Power operation safety protection positioning method, system and device integrated with beidou navigation

Abstract

The application discloses a power operation safety protection positioning method, system and equipment fused with Beidou navigation, relates to the field of radio wave positioning, and comprises the following steps: identifying a weak positioning area in a power operation area; acquiring a current power operation task; if the path of the power operation task includes the weak positioning area, activating a sliding mode switching positioning module; the sliding mode switching positioning module is used for behavior pattern recognition of a target performing the power operation task, and outputs behavior pattern features; a multi-mode confidence relationship matrix is established according to the behavior pattern features, a mode switching strategy is outputted according to the multi-mode confidence relationship matrix; and a multi-mode fusion device analyzes multi-mode positioning data according to the mode switching strategy and outputs a protection positioning result. The application solves the technical problems of unstable positioning signals and decreased precision in complex environment switching existing in the power operation safety protection positioning, and achieves the technical effects of improving the accuracy and reliability of the power operation safety protection positioning.

Description

Technical Field

[0001] The present application relates to fields related to radio wave positioning, and in particular to methods, systems and equipment for positioning safety protection of power operations integrated with Beidou navigation. Background Art

[0002] Positioning for power operation safety and protection is crucial for ensuring the safety of workers and the stable operation of power facilities, especially in complex operating environments. Accurate positioning is crucial for ensuring safe operations. Currently, the primary approach to addressing this problem relies on a single positioning technology, such as GPS or indoor positioning systems. However, due to the limitations of different positioning technologies, a single technology struggles to adapt to complex switching environments, resulting in unstable positioning signals and decreased positioning accuracy, thus compromising the reliability of power operation safety and protection.

[0003] In the current related technologies, the safety protection positioning of power operations has technical problems such as unstable positioning signals and decreased accuracy when switching in complex environments. Summary of the Invention

[0004] The present application provides a method, system and equipment for electric power operation safety protection positioning that integrates Beidou navigation. It first identifies the weak positioning area for indoor and outdoor switching in the electric power operation area. After obtaining the operation task, if the path contains this area, the sliding mode switching positioning module is activated. The module identifies the behavior pattern of the operation target and outputs the characteristics. Based on this, a multi-modal confidence relationship matrix is ​​established to output the modal switching strategy. Finally, the multi-modal fusion device performs multi-modal positioning data analysis according to this strategy and outputs the protection positioning results. These technical means solve the technical problems of unstable positioning signals and decreased accuracy when switching in complex environments in the existing electric power operation safety protection positioning, and achieve the technical effect of improving the accuracy and reliability of electric power operation safety protection positioning.

[0005] The present application provides a method for safety protection positioning of electric power operations integrated with Beidou navigation, comprising: identifying a weak positioning area in an electric power operation area, wherein the weak positioning area is a transition area when switching from indoors to outdoors, or from outdoors to indoors; obtaining a current electric power operation task, and if the path of the electric power operation task includes the weak positioning area, activating a sliding mode switching positioning module, wherein the sliding mode switching positioning module is connected to a multimodal fusion device, and the multimodal fusion device has Beidou navigation integrated therein; the sliding mode switching positioning module is used to identify the behavior pattern of a target performing the electric power operation task and output the behavior pattern characteristics; establishing a multimodal confidence relationship matrix according to the behavior pattern characteristics, and outputting a modal switching strategy using the multimodal confidence relationship matrix; the multimodal fusion device performs multimodal positioning data analysis according to the modal switching strategy and outputs a protection positioning result.

[0006] In a possible implementation, the sliding mode switching positioning module is set in an edge device, and the behavior pattern of the target performing the power operation task is identified through the edge device, and the following processing is performed: obtaining the target of the power operation task and collecting real-time behavior data of the target; identifying the spatial change rate, action mutation point and movement direction change of the real-time behavior data, and outputting behavior pattern category characteristics; collecting the number of targets and target dynamic rate performing the power operation task, and outputting behavior pattern target characteristics; outputting behavior pattern characteristics according to the behavior pattern category characteristics and the behavior pattern target characteristics.

[0007] In a possible implementation, a multimodal confidence relationship matrix is ​​established according to the behavioral pattern characteristics, and the following processing is performed: data quality parameters of each modality are obtained from the multimodal fusion device, including RSSI signal strength, data update frequency, historical positioning MSE error, and data packet loss rate; the behavioral pattern characteristics and the data quality parameters of each modality are input as a joint vector into the modal confidence assessment model, and the confidence score of each modality under the current behavior is output; and a multimodal confidence relationship matrix is ​​constructed based on the confidence scores of each modality.

[0008] In a possible implementation, the following processing is performed: the modal confidence assessment model includes a first-layer fully connected network and a second-layer fully connected network, the first-layer fully connected network is a Tanh activation function, and the second-layer fully connected network is a Sigmoid activation function; the first-layer fully connected network and the second-layer fully connected network are trained with mean square error loss using historical sample data, and the first-layer fully connected network and the second-layer fully connected network are output, and the historical sample data includes sample data quality parameters and corresponding confidence labeled samples under different power operation behavior modes.

[0009] In a possible implementation, the modal switching strategy is output using the multimodal confidence relationship matrix, and the following processing is performed: confidence switching rules are set, and the confidence switching rules include a modal retention interval, a modal replacement interval, and a modal suppression interval; according to the confidence size in the multimodal confidence relationship matrix, each modality is marked according to the confidence switching rule, and a modal marking result is output, and the modal marking result includes a retained modality, a backup modality, and a suppressed modality; according to the modal marking result, the current modal activation status is compared, and a modal switching strategy is output.

[0010] In a possible implementation, the following processing is performed: the confidence decrease rate of the modal retention interval is less than a first preset decrease rate, the confidence decrease rate of the modal suppression interval is greater than a second preset decrease rate, and the confidence decrease rate of the modal replacement interval is greater than or equal to the first preset decrease rate and less than or equal to the second preset decrease rate; wherein, the first preset decrease rate is less than the second preset decrease rate.

[0011] In a possible implementation, the current modal activation state is compared according to the modal labeling result, a modal switching strategy is output, and the following processing is performed: if the retained modality in the modal labeling result is different from the current modal activation state, a modal switching instruction is obtained, and the modal fusion device is controlled to output modal positioning data according to the retained modality; if the confidence fluctuation of the retained modality and the backup modality in the modal labeling result is greater than a preset threshold, the retained modality and the backup modality are combined, and the modal fusion device is controlled to output modal positioning data according to the combined modality; if the suppressed modality in the modal labeling result has the same modality as the current modality, the suppressed modality is eliminated.

