A dynamic fusion method and system of inertial navigation data

Abstract

The application discloses a kind of inertial navigation data dynamic fusion method and system, method includes: by inertial navigation system real-time acquisition of original inertial navigation data of different positions of traverse, first data is obtained after preprocessing;The inertial navigation system data of different positions in first data is carried out time synchronization processing, to obtain time synchronization inertial navigation data;Based on the accuracy and stability of inertial navigation system historical data, each inertial navigation system is allocated dynamic weight;The state of next time step is predicted by linear dynamic system equation, when new time synchronization inertial navigation data arrives, state estimation and error covariance matrix are updated using Kalman gain, dynamic weight and new time synchronization inertial navigation data, to output the fusion data set processed by Kalman filter.This application can effectively reduce the influence of environmental noise and equipment error on monitoring result;By dynamically adjusting weight and fusion strategy, it can better adapt to various complex and changeable environmental conditions, and maintain high stability.

Description

Technical Field

[0001] This invention relates to the field of power information technology processing technology, and in particular to a dynamic fusion method and system for inertial navigation data. Background Technology

[0002] In the power sector, real-time conductor monitoring is crucial for ensuring the safe and stable operation of the system. With technological advancements, Inertial Navigation Systems (INS) have been widely applied to various navigation and positioning tasks. However, in real-time conductor monitoring, data from a single INS may contain errors and uncertainties due to environmental noise, equipment malfunctions, and various other factors. To improve data accuracy and reliability, multiple INS systems are typically installed at different locations along the conductor.

[0003] Traditional data fusion methods often rely on simple weighted averaging or voting mechanisms. These methods may not adequately consider the performance and characteristics of each inertial navigation system, thus lacking the ability to dynamically adjust and optimize data fusion weights and strategies based on real-time conditions. Therefore, effectively fusing data from different locations and systems to obtain an accurate and comprehensive guideline state assessment has become a pressing problem. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a dynamic fusion method and system for inertial navigation data to solve the problem that current data fusion lacks the ability to dynamically adjust, cannot optimize the weights and strategies of data fusion according to real-time conditions, and cannot effectively fuse data from different locations and systems for state assessment.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a method for dynamic fusion of inertial navigation data, comprising:

[0008] The raw inertial navigation data at different positions of the conductor is acquired in real time through the inertial navigation system, and the first data is obtained after preprocessing.

[0009] Time synchronization processing is performed on the inertial navigation system data at different locations in the first data to obtain time-synchronized inertial navigation data;

[0010] Based on the accuracy and stability of historical data of inertial navigation systems, dynamic weights are assigned to each inertial navigation system.

[0011] The state at the next time step is predicted by linear dynamic system equations. When new time-synchronized inertial navigation data arrives, the state estimate and error covariance matrix are updated using Kalman gain, dynamic weights, and the new time-synchronized inertial navigation data to output a fused dataset processed by a Kalman filter.

[0012] As a preferred embodiment of the dynamic fusion method for inertial navigation data described in this invention, wherein:

[0013] The raw inertial navigation data at different positions of the conductor is acquired in real time through the inertial navigation system, including:

[0014] Define the optimization objectives and constraints for the inertial navigation system;

[0015] Based on the optimization objective and constraints, the particle swarm optimization algorithm is used to search for the optimal solution, evaluate the objective function values ​​of different system layout schemes, and determine the final layout position of the system based on the optimal solution.

[0016] Based on the final layout, inertial navigation systems are installed at the corresponding positions of the traverse lines to obtain real-time data at each position.

[0017] As a preferred embodiment of the dynamic fusion method for inertial navigation data described in this invention, wherein:

[0018] The preprocessing includes: setting the window size of the moving average filter to N;

[0019] Initialize a buffer of size N to store the N most recent data points;

[0020] When a new data point arrives, add the new data point to the end of the buffer;

[0021] If the buffer is full, remove the first element of the buffer so that the buffer always contains the latest N data points;

[0022] Calculate the moving average of the latest N data points contained in the buffer to obtain the first data.

[0023] As a preferred embodiment of the dynamic fusion method for inertial navigation data described in this invention, wherein:

[0024] The time synchronization process includes:

[0025] Create an empty list of time-synchronized data and determine the unit of synchronization;

[0026] For each dataset, create a mapping table from timestamps to data values;

[0027] For each time point t within the common time range, if time point t exists in the data mapping of the dataset, the data value is added to the position of the corresponding time point in the time synchronization data list; if it does not exist, the data value of the time point is estimated using linear interpolation, added to the time synchronization data list, and the time synchronization inertial navigation data is output.

