real-time correction processing on the preset security threshold according to the dynamic security threshold adjustment coefficient to generate a dynamic security threshold; Compare the comprehensive risk assessment index with the dynamic security threshold; When the comprehensive risk assessment index is greater than the dynamic security threshold, generate a first risk assessment result; When the comprehensive risk assessment index is less than or equal to the dynamic security threshold, generate a second risk assessment result; Determine the security status assessment result according to the first risk assessment result or the second risk assessment result; Among them, the dynamic security threshold adjustment coefficient is determined by the following method: Calculate the environmental complexity parameter according to the obstacle density index and the dynamic change rate index in the environmental perception data; Calculate the device stability parameter according to the load fluctuation index and the energy consumption trend index in the device status data; Use the weighted sum of the environmental complexity parameter and the device stability parameter as the dynamic security threshold adjustment coefficient.

7. The safety monitoring method for a VR flight device according to claim 6, characterized in that, The dynamic security policy matching process based on the security status assessment result to obtain a security control instruction includes: When the security status assessment result is the first risk assessment result, call the pre-stored emergency control strategy library to match the first security control instruction corresponding to the current environmental risk level and the device operation risk level; When the security status assessment result is the second risk assessment result, call the pre-stored conventional control strategy library to match the second security control instruction corresponding to the current environmental risk level and the device operation risk level; Generate the security control instruction according to the first security control instruction or the second security control instruction; Among them, the emergency control strategy library includes the device emergency speed reduction strategy, the environmental obstacle avoidance path adjustment strategy, and the energy supply strengthening strategy; The conventional control strategy library includes the device operation parameter optimization strategy, the environmental monitoring frequency adjustment strategy, and the energy consumption balanced distribution strategy; The security control instruction includes the priority parameter for executing the device emergency speed reduction strategy, the path offset parameter for the environmental obstacle avoidance path adjustment strategy, and the energy distribution ratio parameter for the energy supply strengthening strategy.

8. The safety monitoring method for a VR flight device according to claim 7, characterized in that, Adjust the operation parameters of the VR flight device according to the security control instruction, including: Analyze the priority parameters in the safety control instruction to determine the execution order of the device control instruction; Adjust the navigation path planning of the VR flight device according to the path offset parameter to generate an updated navigation path; Reallocate the energy supply mode of the device according to the energy distribution ratio parameter; Synchronize the execution order, the updated navigation path, and the reallocated energy supply mode to the control system of the VR flight device to complete the adjustment of the operating parameters; Among them, the adjustment of the navigation path planning includes the dynamic correction of the obstacle avoidance path and the gradient change control of the flight speed; The adjustment of the energy supply mode includes the switching ratio control of the main energy module and the standby energy module and the smooth transition processing of the energy output power; Among them, the triggering of the warning feedback operation includes: Generate a warning level identifier according to the safety status evaluation result; When the warning level identifier is a high-risk warning, start the multi-level warning notification process, and the multi-level warning notification process includes sending an emergency warning signal to the user terminal, displaying dynamic obstacle avoidance prompt information on the VR interface, and activating the automatic braking mechanism of the device; When the warning level identifier is a low-risk warning, start the regular warning notification process, including sending a status reminder message to the user terminal, displaying parameter optimization suggestion information on the VR interface, and recording the device operation log; Among them, the emergency warning signal includes a combination of sound warning, vibration feedback, and visual flashing prompt; The dynamic obstacle avoidance prompt information marks the obstacle position and recommends the avoidance direction on the VR interface through augmented reality technology; The automatic braking mechanism calculates the braking distance parameter according to the current operating speed of the device, the distance of environmental obstacles, and the preset braking response time, and gradually reduces the operating speed of the device to within the safety threshold through segmented deceleration control; The segmented deceleration control includes dynamically adjusting the braking gradient according to the obstacle distance to ensure that the speed adjustment is completed within the braking distance parameter range.

9. The safety monitoring method for a VR flight device according to claim 8, characterized in that, The method further includes: After completing the warning feedback operation, continuously monitor the subsequent operation data set of the VR flight device; Perform a safety status verification process on the subsequent operation data set to generate a verification result; When the verification result indicates that the safety status has not returned to normal, iteratively execute the dynamic safety policy matching process to update the safety control instruction; When the verification result indicates that the safety status has returned to normal, terminate the warning feedback operation and restore the default operating parameters; Among them, the safety status verification process includes: Compare the adjusted operating parameters with the safety reference parameters updated according to the dynamic safety threshold adjustment coefficient; Analyze whether the change trend of the obstacle distribution in the environmental perception data meets the preset dynamic obstacle avoidance requirements; Verify whether the load fluctuation stability index in the device status data has returned to within the safety threshold range.

10. A VR service system, characterized in that, The VR service system includes a processor and a readable storage medium, and the readable storage medium stores a program, which when executed by the processor implements the safety monitoring method for the VR flight device according to any one of claims 1-9.

Images