[0012] In a possible implementation, the multimodal fusion device performs multimodal positioning data analysis according to the modal switching strategy, outputs a protection positioning result, and performs the following processing: the multimodal fusion device performs multimodal positioning data analysis according to the modal switching strategy, and outputs auxiliary positioning data of the weak positioning area; according to the multimodal fusion device, obtains indoor positioning data or outdoor positioning data corresponding to the target performing the power operation task; assists in updating the indoor positioning data or outdoor positioning data according to the auxiliary positioning data, and outputs a protection positioning result.

[0013] The present application also provides a power operation safety protection positioning system integrated with Beidou navigation, including: a weak positioning area identification module, used to identify the weak positioning area in the power operation area, the weak positioning area is the transition area when switching from indoors to outdoors, or from outdoors to indoors; a sliding mode switching positioning activation module, used to obtain the current power operation task, if the path of the power operation task includes the weak positioning area, activate the sliding mode switching positioning module, the sliding mode switching positioning module is connected to the multimodal fusion device, and the multimodal fusion device has Beidou navigation integrated in it; a behavior pattern recognition module, used to identify the behavior pattern of the target performing the power operation task through the sliding mode switching positioning module, and output the behavior pattern characteristics; a modal switching strategy output module, used to establish a multimodal confidence relationship matrix according to the behavior pattern characteristics, and output the modal switching strategy with the multimodal confidence relationship matrix; a protection positioning module, used for the multimodal fusion device to perform multimodal positioning data analysis according to the modal switching strategy, and output the protection positioning result.

[0014] The present application also provides an electronic device, comprising: a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory to implement a power operation safety protection positioning method integrating Beidou navigation.

[0015] The electric power operation safety protection positioning method, system and equipment proposed in this application that integrates Beidou navigation first identify the weak positioning area in the electric power operation area. The weak positioning area is the transition area when switching from indoors to outdoors, or from outdoors to indoors. Then, the current electric power operation task is obtained. If the path of the electric power operation task includes the weak positioning area, the sliding mode switching positioning module is activated. The sliding mode switching positioning module is connected to a multimodal fusion device. The multimodal fusion device integrates Beidou navigation. The sliding mode switching positioning module is used to identify the behavior pattern of the target performing the electric power operation task and output the behavior pattern characteristics. Then, a multimodal confidence relationship matrix is ​​established according to the behavior pattern characteristics. The modal switching strategy is output based on the multimodal confidence relationship matrix. Finally, the multimodal fusion device performs multimodal positioning data analysis according to the modal switching strategy and outputs the protection positioning result. The technical effect of improving the accuracy and reliability of electric power operation safety protection positioning is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention are briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in precise order. Instead, various steps may be processed in reverse order or simultaneously as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0017] Figure 1 A flow chart of a method for electric power operation safety protection and positioning integrated with Beidou navigation provided in an embodiment of the present application.

[0018] Figure 2 A schematic diagram of the structure of the power operation safety protection positioning system integrated with Beidou navigation provided in an embodiment of the present application.

[0019] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0020] Explanation of the accompanying drawings: weak positioning area identification module 10, sliding mode switching positioning activation module 20, behavior pattern recognition module 30, mode switching strategy output module 40, protection positioning module 50, input device 301, processor 302, memory 303, output device 304. DETAILED DESCRIPTION

[0021] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0022] In order to make the purpose, technical solutions and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limiting this application. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0023] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict, and the terms “first\second” involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. The terms “including” and “having” and any variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or that are inherent to these processes, methods, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. The terms used herein are for the purpose of describing the embodiments of this application only.

[0024] The embodiment of the present application provides a power operation safety protection positioning method integrating Beidou navigation, such as Figure 1 As shown, the method includes:

[0025] Step S100: Identify a weak positioning area in the power operation area, where the weak positioning area is a transition area when switching from indoors to outdoors or from outdoors to indoors.

[0026] Specifically, weak positioning areas refer to transition zones between indoors and outdoors, or vice versa. Positioning signals in these areas are unstable, resulting in reduced positioning accuracy. A high-precision sensor network, including Wi-Fi signal strength sensors, Bluetooth signal sensors, and geomagnetic sensors, is deployed within the power operation area. These sensors monitor signal strength and magnetic field variations in real time to identify differences between indoor and outdoor environments. Geographic Information System (GIS) technology is used to create geographic information modeling of the power operation area, dividing the area into indoor and outdoor sections and marking transition zones. The GIS system can integrate sensor data to accurately identify weak positioning areas. Alternatively, machine learning algorithms can be used to analyze sensor data and train models to automatically identify weak positioning areas. For example, support vector machines (SVMs) or decision tree algorithms can be used to automatically delineate indoor-outdoor transition zones based on signal strength and magnetic field variation characteristics. Alternatively, the GIS system can embed machine learning algorithms to combine sensor data with geographic information data to more accurately identify weak positioning areas. The GIS system not only provides geographic information visualization but also uses machine learning algorithms to analyze sensor data and automatically identify weak positioning areas.

[0027] For example, within a power substation, a Wi-Fi signal strength sensor is deployed every 10 meters, and a geomagnetic sensor is deployed every 20 meters around the substation. These sensors collect data in real time and upload it to a central server. In the GIS system, the substation's indoor areas are marked in blue, outdoor areas in green, and transition areas in yellow. Using sensor data, the GIS system can dynamically update the boundaries of these areas. Alternatively, a support vector machine (SVM) model can be trained using historical data, with input features including Wi-Fi signal strength, Bluetooth signal strength, and geomagnetic intensity. The model output is "indoor," "outdoor," or "transition area." Alternatively, in the GIS system, sensor data (such as Wi-Fi signal strength and geomagnetic intensity) can be fused with geographic information data (such as building layout and road conditions) to form a comprehensive dataset. The SVM algorithm is trained on this comprehensive dataset to generate a recognition model. The trained model is applied to real-time data to automatically identify areas with weak localization.

[0028] Step S200: Obtain the current power operation task. If the path of the power operation task includes the weak positioning area, activate the sliding mode switching positioning module. The sliding mode switching positioning module is connected to the multimodal fusion device, and the multimodal fusion device has Beidou navigation integrated therein.