[0028] As a preferred embodiment of the dynamic fusion method for inertial navigation data described in this invention, it further includes:

[0029] The fused dataset is monitored in real time, and the operating status of the conductor is determined by a preset threshold.

[0030] If the monitored data exceeds the preset threshold, an alarm will be automatically triggered, and a potential abnormality in the conductor status will be indicated.

[0031] As a preferred embodiment of the dynamic fusion method for inertial navigation data described in this invention, wherein:

[0032] The dynamic weight is represented as:

[0033]

[0034] Among them, w i (t-1) represents the weight of the i-th inertial navigation system at the previous time t-1; β∈(0,1) represents the forgetting factor; α∈(0,1) represents the adjustment coefficient; This represents the accuracy evaluation index of the i-th inertial navigation system at time t; This is the stability evaluation index of the i-th inertial navigation system at time t, where N is a constant representing the number of data points used to calculate the accuracy and stability evaluation index, t represents the time index, i represents the current time, i represents the index of the inertial navigation system, j represents the index of the data point, and p... ij p represents the predicted value of the i-th inertial navigation system for the j-th data point. ij (t) represents the prediction value of the i-th inertial navigation system for the j-th data point at time t, r ij r represents the actual observation value of the j-th data point for the i-th inertial navigation system. ij (t) represents the actual observed value at time t; ij This represents the output value of the i-th inertial navigation system when processing the _i-th data point, o ii (t) represents the output value of the inertial navigation system when it processes the j-th data point at time t; It is the average value of all output values ​​of the inertial navigation system at time t.

[0035] As a preferred embodiment of the dynamic fusion method for inertial navigation data described in this invention, wherein:

[0036] It also includes initializing the relevant parameters of the Kalman filter to obtain the linear dynamic system equations. The relevant parameters include: initial state, initial error covariance, state transition matrix, observation matrix, process noise covariance matrix, and observation noise covariance matrix.

[0037] Secondly, the present invention provides a dynamic fusion system for inertial navigation data, comprising:

[0038] The acquisition module is used to acquire raw inertial navigation data at different positions of the conductor in real time through the inertial navigation system, and obtain the first data after preprocessing;

[0039] The synchronization module is used to perform time synchronization processing on the inertial navigation system data at different locations in the first data to obtain time-synchronized inertial navigation data;

[0040] The allocation module is used to assign dynamic weights to each inertial navigation system based on the accuracy and stability of historical data.

[0041] The fusion module is used to predict the state of the next time step through linear dynamic system equations. When new time-synchronized inertial navigation data arrives, it updates the state estimate and error covariance matrix using Kalman gain, dynamic weights, and the new time-synchronized inertial navigation data to output a fused dataset processed by a Kalman filter.

[0042] Thirdly, the present invention provides a computing device, comprising:

[0043] Memory and processor;

[0044] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the dynamic fusion method of inertial navigation data.

[0045] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the dynamic fusion method for inertial navigation data.

[0046] Compared with existing technologies, the advantages of this invention are as follows: This invention utilizes the historical data accuracy and stability information of each inertial navigation system to obtain more accurate fusion results, which helps to more precisely assess the real-time status of the guide rail and promptly identify and address potential safety hazards. Using a Kalman filter for data fusion effectively reduces the impact of environmental noise and equipment errors on the monitoring results. Simultaneously, by dynamically adjusting weights and fusion strategies, the system can better adapt to various complex and changing environmental conditions, thereby maintaining higher stability. Real-time processing and fusion of data from multiple inertial navigation systems is achieved without manual intervention or post-processing, improving the efficiency and response speed of real-time guide rail monitoring. Attached Figure Description

[0047] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram of the overall process logic of the dynamic fusion method for inertial navigation data according to the first embodiment of the present invention. Detailed Implementation

[0049] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0050] Example 1

[0051] Reference Figure 1 As an embodiment of the present invention, a dynamic fusion method for inertial navigation data is provided, comprising:

[0052] S100: The original inertial navigation data at different positions of the conductor is acquired in real time through the inertial navigation system, and the first data is obtained after preprocessing;

[0053] S200: Perform time synchronization processing on the inertial navigation system data at different locations in the first data to obtain time-synchronized inertial navigation data;

[0054] S300: Based on the accuracy and stability of historical data of inertial navigation systems, dynamic weights are assigned to each inertial navigation system;

[0055] S400: Predicts the state of the next time step using linear dynamic system equations. When new time-synchronized inertial navigation data arrives, it updates the state estimate and error covariance matrix using Kalman gain, dynamic weights, and the new time-synchronized inertial navigation data to output a fused dataset processed by a Kalman filter.