[0029] Specifically, a power operation task management system is established, which records the path, schedule and operation area of ​​each operation task. The task management system communicates with the multimodal fusion device to obtain task information in real time. Among them, the multimodal fusion device is a device that integrates multiple positioning technologies (such as Beidou navigation, Wi-Fi positioning, Bluetooth positioning, etc.), which can fuse data from different modes to perform high-precision positioning. A path analysis algorithm is developed to check whether the operation task path passes through a weak positioning area, wherein the algorithm can automatically determine whether the path contains a weak positioning area based on the GIS map and task path data. When the path analysis algorithm confirms that the path contains a weak positioning area, the sliding mode switching positioning module is activated through the communication interface. The sliding mode switching positioning module is an intelligent module that can dynamically switch between different positioning modes according to the target's behavior pattern and sensor data to improve positioning accuracy.

[0030] For example, the task management system stores detailed information about each work task, including the operator, task start and end points, and expected path. The path analysis algorithm inputs the task path coordinates and GIS map data. The algorithm determines whether the path passes through the weak positioning area by comparing the path coordinates with the coordinates of the weak positioning area in the GIS map. For example, the starting point of task T1 is point A indoors and the end point is point B outdoors. The path coordinates of task T1 overlap with the coordinates of the yellow area in the GIS map, and the algorithm output is "the path contains a weak positioning area." When the path analysis algorithm outputs "the path contains a weak positioning area," the task management system sends an activation command to the multimodal fusion device through the serial port or network interface to activate the sliding mode switching positioning module.

[0031] In step S300 , the sliding mode switching positioning module is used to perform behavior pattern recognition on a target that performs the power operation task and output behavior pattern features.

[0032] Specifically, the sliding mode switching positioning module collects data from various sensors, including accelerometers, gyroscopes, and barometers, to identify the target's behavioral patterns. Machine learning algorithms (such as hidden Markov models (HMMs) or deep learning algorithms) analyze this sensor data and extract features, such as the mean and variance of acceleration and the rotation angle of the gyroscope. These features serve as input features for behavioral pattern recognition, identifying the target's behavior patterns, such as walking, running, and standing still.

[0033] For example, the sliding mode switching positioning module collects accelerometer data at a rate of 100 data points per second. It also collects gyroscope data at a rate of 50 data points per second. The HMM algorithm analyzes the accelerometer and gyroscope data, extracting the mean and variance from the accelerometer data and the rotation angle from the gyroscope data. These features serve as input to the HMM algorithm for behavioral pattern recognition. Based on these data features, the algorithm identifies the target's behavioral pattern. For example, when the accelerometer data has a mean of 0.5g and a variance of 0.1g, and the gyroscope data has a rotation angle of 10 degrees per second, it is identified as a "walking" pattern.

[0034] In one possible implementation, the sliding mode switching positioning module is provided in the edge device, and the behavior pattern of the target performing the power operation task is identified through the edge device. Step S300 further includes step S310, obtaining the target of the power operation task and collecting real-time behavior data of the target. Specifically, the sliding mode switching positioning module is provided in the edge device, and the behavior pattern of the target performing the power operation task is identified through the edge device. A variety of sensors are deployed on the edge device, such as accelerometers, gyroscopes, barometers, GPS / Beidou positioning modules, etc., for collecting real-time behavior data of the target. The acceleration, angular velocity, air pressure change, position coordinates and other data of the target are collected in real time. For example, the accelerometer collects 100 data points per second, the gyroscope collects 50 data points per second, and the GPS / Beidou positioning module collects 1 position coordinate data per second.

[0035] For example, an accelerometer, gyroscope, and GPS / Beidou positioning module are installed on the smart helmet or handheld terminal worn by power workers. The acceleration data collected by the accelerometer is [a x ,a y ,a z ], the angular velocity data collected by the gyroscope is [ω x ,ω y ,ω z ], the location coordinates collected by the GPS / Beidou positioning module are (lat, lon).

[0036] Step S320, identify the spatial change rate, action mutation point and movement direction change of the real-time behavior data, and output the behavior pattern category features. Specifically, calculate the change rate of acceleration and angular velocity to identify the target's motion state. For example, the acceleration change rate , rate of change of angular velocity ,in, represents the acceleration value at time point t, represents the acceleration value at time point t−1, represents the angular velocity value at time point t, represents the angular velocity value at time point t−1, Represents the time interval, that is, the length of time from time point t−1 to time point t. Use signal processing algorithms (such as wavelet transform or threshold detection) to identify the mutation points of acceleration and angular velocity, which indicate the start or end of the action. By analyzing the position coordinate changes of the GPS / Beidou positioning module, calculate the change rate of the motion direction. For example, the change rate of the motion direction , where θ is the angle of the direction of motion, represents the angle of motion direction at time point t, Represents the angle of the movement direction at time point t−1. Based on the above calculation results, the behavior pattern category features are output, such as "walk", "stop", "turn", etc.

[0037] Step S330, collect the target number and target dynamic rate of executing the power operation task, and output the target characteristics of the behavior pattern. Specifically, through the sensor network of the edge device, count the number of targets performing tasks in the same operation area. For example, use the Wi-Fi signal strength or Bluetooth signal strength to detect and count the number of targets. Calculate the dynamic rate of the target, that is, the frequency of the target's motion state change per unit time. For example, the dynamic rate can be calculated by counting the number of action mutation points of the target within a certain period of time. Based on the target number and dynamic rate, output the target characteristics of the behavior pattern, such as "single-person operation", "multi-person operation", "high dynamic rate", etc.

[0038] For example, within a certain operating area, three targets are detected performing tasks based on Wi-Fi signal strength. Target A has five sudden motion changes within one minute, with a dynamic rate of 5 per minute.

[0039] Step S340: Outputting behavior pattern features based on the behavior pattern category features and the behavior pattern target features. Specifically, the behavior pattern category features and the behavior pattern target features are integrated to form a comprehensive behavior pattern feature. For example, category features such as "walk," "stop," and "turn" can be combined with target features such as "single-person operation," "multi-person operation," and "high dynamic rate."

[0040] For example, the behavioral pattern category characteristic of target A is "walking," and the behavioral pattern target characteristics are "single-person operation" and "high dynamic rate." The final output behavioral pattern characteristics are "single-person operation, high dynamic rate, walking." This implementation method, by subdividing behavioral pattern characteristics into category characteristics and target characteristics, achieves high-precision behavioral pattern recognition and adaptability to dynamic environments. It also leverages edge computing to improve real-time performance and system response speed, thereby enhancing the safety protection capabilities of power operations, enabling timely detection and early warning of potential safety risks, and effectively ensuring the safety of operators.

[0041] Step S400: establishing a multimodal confidence relationship matrix according to the behavior pattern characteristics, and outputting a modal switching strategy using the multimodal confidence relationship matrix.

[0042] Specifically, the multimodal confidence relationship matrix is ​​a matrix used to represent the confidence of each positioning mode under different behavior modes. Develop a confidence evaluation algorithm to evaluate the confidence of each positioning mode based on the quality of sensor data and the characteristics of the behavior mode. For example, the confidence is evaluated based on the Wi-Fi signal strength, Beidou signal strength, etc. Construct a multimodal confidence relationship matrix, in which the rows of the matrix represent different positioning modes, the columns represent different behavior modes, and the matrix elements represent the confidence of a certain positioning mode under a specific behavior mode. Based on the multimodal confidence relationship matrix, a modal switching strategy is generated to guide the multimodal fusion device to select the optimal positioning mode under different behavior modes, that is, to select the optimal positioning mode for switching based on the mode with the highest confidence.

[0043] For example, the confidence assessment algorithm for Wi-Fi positioning is as follows: when the signal strength is greater than -50dBm, the confidence is 0.9; when the signal strength is between -50dBm and -70dBm, the confidence is 0.6; and when the signal strength is less than -70dBm, the confidence is 0.3. For Beidou positioning, when the number of satellites is greater than 10, the confidence is 0.9; when the number of satellites is between 5 and 10, the confidence is 0.6; and when the number of satellites is less than 5, the confidence is 0.3. An example of a multimodal confidence relationship matrix is ​​shown in Table 1.

[0044] Table 1: Multimodal confidence relationship matrix example 1

[0045]

[0046] According to the multimodal confidence relationship matrix in Table 1, when the target is in the "walking" mode, the Beidou positioning mode is selected; when the target is in the "stationary" mode, the Wi-Fi positioning mode is selected; when the target is in the "running" mode, the Beidou positioning mode is selected.

[0047] In one possible implementation, a multimodal confidence relationship matrix is ​​established based on the behavioral pattern characteristics. Step S400 further includes step S410, where data quality parameters for each modality are obtained from the multimodal fusion device, including RSSI signal strength, data update frequency, historical positioning mean square error (MSE) error, and data packet loss rate. Specifically, received signal strength indicator (RSSI) values ​​are obtained from modalities such as Wi-Fi, Bluetooth, or Beidou to assess signal strength and stability. The data update frequency of each modality, i.e., the number of updates per second, is recorded to assess data real-time performance. The historical mean square error (MSE) of each modality is calculated to assess positioning accuracy. The data packet loss rate of each modality, i.e., the proportion of lost data packets to total data packets, is recorded to assess data reliability.

[0048] For example, the RSSI for Wi-Fi mode is -60dBm, the RSSI for Bluetooth mode is -70dBm, and the RSSI for Beidou mode is -50dBm. Wi-Fi mode updates 10 times per second, Bluetooth mode updates 5 times per second, and Beidou mode updates once per second. The mean square error (MSE) for Wi-Fi mode is 1.2 meters, the MSE for Bluetooth mode is 2.0 meters, and the MSE for Beidou mode is 0.5 meters. The packet loss rate for Wi-Fi mode is 5%, the packet loss rate for Bluetooth mode is 10%, and the packet loss rate for Beidou mode is 2%.

[0049] In step S420, the behavioral pattern features and the data quality parameters of each modality are input into the modal confidence assessment model as a joint vector, which outputs a confidence score for each modality under the current behavior. Specifically, the behavioral pattern features (e.g., "single-person operation, high dynamic rate, walking") and data quality parameters (RSSI, update frequency, MSE, packet loss rate) are combined into a joint vector. This joint vector is evaluated using a machine learning model (e.g., support vector machine (SVM), random forest, or neural network) to output a confidence score for each modality under the current behavior.

[0050] For example, for target A, the joint vector is [current operation mode, target dynamic rate, target behavior pattern, Wi-Fi RSSI value, Wi-Fi data update frequency, Wi-Fi historical positioning MSE error, Wi-Fi data packet loss rate, Bluetooth RSSI value, Bluetooth data update frequency, Bluetooth historical positioning MSE error, Bluetooth data packet loss rate, Beidou RSSI value, Beidou data update frequency, Beidou historical positioning MSE error, Beidou data packet loss rate]. The specific data is: [single-person operation, high dynamic rate, walking, −60, 10, 1.2, 5, −70, 5, 2.0, 10, −50, 1, 0.5, 2]. The joint vector is evaluated using a random forest model, outputting a confidence score for each mode. For example, the confidence level of the Wi-Fi modality is 0.7, the confidence level of the Bluetooth modality is 0.4, and the confidence level of the Beidou modality is 0.9.

[0051] Step S430, constructing a multimodal confidence relationship matrix based on the confidence scores of each modality. Specifically, a multimodal confidence relationship matrix is ​​constructed based on the confidence scores of each modality. The rows of the matrix represent different behavior patterns, the columns represent different modes, and the matrix elements represent the confidence of a certain positioning mode under a specific behavior pattern. Among them, an example of a multimodal confidence relationship matrix is ​​shown in Table 2. This implementation method combines the behavior pattern characteristics and data quality parameters to construct a multimodal confidence relationship matrix, thereby realizing an accurate assessment of the confidence of different modalities under different behavior patterns. This not only improves the accuracy and reliability of positioning, but also enhances the system's adaptability, and can dynamically adjust the positioning mode according to real-time data, thereby improving the safety and efficiency of power operations.

[0052] Table 2: Multimodal confidence relationship matrix example 2

[0053]

[0054] In one possible implementation, step S420 further includes step S421, wherein the modal confidence assessment model includes a first-layer fully connected network and a second-layer fully connected network, wherein the first-layer fully connected network uses a Tanh activation function, and the second-layer fully connected network uses a Sigmoid activation function. Specifically, the structure of the modal confidence assessment model includes a first-layer fully connected network and a second-layer fully connected network, wherein the first-layer fully connected network is a fully connected network layer using a Tanh activation function, and is used to perform a nonlinear transformation on the input joint vector and extract features. The second-layer fully connected network is a fully connected network layer using a Sigmoid activation function, and is used to map the extracted features to the [0,1] interval and output a confidence score for each modality.

[0055] For example, the first layer has 128 neurons and uses the Tanh activation function. After the input joint vector passes through the first layer's fully connected network, it outputs a 128-dimensional feature vector. The second layer has three neurons (corresponding to Wi-Fi, Bluetooth, and BeiDou modalities) and uses the Sigmoid activation function. After inputting the 128-dimensional feature vector, it outputs three confidence scores, for example, [0.7, 0.4, 0.9].