[0056] It should be noted that the inertial navigation system (INS) is the same as the inertial navigation system. Step S100, which acquires data in real time, ensures the monitoring system always has the latest information and can quickly respond to any changes in the traverse state. Acquiring data from multiple locations provides a comprehensive view of the traverse, helping to detect local anomalies or overall trends. Preprocessing in S200 helps remove noise, outliers, or invalid data from the original data, thereby improving data quality. Preprocessing also unifies data from different formats or units. In S300, dynamic weights are assigned to filter systems with better historical performance, allowing the dynamic weights to be adjusted according to the real-time performance of each INS, improving the adaptability of the fusion method. In S400, the Kalman filter combines previous estimates with new observation data to provide more accurate state estimates, achieving a cycle of prediction and correction. By continuously updating the state estimate and error covariance matrix, the Kalman filter effectively reduces errors in the data fusion process. The Kalman filter is robust to noise and incomplete observations, making the fused data more reliable.

[0057] In this embodiment of the application, the step S100 above, which involves acquiring raw inertial navigation data at different positions of the conductor in real time using an inertial navigation system, includes the following steps A1-A3:

[0058] A1: Set the optimization objectives and constraints for the inertial navigation system;

[0059] A2: Based on the optimization objective and constraints, the particle swarm optimization algorithm is used to search for the optimal solution, evaluate the objective function values ​​of different system layout schemes, and determine the final layout position of the system based on the optimal solution.

[0060] A3: Based on the final layout, install inertial navigation systems at the corresponding positions of the conductors to obtain real-time data at each position.

[0061] It should be noted that step A1 is mainly used to determine monitoring needs;

[0062] In one implementation, the optimization objective could be whether to maximize the overall monitoring coverage of the conductor.

[0063] In another implementation, the optimization objective may also be to minimize the monitoring blind spot or to optimize data accuracy, etc.

[0064] For example, a specific quantitative indicator can be set, such as a monitoring coverage rate of over 95% or a data error rate controlled within 5%.

[0065] In one implementation, constraints can be used to determine the physical limitations of the inertial navigation system, such as maximum working distance and signal reception range, and to determine the feasibility of the installation location. For example, some areas may not be suitable for installation due to environmental factors (such as terrain, buildings, etc.).

[0066] In another implementation, the constraints may also be to determine the cost budget, including equipment costs, installation costs, maintenance costs, etc., and to set minimum and maximum spacing between devices to avoid signal interference and ensure full coverage.

[0067] It should be noted that step A2 uses the particle swarm optimization algorithm, where each particle represents a possible arrangement of the inertial navigation system. The position and velocity of the particles are randomly initialized. An objective function is defined to quantitatively evaluate the merits of each arrangement. This function is used to calculate the fitness value of each particle, which is the objective function value corresponding to each arrangement.

[0068] Specifically, let f(x) be the objective function, where x represents a particle, i.e., a specific arrangement scheme of the inertial navigation system. Each arrangement scheme consists of a series of position coordinates, denoted as x = [x1, x2, ..., x...]. n ], where n is the number of inertial navigation systems, x i It is the position of the i-th inertial navigation system on the guide wire.

[0069] The objective function can be defined as:

[0070]

[0071] Here, x represents a particle, which is a specific arrangement scheme of the inertial navigation system. This scheme consists of a series of position coordinates, denoted as x = [x1, x2, ..., x...]. n ], where n is the number of inertial navigation systems, x i Let be the position of the i-th inertial navigation system on the guide wire, α1 be a positive number representing the importance weight of coverage in the objective function, and β1 be a positive number representing the importance weight of cost in the objective function. C represents the total length of the coverage wires of all inertial navigation systems. i Let be the length of the traverse line covered by the i-th inertial navigation system, and L be the total length of the traverse line. C is used to calculate the coverage ratio, which is obtained by dividing the sum of the traverse line lengths covered by all inertial navigation systems by the total length of the traverse line. d k represents the fixed cost of a single device.inl and k m These are proportionality constants, used to convert location characteristics (distance from the starting point) into installation and maintenance costs, respectively. Specifically, k inl It is a proportionality constant of installation costs, k m It is a proportionality constant of maintenance costs, |x i -x st | represents the position x of the i-th inertial navigation system. i With the starting point x of the conductor st The distance is used to determine the particle's position. In each iteration, the particle's position is updated based on its current position and velocity. The updated particle's fitness value is recalculated, and the particle's velocity and position are adjusted based on the fitness value and the individual's and the group's historical best positions. The number of iterations is set, and iteration stops when the termination condition is met, outputting the current optimal arrangement.