[0056] Step S422: Use historical sample data to train the first and second fully connected networks using mean squared error loss, outputting the first and second fully connected networks. The historical sample data includes sample data quality parameters and corresponding confidence-labeled samples for different power operation behavior patterns. Specifically, sample data quality parameters and corresponding confidence-labeled samples for different power operation behavior patterns are collected. This sample data is used to train the model. Using mean squared error (MSE) as the loss function, the model is trained using a backpropagation algorithm to optimize model parameters, ensuring that the predicted confidence score is as close as possible to the labeled confidence score.

[0057] For example, the historical sample data includes 1000 samples, each of which contains behavioral pattern characteristics, RSSI signal strength, data update frequency, historical positioning MSE error, data packet loss rate, and corresponding confidence annotation. The model is trained using the mean square error loss function. Assuming that the labeled confidence is [0.7, 0.4, 0.9] and the model prediction confidence is [0.65, 0.38, 0.88], the mean square error is: MSE = 0.0011. The backpropagation algorithm adjusts the model's weights and biases to minimize the mean squared error. This approach, through the design of a modal confidence assessment model consisting of a two-layer fully connected network and training with a mean squared error loss using historical sample data, enables accurate confidence assessment of different modalities under different behavior patterns.

[0058] In one possible implementation, the modal switching strategy is output using the multimodal confidence relationship matrix, and step S400 further includes step S440, setting confidence switching rules, wherein the confidence switching rules include a modal retention interval, a modal replacement interval, and a modal suppression interval. Specifically, a confidence range is defined, and when the confidence of a modality is within this range, the modality is marked as a retained modality, i.e., the modality that is primarily used. A confidence range is defined, and when the confidence of a modality is within this range, the modality is marked as a backup modality, i.e., a modality that can be switched to when the main modality is unavailable. A confidence range is defined, and when the confidence of a modality is within this range, the modality is marked as a suppressed modality, i.e., a modality that is not recommended for use. For example, the confidence switching rules are: modal retention interval: confidence greater than 0.8; modal replacement interval: confidence between 0.5 and 0.8; modal suppression interval: confidence less than 0.5.

[0059] Step S450, according to the confidence level in the multimodal confidence relationship matrix, each modality is marked according to the confidence switching rule, and a modality marking result is output. The modality marking result includes a retained modality, a spare modality, and a suppressed modality. Specifically, each modality is evaluated according to the confidence value in the multimodal confidence relationship matrix. According to the confidence switching rule, each modality is marked as a retained modality, a spare modality, or a suppressed modality. For example, for the multimodal confidence relationship matrix in Table 2, the modality marking result is shown in Table 3.

[0060] Table 3: Example of modality labeling results

[0061]

[0062] Step S460, compare the current mode activation state according to the mode marking result, and output the mode switching strategy. Specifically, compare the mode marking result with the activation state of the current mode to determine whether mode switching is required. Based on the comparison result, output the mode switching strategy to decide whether to switch to the backup mode or suppress the current mode. This implementation method sets the confidence switching rules and marks each mode according to the multi-modal confidence relationship matrix, and finally outputs the mode switching strategy. It not only improves the accuracy and reliability of positioning, but also enhances the system's adaptability. It can dynamically adjust the positioning mode according to real-time data, thereby improving the safety and efficiency of power operations.

[0063] In one possible implementation, step S440 further includes step S441, wherein the confidence decrease rate of the modal retention interval is less than a first preset decrease rate, the confidence decrease rate of the modal suppression interval is greater than a second preset decrease rate, and the confidence decrease rate of the modal replacement interval is greater than or equal to the first preset decrease rate and less than or equal to the second preset decrease rate; wherein the first preset decrease rate is less than the second preset decrease rate.

[0064] Specifically, the confidence decline rate refers to the rate at which the confidence of a modal changes over time. It is the change in confidence divided by the time interval. For example, if the confidence of a modal decreases from 0.8 to 0.7 in 1 second, then the decline rate is (0.8-0.7) / 1 = 0.1 / second.

[0065] A confidence range is defined. When the confidence of a modality is within this range, the modality is marked as a retained modality. At the same time, a confidence decrease rate threshold is defined. When the confidence decrease rate of a modality is less than a first preset decrease rate, the modality is still retained. A confidence range is defined. When the confidence of a modality is within this range, the modality is marked as a suppressed modality. At the same time, a confidence decrease rate threshold is defined. When the confidence decrease rate of a modality is greater than a second preset decrease rate, the modality is suppressed. A confidence range is defined. When the confidence of a modality is within this range, the modality is marked as a backup modality. At the same time, a confidence decrease rate threshold is defined. When the confidence decrease rate of a modality is greater than or equal to the first preset decrease rate and less than or equal to the second preset decrease rate, the modality is marked as a backup modality.

[0066] For example, assume the first preset descent rate is 0.05 / second and the second preset descent rate is 0.1 / second. If the Wi-Fi positioning confidence is 0.7, the descent rate is 0.03 / second. 0.7 is between 0.5 and 0.8, which falls within the modality replacement range; 0.03 / second is less than the first preset descent rate (0.05 / second). According to the rules, although the Wi-Fi positioning confidence is within the modality replacement range, its descent rate is less than the first preset descent rate, so it is marked as the retained modality.

[0067] This implementation method uses the confidence drop rate to more accurately determine the stability of the mode. For example, even if the confidence of a mode is within the mode replacement range, if its drop rate is low, it means that the mode is still relatively stable and can be retained. At the same time, the system can dynamically adjust the mode based on real-time data. For example, when the confidence drop rate of a mode exceeds the threshold, the system can promptly mark it as a suppressed mode to avoid using the unstable mode. Through more precise mode selection, the system can improve the accuracy and reliability of positioning, thereby improving the safety and efficiency of power operations.

[0068] In one possible implementation, the current mode activation state is compared with the modality tagging result, and a modality switching strategy is output. Step S460 further includes step S461: If the retained mode in the modality tagging result is different from the current mode activation state, a modality switching instruction is obtained, and the modality fusion device is controlled to output modal positioning data according to the retained mode. Specifically, the retained mode in the modality tagging result is compared with the current mode activation state. If the retained mode is different from the currently activated mode, a modality switching instruction is obtained. Based on the modality switching instruction, the modality fusion device is controlled to output modal positioning data according to the retained mode, ensuring that the system always uses the most reliable modality for positioning. For example, the modality tagging result is: retained mode: Beidou positioning; backup mode: Wi-Fi positioning; suppressed mode: Bluetooth positioning. The current mode activation state is: currently activated mode: Wi-Fi positioning. The comparison result is: the retained mode (Beidou positioning) is different from the currently activated mode (Wi-Fi positioning). The mode switching command is: switch to Beidou positioning. Control the modal fusion device to output modal positioning data according to Beidou positioning.