[0072] It should be noted that, in the specific operation of step A3, an on-site survey is required to confirm the feasibility of the installation location, check whether there is sufficient space, suitable power supply and network connection, etc., and install the inertial navigation system at the corresponding position of the wire according to the optimal layout scheme, ensuring that the equipment is installed firmly and in compliance with safety regulations. The installed inertial navigation system is debugged to ensure that it works normally, and tests are conducted to verify whether the system can accurately acquire real-time data and transmit it to the data center. The inertial navigation system is regularly maintained and inspected to ensure its long-term stable operation. A monitoring system is set up to monitor the working status and data quality of the equipment in real time, and to promptly identify and deal with potential problems.

[0073] For example, steps A1-A3 are applied to installing five inertial navigation systems on a 10-kilometer power line, with the goal of maximizing the overall monitoring coverage of the power line.

[0074] Step A1: Set the optimization goal to maximize the overall monitoring coverage of the five inertial navigation systems. Considering budget and technical limitations, set the minimum installation distance between each inertial navigation system to 1 kilometer and the maximum installation distance to 3 kilometers.

[0075] Step A2: Use the particle swarm optimization algorithm to search for the optimal placement of the inertial navigation systems that satisfy these conditions. After several iterations, the algorithm may find a solution. For example, installing inertial navigation systems at 1 km, 3.5 km, 5.5 km, 7.5 km, and 9.5 km along the guide wire.

[0076] Step A3: Based on this optimal solution, install inertial navigation systems at the above-mentioned locations along the conductor, ensuring that the monitoring range of each system can cover the entire conductor to the maximum extent while satisfying all constraints.

[0077] It should be noted that the inertial navigation system (INS) layout determined by the optimization algorithm in steps A1-A3 ensures that the entire guideway is effectively monitored, reducing or eliminating blind spots. A reasonable layout reduces signal interference and errors, thereby improving the accuracy and reliability of the data collected from the INS. While meeting monitoring requirements, budget and technical feasibility were also considered, ensuring the practicality and economy of the solution.

[0078] In this embodiment of the application, the preprocessing in step S100 above includes the following steps B1-B5:

[0079] B1: Set the window size of the moving average filter to N;

[0080] B2: Initialize a buffer of size N to store the N most recent data points;

[0081] B3: When a new data point arrives, add the new data point to the end of the buffer;

[0082] B4: If the buffer is full, remove the first element of the buffer so that the buffer always contains the latest N data points;

[0083] B5: Calculate the moving average of the latest N data points contained in the buffer to obtain the first data.

[0084] It should be noted that step B1 determines the number of data points the moving average filter will consider. The window size N determines how many consecutive data points the filter will use to calculate the average. A larger N value will result in a better smoothing effect on the data, but a slower response time; a smaller N value has the opposite effect. Step B2 uses a buffer to store the most recent N data points, which will be used to calculate the moving average. The buffer can be an array, list, or other data structure. In step B3, whenever a new data point arrives, it is added to the end of the buffer. If the buffer is full (i.e., it already contains N data points), the first element (the oldest data point) is removed to ensure that the buffer always contains the most recent N data points. Step B4 sums all the data points in the buffer and then divides by the number of data points (i.e., N) to obtain the moving average, which reflects the central trend of the most recent N data points. Step B5 uses the calculated moving average as a preprocessing result for the current data points, which can be used for subsequent data analysis, signal processing, or control system input.