[0069] In step S462, if the confidence fluctuation of the retained modality and the backup modality in the modality labeling result is greater than a preset threshold, the retained modality and the backup modality are combined, and the modal fusion device is controlled to output modal positioning data according to the combined modality. Specifically, a test is performed to determine whether the confidence fluctuation of the retained modality and the backup modality is greater than a preset threshold. If the confidence fluctuation is greater than the preset threshold, the retained modality and the backup modality are combined. The modal fusion device is controlled to output modal positioning data according to the combined modality to improve the stability and reliability of positioning and reduce the impact of the instability of a single modality on positioning accuracy.

[0070] Step S463: If the suppressed mode in the modal marking result is the same as the current mode, the suppressed mode is eliminated. Specifically, the suppressed mode in the modal marking result is compared with the current modal activation state. If the suppressed mode is the same as the currently activated mode, the suppressed mode is eliminated to avoid using an unstable mode for positioning. This implementation method dynamically adjusts the output mode of the modal fusion device by comparing the modal marking result with the current modal activation state in detail, which not only improves the accuracy and reliability of positioning, but also enhances the system's adaptability, and can dynamically adjust the positioning mode according to real-time data, thereby improving the safety and efficiency of power operations.

[0071] In step S500 , the multimodal fusion device performs multimodal positioning data analysis according to the modal switching strategy and outputs a protection positioning result.

[0072] Specifically, a multimodal positioning algorithm is developed. Based on a modal switching strategy, data from different positioning modalities is integrated to achieve high-precision positioning. For example, a Kalman filter algorithm is used to fuse Beidou and Wi-Fi positioning data. This fused positioning data is then combined with safety protection rules to output a safety positioning result, which is used to guide operators in safe operations. For example, an alarm is issued when a target approaches a dangerous area.

[0073] For example, a Kalman filter algorithm is used to fuse Beidou positioning data with Wi-Fi positioning data. The Kalman filter dynamically adjusts the weights based on the confidence level of the Beidou and Wi-Fi positioning data, outputting high-precision positioning results. This positioning result is then integrated with safety regulations for power operation areas. For example, if a target approaches a high-voltage area, the system issues an alarm, alerting operators to safety.

[0074] In one possible implementation, the multimodal fusion device analyzes multimodal positioning data based on the modal switching strategy and outputs a protective positioning result. Step S500 further includes step S510, in which the multimodal fusion device analyzes multimodal positioning data based on the modal switching strategy and outputs auxiliary positioning data for the weak positioning area. Specifically, based on the modal switching strategy, the multimodal fusion device analyzes positioning data from different modalities to generate auxiliary positioning data. This auxiliary positioning data is used to provide more accurate positioning information in weak positioning areas (such as the transition zone between indoors and outdoors). For example, if the retained mode is Beidou positioning and the backup mode is Wi-Fi positioning, in the weak positioning area, the confidence level of Beidou positioning is 0.8, and the confidence level of Wi-Fi positioning is 0.7. The generated auxiliary positioning data includes high-precision location information from Beidou positioning and auxiliary information from Wi-Fi positioning, and is used to improve positioning accuracy.

[0075] In step S520, the multimodal fusion device obtains indoor or outdoor positioning data corresponding to the target performing the power operation task. Specifically, indoor positioning technology (such as Wi-Fi, Bluetooth, and UWB) is used to obtain the target's indoor positioning data, which contains inaccurate information in weakly positioned areas. Outdoor positioning technology (such as Beidou and GPS) is used to obtain the target's outdoor positioning data, which also contains inaccurate information in weakly positioned areas.

[0076] Step S530, the indoor positioning data or outdoor positioning data is updated according to the auxiliary positioning data, and the protection positioning result is output. Specifically, the indoor or outdoor positioning data is updated according to the auxiliary positioning data to improve the positioning accuracy. The auxiliary positioning data provides more accurate positioning information, especially in weak positioning areas. The final protection positioning result is output to ensure that high-precision positioning information can be provided even in weak positioning areas. This implementation method significantly improves the accuracy and reliability of positioning by generating auxiliary positioning data in weak positioning areas and updating indoor or outdoor positioning data based on the auxiliary positioning data. This not only enhances the system's adaptability, but also ensures high-precision positioning in complex environments, significantly improving the safety and efficiency of power operations.

[0077] The embodiment of the present application adopts a technical means of first identifying the weak positioning area of ​​indoor and outdoor switching in the power operation area, and activating the sliding mode switching positioning module if the path contains this area after obtaining the operation task. The module recognizes the behavior pattern of the operation target and outputs the characteristics, and then establishes a multimodal confidence relationship matrix based on this to output the modal switching strategy. Finally, the multimodal fusion device performs multimodal positioning data analysis according to this strategy and outputs the protection positioning results. It solves the technical problems of unstable positioning signals and decreased accuracy when switching in complex environments in the existing power operation safety protection positioning, and achieves the technical effect of improving the accuracy and reliability of power operation safety protection positioning.

[0078] In the above, refer to Figure 1 The power operation safety protection positioning method integrating Beidou navigation according to the embodiment of the present invention is described in detail. Figure 2 The following describes a power operation safety protection positioning system integrated with Beidou navigation according to an embodiment of the present invention.

[0079] The electric power operation safety protection positioning system integrated with Beidou navigation, according to an embodiment of the present invention, is designed to address the technical issues of unstable positioning signals and decreased accuracy during complex environment switching, thereby improving the accuracy and reliability of electric power operation safety protection positioning. The electric power operation safety protection positioning system integrated with Beidou navigation includes a weak positioning area identification module 10, a sliding mode switching positioning activation module 20, a behavioral pattern recognition module 30, a mode switching strategy output module 40, and a protection positioning module 50.