[0085] For example, taking the design of an inertial navigation system installation scheme along a 10-kilometer power line as an example, in order to optimize the installation location and reduce the impact of noise data on decision-making, it is necessary to smooth the environmental data along the line to obtain more accurate installation point recommendations. Step B1: Select a window size N of 5, meaning that the moving average of relevant parameters (such as wind speed, vibration frequency, etc.) will be calculated based on the 5 most recent data points. Step B2: Initialize an array of size 5 as a buffer to store the 5 most recent environmental data points. Step B3: Whenever a new data point is acquired from the inertial navigation system or environmental monitoring equipment, it is added to the end of the buffer, and the oldest data point is removed (if the buffer is full). Step B4: Sum the 5 data points in the buffer and then divide by 5 to obtain the moving average of these environmental parameters. Step B5: Use this moving average as the representative value of the environmental parameters at the current point for analysis and determination of the optimal inertial navigation system installation location to ensure the stability and safety of the power line. In this way, the environmental conditions of each potential installation point can be evaluated more accurately, avoiding misjudgments due to transient noise or abnormal data. Moving average processing helps make decisions based on smoother and more stable datasets, thereby improving the rationality and effectiveness of inertial navigation system installation.

[0086] It should be noted that steps B1-B5, by calculating the average of multiple data points, can effectively reduce the impact of random noise on the data, making the data smoother. Moving averages generally reflect the true trend of the data better than a single data point, thus improving the accuracy of subsequent analysis and decision-making. Because the average is calculated based on multiple data points, the impact of a single outlier on the results is reduced.

[0087] In this embodiment of the application, the time synchronization processing in step S200 includes the following steps:

[0088] C1: Creates an empty list of time synchronization data and specifies the unit of synchronization;

[0089] C2: For each dataset, create a mapping table from timestamps to data values;

[0090] C3: For each time point t within the common time range, if time point t exists in the data mapping of the dataset, add the data value to the position of the corresponding time point in the time synchronization data list; if it does not exist, use linear interpolation to estimate the data value of the time point, add it to the time synchronization data list, and output the time synchronization inertial navigation data.

[0091] It should be noted that step C1 prepares an empty data structure to store the time-synchronized data. Simultaneously, the unit of time synchronization needs to be determined, such as seconds or milliseconds, depending on the data accuracy requirements and acquisition frequency. Step C2, since different datasets may have different sampling frequencies and timestamps, requires creating a mapping table for each dataset to associate timestamps with corresponding data values. This facilitates subsequent lookup and interpolation when synchronizing data by time. Step C3, for each time point t within the common time range of all datasets, performs the following operations: First, check if time point t exists in the timestamp mapping table of each dataset. If time point t exists in the timestamp mapping table of the dataset, directly extract the corresponding data value. If time point t does not exist, estimate the data value at that time point using linear interpolation. Linear interpolation calculates the estimated value at time t based on the two closest actual data points to time point t. Add the extracted or interpolated data value to the corresponding time point position in the time-synchronized data list. Output the time-synchronized inertial navigation data, which now has the same timestamp and can be directly used for comparison and analysis.

[0092] For example, consider five inertial navigation systems installed on a 10-kilometer power line; each system records data at a different frequency, and their timestamps may not be perfectly aligned. The goal is to synchronize this data to accurately analyze the dynamic behavior of the entire power line. Step C1: Create an empty time-synchronized data list and decide to use milliseconds as the unit of time synchronization. Step C2: For the datasets collected from each inertial navigation system, create a mapping table from timestamps to data values ​​(such as acceleration, angular velocity, etc.). Step C3: Determine a common time range, such as from a specific start time to an end time. For each millisecond within this range, check the data mapping table for each inertial navigation system. If a system has data recorded at that time point, extract that data; otherwise, use linear interpolation of the system's two most recent data points to estimate the data value at that time point. Finally, integrate the data from all systems into a single time-synchronized data list.

[0093] It should be noted that steps C1-C3 now align all data to the same timestamp, facilitating cross-system comparisons and analysis. Linear interpolation ensures a reasonable estimate even at points in time when some systems do not directly record data, thus maintaining data continuity and integrity. Time-synchronized data can more accurately reflect the actual dynamic behavior of power lines.