[0080] A weak positioning area identification module 10 is used to identify a weak positioning area in the power operation area, where the weak positioning area is a transition area when switching from indoors to outdoors, or from outdoors to indoors; a sliding mode switching positioning activation module 20 is used to obtain the current power operation task. If the path of the power operation task includes the weak positioning area, the sliding mode switching positioning module is activated. The sliding mode switching positioning module is connected to a multimodal fusion device, and the multimodal fusion device has Beidou navigation integrated therein; a behavior pattern recognition module 30 is used to perform behavior pattern recognition on the target performing the power operation task through the sliding mode switching positioning module and output behavior pattern characteristics; a modal switching strategy output module 40 is used to establish a multimodal confidence relationship matrix according to the behavior pattern characteristics and output a modal switching strategy using the multimodal confidence relationship matrix; a protection positioning module 50 is used for the multimodal fusion device to perform multimodal positioning data analysis according to the modal switching strategy and output a protection positioning result.

[0081] The specific configuration of the behavior pattern recognition module 30 will be described in detail below. As described above, the sliding mode switching positioning module is set in the edge device, and the behavior pattern recognition of the target performing the power operation task is performed through the edge device. The behavior pattern recognition module 30 can further include: a real-time behavior data acquisition unit for acquiring the target of the power operation task and collecting the real-time behavior data of the target; a behavior pattern category feature output unit for identifying the spatial change rate, action mutation point and movement direction change of the real-time behavior data, and outputting the behavior pattern category feature; a behavior pattern target feature output unit for collecting the number of targets and target dynamic rate performing the power operation task and outputting the behavior pattern target feature; a behavior pattern feature output unit for outputting the behavior pattern feature according to the behavior pattern category feature and the behavior pattern target feature.

[0082] The specific configuration of the modal switching strategy output module 40 will be described in detail below. As described above, a multimodal confidence relationship matrix is ​​established according to the behavioral pattern characteristics. The modal switching strategy output module 40 may further include: a data quality parameter acquisition unit for acquiring data quality parameters of each modality from the multimodal fusion device, including RSSI signal strength, data update frequency, historical positioning MSE error, and data packet loss rate; a modal confidence assessment unit for inputting the behavioral pattern characteristics and the data quality parameters of each modality as a joint vector into a modal confidence assessment model, and outputting a confidence score for each modality under the current behavior; and a multimodal confidence relationship matrix construction unit for constructing a multimodal confidence relationship matrix based on the confidence scores of each modality.

[0083] Among them, the modal confidence assessment unit can further include: a modal confidence assessment model construction subunit for constructing a modal confidence assessment model, the modal confidence assessment model including a first-layer fully connected network and a second-layer fully connected network, the first-layer fully connected network is a Tanh activation function, and the second-layer fully connected network is a Sigmoid activation function; a training subunit for performing mean square error loss training on the first-layer fully connected network and the second-layer fully connected network using historical sample data, and outputting the first-layer fully connected network and the second-layer fully connected network, the historical sample data including sample data quality parameters and corresponding confidence labeled samples under different power operation behavior modes.

[0084] Among them, the modal switching strategy is outputted using the multimodal confidence relationship matrix, and the modal switching strategy output module 40 may further include: a confidence switching rule setting unit for setting the confidence switching rule, wherein the confidence switching rule includes a modal retention interval, a modal replacement interval, and a modal suppression interval; a modal marking unit for marking each modality according to the confidence size in the multimodal confidence relationship matrix and the confidence switching rule, and outputting a modal marking result, wherein the modal marking result includes a retained modality, a standby modality, and a suppressed modality; a modal switching strategy output unit for comparing the current modal activation state according to the modal marking result, and outputting a modal switching strategy.

[0085] In which, the confidence switching rule setting unit may further include: the confidence decrease rate of the modal retention interval is less than the first preset decrease rate, the confidence decrease rate of the modal suppression interval is greater than the second preset decrease rate, and the confidence decrease rate of the modal replacement interval is greater than or equal to the first preset decrease rate and less than or equal to the second preset decrease rate; wherein, the first preset decrease rate is less than the second preset decrease rate.

[0086] Among them, the current modal activation state is compared according to the modal labeling result, and a modal switching strategy is output. The modal switching strategy output unit may further include: a first modal positioning data output subunit is used to obtain a modal switching instruction if the retained modality in the modal labeling result is different from the current modal activation state, and control the modal fusion device to output modal positioning data according to the retained modality; a second modal positioning data output subunit is used to combine the retained modality and the backup modality if the confidence fluctuation of the retained modality and the backup modality in the modal labeling result is greater than a preset threshold, and control the modal fusion device to output modal positioning data according to the combined modality; a suppressed modality elimination subunit is used to eliminate the suppressed modality if the suppressed modality in the modal labeling result has the same modality as the current modality.

[0087] The specific configuration of the protection positioning module 50 will be described in detail below. As described above, the multimodal fusion device performs multimodal positioning data analysis according to the modal switching strategy and outputs a protection positioning result. The protection positioning module 50 may further include: an auxiliary positioning data output unit configured to cause the multimodal fusion device to perform multimodal positioning data analysis according to the modal switching strategy and output auxiliary positioning data for the weak positioning area; an indoor / outdoor positioning data acquisition unit configured to acquire, using the multimodal fusion device, indoor positioning data or outdoor positioning data corresponding to the target performing the power operation task; and a protection positioning result output unit configured to assist in updating the indoor positioning data or outdoor positioning data according to the auxiliary positioning data and output the protection positioning result.

[0088] The electric power operation safety protection positioning system integrated with Beidou navigation provided in an embodiment of the present invention can execute the electric power operation safety protection positioning method integrated with Beidou navigation provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0089] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention.

[0090] Based on the foregoing embodiments, an embodiment of the present application further provides an electronic device. Figure 3 1 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention, showing a block diagram of an exemplary electronic device suitable for implementing an embodiment of the present invention. Figure 3 The electronic device shown is merely an example and should not limit the functionality and scope of use of the embodiments of the present invention. The electronic device is implemented as a general-purpose computing device, and its components may include, but are not limited to, an input device 301, a processor 302, a memory 303, and an output device 304. The processor 302 may be one or more; the memory 303 may include a computer-readable medium and at least one program product, which has a set (at least one) of program modules configured to perform the functions of the various embodiments of the present application.

[0091] The memory 303 shown in the embodiment of the present invention can adopt any combination of one or more computer-readable media; the computer-readable storage medium can be but not limited to infrared, semiconductor systems, devices or components, or any combination of the above, for storing software programs, computer executable programs and modules, such as the program instructions / modules corresponding to the power operation safety protection and positioning method integrated with Beidou navigation in the embodiment of the present invention. The processor 302 executes various functional applications and data processing of the computer device by running the software programs, instructions and modules stored in the memory 303, thereby realizing the above-mentioned power operation safety protection and positioning method integrated with Beidou navigation.