[0094] In this embodiment of the application, the dynamic weight in step S300 above is represented as:

[0095]

[0096] Among them, w i (t-1) represents the weight of the i-th inertial navigation system at the previous time t-1; β∈(0,1) represents the forgetting factor; α∈(0,1) represents the adjustment coefficient; This represents the accuracy evaluation index of the i-th inertial navigation system at time t; This is the stability evaluation index of the i-th inertial navigation system at time t, where N is a constant representing the number of data points used to calculate the accuracy and stability evaluation index, t represents the time index, i represents the current time, i represents the index of the inertial navigation system, j represents the index of the data point, and p... ij p represents the predicted value of the i-th inertial navigation system for the j-th data point. ij (t) represents the prediction value of the i-th inertial navigation system for the j-th data point at time t, r ij r represents the actual observation value of the j-th data point for the i-th inertial navigation system. ij (t) represents the actual observed value at time t; ij This represents the output value of the i-th inertial navigation system when processing the j-th data point, o ij (t) represents the output value of the inertial navigation system when it processes the j-th data point at time t; It is the average value of all output values ​​of the inertial navigation system at time t.

[0097] It should be noted that by setting dynamic weights, this method can adjust the weights of each inertial navigation system in real time based on its performance and accuracy. This means that a system that performs better at a given moment will receive a higher weight, thereby improving the overall accuracy and reliability of the data.

[0098] If an inertial navigation system malfunctions or its performance degrades, its weights are automatically reduced, thus minimizing the impact on overall data accuracy. This automatic adjustment capability enhances the system's fault tolerance. This method considers both accuracy and stability as evaluation metrics, helping to find a balance between the two. This not only provides accurate data but also ensures data stability, avoiding overall data distortion caused by abnormal fluctuations in individual data points. By comprehensively considering the data and performance of multiple inertial navigation systems, data from different systems can be more effectively integrated; the settings of the forgetting factor and adjustment coefficients can flexibly adapt to different application scenarios and needs. By adjusting these parameters, the speed and magnitude of weight updates can be controlled to adapt to different environments and data changes.

[0099] In this embodiment of the application, step S400 further includes initializing the relevant parameters of the Kalman filter to obtain the linear dynamic system equation. The relevant parameters include: initial state, initial error covariance, state transition matrix, observation matrix, process noise covariance matrix, and observation noise covariance matrix.

[0100] It should be noted that the initial state is usually an estimate, representing the most likely state of the system at the beginning. The parameters therein primarily serve the following purposes:

[0101] Initial error covariance: represents the uncertainty of the initial state estimation;

[0102] State transition matrix: describes how the system state evolves over time;

[0103] Observation matrix: maps the system state to the observation space;

[0104] Process noise covariance matrix: represents the uncertainty in the system's state transition process;

[0105] Observation noise covariance matrix: represents the uncertainty in the observation process.

[0106] The linear dynamic system equations predict the system state at the next time step based on the current state estimate and the state transition matrix. When new time-synchronized inertial navigation data arrives, the Kalman gain (a factor that weighs the prediction and observation) is used to combine the predicted state with the new observation data, thereby updating the state estimate. Simultaneously, the error covariance matrix also needs to be updated to reflect the uncertainty of the new state estimate.

[0107] For example, consider a 10-kilometer-long power line where five inertial navigation systems (INS1, INS2, INS3, INS4, INS5) are installed at equal intervals. Each system provides position, velocity, and acceleration data for the power line. These data vary due to environmental noise, sensor errors, and calibration discrepancies. The goal is to fuse these data using a Kalman filter to obtain a more accurate state estimate.

[0108] Let the initial position of the conductor be x0, where x0 = 0 is the starting point of the conductor, the initial velocity be v0 = 0, and the initial acceleration be a0 = 0. Combine these values ​​into an initial state vector. T represents the matrix transpose.

[0109] Define an initial error covariance matrix P0 to represent the uncertainty in the initial state estimation. For example, a diagonal matrix can be defined, where the elements on the diagonal represent the initial estimated variances of position, velocity, and acceleration, respectively.

[0110] Parameter settings, specifically including:

[0111] The state transition matrix F describes how the conductor's state evolves over time. For example, for a simple linear dynamic model, the state transition matrix might be a 3×3 matrix that considers the time evolution of position, velocity, and acceleration. Where Δt is the sampling time interval.

[0112] The observation matrix H maps the system state to the observation space. For each inertial navigation system, the observation matrix may be a 3×3 identity matrix because we directly observe position, velocity, and acceleration.

[0113] The process noise covariance matrix Q represents the uncertainty in the state transition process. This matrix is ​​typically set based on the system dynamics and noise characteristics.

[0114] The observation noise covariance matrix R represents the uncertainty in the observation process. For each inertial navigation system, this matrix can be set according to the accuracy of its sensors.