[0092] The above specific embodiments do not constitute a limitation to the scope of protection of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of this application should be included in the scope of protection of this application. In some cases, the actions or steps recorded in this application can be performed in an order different from that in the embodiments and can still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. The power operation safety protection positioning method integrating Beidou navigation is characterized by: The method comprises: Identify weak positioning areas in the power operation area, where the weak positioning areas are transition areas when switching from indoors to outdoors or from outdoors to indoors; Obtaining a current power operation task. If the path of the power operation task includes the weak positioning area, activating a sliding mode switching positioning module, wherein the sliding mode switching positioning module is connected to a multimodal fusion device, wherein the multimodal fusion device has integrated Beidou navigation. The sliding mode switching positioning module is used to identify the behavior pattern of the target performing the power operation task and output the behavior pattern characteristics; Establishing a multimodal confidence relationship matrix according to the behavior pattern characteristics, and outputting a modal switching strategy using the multimodal confidence relationship matrix; The multimodal fusion device performs multimodal positioning data analysis according to the modal switching strategy and outputs a protection positioning result; The sliding mode switching positioning module is provided in an edge device, and the behavior pattern recognition of the target performing the power operation task is performed by the edge device. The method includes: Obtaining the target of the power operation task and collecting real-time behavior data of the target; Identify the spatial change rate, action mutation point, and movement direction change of the real-time behavior data, and output behavior pattern category features; Collecting the target number and target dynamic rate of executing the power operation task, and outputting the target characteristics of the behavior pattern; According to the behavior pattern category feature and the behavior pattern target feature, the behavior pattern feature is output.

2. The power operation safety protection positioning method integrated with Beidou navigation as claimed in claim 1 is characterized in that: A multimodal confidence relationship matrix is ​​established according to the behavior pattern characteristics, the method comprising: Obtaining data quality parameters of each modality from the multimodal fusion device, including RSSI signal strength, data update frequency, historical positioning MSE error, and data packet loss rate; Inputting the behavior pattern characteristics and the data quality parameters of each mode as a joint vector into the modal confidence assessment model, and outputting the confidence score of each mode under the current behavior; A multimodal confidence relationship matrix is ​​constructed based on the confidence scores of each modality.

3. The power operation safety protection positioning method integrated with Beidou navigation as claimed in claim 2 is characterized in that: The modal confidence assessment model includes a first-layer fully connected network and a second-layer fully connected network, the first-layer fully connected network is a Tanh activation function, and the second-layer fully connected network is a Sigmoid activation function; The first-layer fully connected network and the second-layer fully connected network are trained with mean square error loss using historical sample data, and the first-layer fully connected network and the second-layer fully connected network are output. The historical sample data includes sample data quality parameters and corresponding confidence-labeled samples under different power operation behavior modes.

4. The power operation safety protection positioning method integrated with Beidou navigation as claimed in claim 1 is characterized in that: Outputting a modal switching strategy using the multimodal confidence relationship matrix, the method comprising: Setting a confidence switching rule, wherein the confidence switching rule includes a mode retention interval, a mode replacement interval, and a mode suppression interval; According to the confidence levels in the multimodal confidence relationship matrix, each modality is marked according to the confidence switching rule, and a modality marking result is output, where the modality marking result includes a retained modality, a standby modality, and a suppressed modality; Compare the current modal activation status according to the modal marking result and output the modal switching strategy.

5. The power operation safety protection positioning method integrated with Beidou navigation as claimed in claim 4 is characterized in that: The confidence decrease rate of the modal retention interval is less than a first preset decrease rate, the confidence decrease rate of the modal suppression interval is greater than a second preset decrease rate, and the confidence decrease rate of the modal replacement interval is greater than or equal to the first preset decrease rate and less than or equal to the second preset decrease rate; The first preset descent rate is smaller than the second preset descent rate.

6. The power operation safety protection positioning method integrated with Beidou navigation as claimed in claim 4 is characterized in that: Compare the current mode activation status according to the modal marking result and output the modal switching strategy. The method includes: If the retained modality in the modality marking result is different from the current modality activation state, obtaining a modality switching instruction and controlling the modality fusion device to output modality positioning data according to the retained modality; If the confidence fluctuation of the retained modality and the backup modality in the modality labeling result is greater than a preset threshold, combining the retained modality and the backup modality, and controlling the modality fusion device to output modality positioning data according to the combined modality; If the suppressed mode in the modality labeling result has the same mode as the current mode, the suppressed mode is eliminated.

7. The power operation safety protection positioning method integrated with Beidou navigation as claimed in claim 1 is characterized in that: The multimodal fusion device performs multimodal positioning data analysis according to the modal switching strategy and outputs a protection positioning result. The method includes: The multimodal fusion device performs multimodal positioning data analysis according to the modal switching strategy, and outputs auxiliary positioning data of the weak positioning area; Acquiring, according to the multimodal fusion device, indoor positioning data or outdoor positioning data corresponding to a target for performing the power operation task; The indoor positioning data or the outdoor positioning data is updated according to the auxiliary positioning data, and the protection positioning result is output.

8. The power operation safety protection positioning system integrated with Beidou navigation is characterized by: The system is used to implement the power operation safety protection positioning method integrated with Beidou navigation according to any one of claims 1 to 7, and the system includes: A weak positioning area identification module is used to identify a weak positioning area in the power operation area, wherein the weak positioning area is a transition area when switching from indoor to outdoor or from outdoor to indoor; A sliding mode switching positioning activation module is used to obtain the current power operation task. If the path of the power operation task includes the weak positioning area, the sliding mode switching positioning module is activated. The sliding mode switching positioning module is connected to the multimodal fusion device, and the multimodal fusion device has Beidou navigation integrated in it. a behavior pattern recognition module, configured to perform behavior pattern recognition on a target performing the power operation task through the sliding mode switching positioning module, and output behavior pattern characteristics; A modal switching strategy output module is used to establish a multimodal confidence relationship matrix according to the behavior pattern characteristics and output a modal switching strategy based on the multimodal confidence relationship matrix; The protection positioning module is used for the multimodal fusion device to perform multimodal positioning data analysis according to the modal switching strategy and output a protection positioning result.

9. An electronic device, characterized in that: The electronic device comprises: a memory for storing executable instructions; The processor is configured to implement the power operation safety protection and positioning method integrating Beidou navigation as described in any one of claims 1 to 7 when executing the executable instructions stored in the memory.

Images