[0115] At each time step, the state transition matrix F and the current state estimate X are used. K-1 To predict the state at the next time step, the error covariance matrix is ​​also predicted. When new observation data is provided by the five inertial navigation systems, the Kalman gain is first calculated, then the state is updated using the observation data, and finally, the error covariance matrix is ​​updated. For the data from the five inertial navigation systems, the above Kalman filtering process can be performed separately, or an extended observation matrix and observation noise covariance matrix can be used to consider the data from all systems simultaneously. The latter requires more complex calculations but can provide a more accurate state estimate.

[0116] It should be noted that, through the above parameter settings and fusion calculations, data from multiple inertial navigation systems can be fused to obtain a more accurate and reliable traverse dynamic state estimate. The Kalman filter can fuse data from multiple inertial navigation systems to provide a more precise and reliable traverse dynamic state estimate. The Kalman filter can process new observation data in real time and update the state estimate immediately. By appropriately setting the process noise and observation noise covariance matrices, the Kalman filter can effectively suppress noise and errors in the data, improving the accuracy of the state estimate.

[0117] In this embodiment of the application, after step S400 is performed, steps D1-D2 are further included:

[0118] D1: Real-time monitoring of the fused dataset to determine the operating status of the conductor based on preset thresholds;

[0119] D2: If the monitored data exceeds the preset threshold, an alarm will be automatically triggered, and a potential abnormality in the conductor status will be indicated.

[0120] It should be noted that steps D1-D2, through real-time monitoring of the fused dataset, allow power system operators to instantly understand the dynamic status of conductors. This helps identify potential problems, enabling preventative maintenance and reducing unexpected power outages and repair costs. When conductor anomalies occur, such as excessive vibration, displacement, or other conditions that could lead to damage, the system can quickly detect them. By setting preset thresholds, the system can detect and address problems before they become severe. Automatic alarm triggering mechanisms can quickly notify maintenance personnel of abnormal conductor conditions, thereby shortening fault response time, reducing the impact of faults on the power system, and improving overall system reliability. Real-time monitoring and alarm systems reduce the frequency of manual inspections and avoid potential risks arising from the failure to detect and address problems in a timely manner.

[0121] The above is an illustrative scheme of a dynamic fusion method for inertial navigation data according to this embodiment. It should be noted that the technical solution of this dynamic fusion system for inertial navigation data belongs to the same concept as the technical solution of the dynamic fusion method for inertial navigation data described above. Details not described in detail in the technical solution of the dynamic fusion system for inertial navigation data in this embodiment can be found in the description of the technical solution of the dynamic fusion method for inertial navigation data described above.

[0122] The dynamic fusion system for inertial navigation data in this embodiment includes:

[0123] The acquisition module is used to acquire raw inertial navigation data at different positions of the conductor in real time through the inertial navigation system, and obtain the first data after preprocessing;

[0124] The synchronization module is used to perform time synchronization processing on the inertial navigation system data at different locations in the first data to obtain time-synchronized inertial navigation data;

[0125] The allocation module is used to assign dynamic weights to each inertial navigation system based on the accuracy and stability of historical data.

[0126] The fusion module is used to predict the state of the next time step through linear dynamic system equations. When new time-synchronized inertial navigation data arrives, it updates the state estimate and error covariance matrix using Kalman gain, dynamic weights, and the new time-synchronized inertial navigation data to output a fused dataset processed by a Kalman filter.

[0127] This embodiment also provides a computing device suitable for dynamic fusion of inertial navigation data, including:

[0128] The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the dynamic fusion method for inertial navigation data as proposed in the above embodiments.

[0129] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the dynamic fusion method for inertial navigation data as proposed in the above embodiments.

[0130] The storage medium proposed in this embodiment and the method for dynamic fusion of inertial navigation data proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0131] Based on the above description of the implementation methods, those skilled in the art will clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0132] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A dynamic fusion method of inertial navigation data, characterized in that, The method comprises the following steps: Real-time acquisition of original inertial navigation data of different positions of the guide line through an inertial navigation system, and pre-processing to obtain first data; Time synchronization processing of the inertial navigation system data at different positions in the first data is performed to obtain time-synchronized inertial navigation data; The time synchronization processing comprises: Creating an empty time-synchronized data list and determining the unit of synchronization; For each data set, a timestamp-to-data value mapping table is created; For each time point t within the common time range, if the time point t exists in the data mapping of the data set, the data value is added to the position corresponding to the time point in the time-synchronized data list; if not, the data value of the time point is estimated using linear interpolation, and is added to the time-synchronized data list, and the time-synchronized inertial navigation data is outputted; Based on the accuracy and stability of the historical data of the inertial navigation system, a dynamic weight is assigned to each inertial navigation system; The state of the next time step is predicted through a linear dynamic system equation, and when new time-synchronized inertial navigation data arrives, the state estimation and error covariance matrix are updated using the Kalman gain, the dynamic weight and the new time-synchronized inertial navigation data, to output the fusion data set processed by the Kalman filter; The dynamic weight is expressed as: where w i (t-1) represents the weight of the i-th inertial navigation system at the previous time t-1; β ∈ (0, 1) represents a forgetting factor; a ∈ (0, 1) represents an adjustment coefficient; represents the accuracy evaluation index of the i-th inertial navigation system at time t; is the stability evaluation index of the i-th inertial navigation system at time t, N represents a constant, representing the number of data points used to calculate the accuracy and stability evaluation indexes, t represents a time index, representing the current time, i represents the index of the inertial navigation system, j represents the index of the data point, p ij represents the predicted value of the i-th inertial navigation system for the j-th data point, p ij (t) represents the predicted value of the i-th inertial navigation system for the j-th data point at time t, r ij represents the actual observation value of the j-th data point for the i-th inertial navigation system, r ij (t) represents the actual observation value at time t; o ij represents the output value of the i-th inertial navigation system when processing the j-th data point, o ij (t) represents the output value of the i-th inertial navigation system when processing the j-th data point at time t; is the average of all output values of the i-th inertial navigation system at time t.

2. The dynamic fusion method of inertial navigation data according to claim 1, characterized in that, Real-time acquisition of original inertial navigation data of different positions of the guide line through an inertial navigation system, comprising: Setting the optimization target and constraint conditions of the inertial navigation system; According to the optimization target and constraint conditions, the optimal solution is searched through a particle swarm optimization algorithm, the objective function value of different arrangement schemes of the system is evaluated, and the final arrangement position of the system is determined according to the optimal solution; According to the final arrangement position, the inertial navigation system is installed at the corresponding position of the guide line to obtain real-time data at each position.

3. The dynamic fusion method of inertial navigation data according to claim 2, characterized in that, The pre-processing comprises setting the window size of the moving average filter to N; An N-sized buffer area is initialized to store the latest N data points; When a new data point arrives, the new data point is added to the end of the buffer area; If the buffer area is full, the first element of the buffer area is removed so that the latest N data points are always contained in the buffer area; The moving average of the latest N data points contained in the buffer area is calculated to obtain the first data.

4. The dynamic fusion method of inertial navigation data according to claim 3, characterized in that, Further comprising: Real-time monitoring of the fusion data set to determine the running state of the guide line through a preset threshold; If the monitored data exceeds the preset threshold, an alarm is automatically triggered, and a potential guide line state anomaly is prompted.

5. The dynamic fusion method of inertial navigation data according to claim 1, further comprising initializing the related parameters of the Kalman filter to obtain the linear dynamic system equation, wherein the related parameters comprise an initial state, an initial error covariance, a state transition matrix, an observation matrix, a process noise covariance matrix and an observation noise covariance matrix. The method comprises the following steps:

6. A system for applying a dynamic fusion method of inertial data as claimed in claim 1, characterized in that, An acquisition module is configured to acquire original inertial navigation data of different positions of the guide line through an inertial navigation system in real time, and to obtain first data after pre-processing; A synchronization module is configured to perform time synchronization processing on the inertial navigation system data at different positions in the first data to obtain time-synchronized inertial navigation data; A distribution module is configured to assign a dynamic weight to each inertial navigation system based on the accuracy and stability of the historical data of the inertial navigation system. ​ A fusion module is configured to predict the state of the next time step by a linear dynamic system equation, and when new time-synchronized inertial navigation data arrives, update the state estimate and error covariance matrix using Kalman gain, dynamic weight and new time-synchronized inertial navigation data, to output a fusion data set processed by a Kalman filter. 7.An electronic device, comprising: a memory and a processor; the memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions, and the computer executable instructions, when executed by the processor, implement the steps of the dynamic fusion method of inertial navigation data according to any one of claims 1 to 5. 8.A computer readable storage medium storing computer executable instructions, and the computer executable instructions, when executed by a processor, implement the steps of the dynamic fusion method of inertial navigation data according to any one of claims 1 to 5.

Images