Virtual-real synchronous inspection visualization method based on improved extended Kalman filtering

Abstract

The invention provides a virtual and real synchronous inspection visualization method based on improved extended Kalman filtering, and relates to the field of indoor and outdoor unmanned inspection scenes and digital twinning application. The method comprises the following steps: converting the motion characteristics of inspection equipment into a motion formula model, arranging inspection objects to form an inspection task, and constructing an inspection system; an improved extended Kalman filtering algorithm is adopted to predict the state of the inspection equipment at the next moment, an inspection task in the inspection scheme is executed through the operation execution process, system errors are corrected in a self-adaptive mode in the inspection operation execution process, and the accuracy of state prediction of the inspection equipment is improved; the twinborn object and target prediction result of the inspection equipment is visualized through a three-dimensional rendering mechanism, the inspection efficiency and controllability are improved, the inspection equipment is not limited by the environment, and no human intervention exists on site; and an intelligent inspection scheme can be adjusted according to different inspection task requirements, and a virtual-real combined synchronization process is completed, so that the requirements in different inspection scenes are met.

Description

Technical Field

[0001] The present invention relates to indoor and outdoor unmanned inspection scenarios and digital twin application fields, and specifically to a virtual-real synchronous inspection visualization method based on an improved extended Kalman filter. Background Art

[0002] With the rapid development of smart manufacturing and industrial Internet of Things technology, the demand for real-time monitoring and remote inspection of factory equipment is becoming more and more urgent. Traditional manual inspection methods have problems such as low efficiency, safety hazards and high labor costs. Remote synchronous inspection technology has emerged, which can achieve real-time monitoring of equipment status and remote fault diagnosis through technologies such as sensors, communication networks and cloud computing, thereby improving the efficiency and reliability of equipment management.

[0003] The core of remote synchronous inspection technology is to establish a real-time communication link between equipment, control system and monitoring center, so that monitoring personnel can understand the operating status of equipment at any time and detect and handle abnormal situations in time. This technology involves multiple key links such as sensor acquisition, data transmission, fault diagnosis, etc., and needs to solve a series of technical problems such as data acquisition accuracy, network communication reliability, and fault prediction accuracy. Among them, solving the prediction problem of inspection equipment is particularly important in remote synchronization means.

[0004] Compared to predictions using neural networks and other means, the Kalman filter can provide a mathematically well-defined framework when dealing with system state estimation with nonlinear characteristics, and can take into account the noise and uncertainty of the system. The advantage of EKF is that it can effectively integrate data from different sensors, and can give accurate predictions and estimate the uncertainty of the state even when the information is partially incomplete. This method is particularly suitable for systems where the model is relatively clear and well understood, and in scenarios with relatively small amounts of data or limited computing resources, the operating efficiency of the Kalman filter is often better than that of the neural network model that requires large-scale data training. However, the Kalman filter can only make predictions in linear systems, and the inspection equipment dynamics model involved in the present invention is a nonlinear motion, resulting in low prediction accuracy. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes a novel virtual-real synchronous inspection visualization model, which effectively improves the inspection efficiency and controllability by improving the extended Kalman filter structure and the adaptive noise adjustment strategy. Specifically, the purpose of the present invention is to propose a virtual-real synchronous inspection visualization method based on the improved extended Kalman filter, including:

[0006] Step 1: Convert the motion law of the inspection equipment in the real inspection scene into a motion formula model;

[0007] Step 2: According to the motion formula model of each inspection device, formulate the state prediction equation required by the extended Kalman filter algorithm for each inspection device;

[0008] Step 3: According to the inspection plan, multiple inspection tasks and inspection point information resources are obtained, wherein the inspection plan is a plan for the inspection equipment to inspect the substation secondary equipment to be inspected; and the multiple inspection tasks are scheduled to obtain an inspection job, wherein the inspection job includes all the inspection tasks obtained according to the inspection plan, and all the inspection tasks are arranged according to priority in the inspection job;

[0009] Step 4: According to the state value of the inspection equipment at the current moment and the state prediction equation of the inspection equipment, the predicted value of the inspection equipment at the next moment and the predicted observation value of the inspection equipment at the next moment are calculated, the state value includes at least speed, acceleration, position, and turning rate, the predicted value includes speed, acceleration, position, and turning rate, the predicted observation value includes the position at the next moment, the speed at the next moment, and the turning rate at the next moment, according to the predicted value, the predicted observation value, the observation value measured by the measuring equipment and the state prediction equation, the process noise covariance matrix, the observation noise covariance matrix and the error covariance matrix are updated in real time, and the inspection operation is analyzed to construct a simulated inspection scene, and in the simulated inspection scene, the inspection operation is performed according to the predicted value;

[0010] Step 5: The execution results of the inspection operation are presented in real time in the simulated inspection scene through a three-dimensional rendering mechanism.

[0011] Optionally, the inspection task includes multiple inspection routes, and substation secondary equipment to be inspected on each inspection route, and the position of the substation secondary equipment on the inspection route, and the inspection equipment for inspecting each inspection route, and the initial position of the inspection equipment. The inspection point information resources include a three-dimensional model of the substation secondary equipment to be inspected, a three-dimensional model of the inspection equipment, and a state prediction equation of the inspection equipment.

[0012] Optionally, step 3 specifically includes:

[0013] Step 3.1: Load the inspection plan, extract the inspection plan, obtain multiple inspection tasks, and obtain inspection point information resources according to the inspection plan;

[0014] Step 3.2: Perform multi-level feedback queue scheduling on the multiple inspection tasks to obtain inspection jobs.

[0015] Optionally, step 4 specifically includes:

[0016] Step 4.1: Analyze the inspection operation and construct a simulated inspection scenario, specifically including: according to the substation secondary equipment to be inspected on each inspection line in the inspection task and the inspection equipment inspected on each inspection line, obtain the three-dimensional model corresponding to the substation secondary equipment to be inspected on each inspection line and the three-dimensional model corresponding to the inspection equipment inspected on each inspection line in the inspection point information resource; according to multiple inspection lines in the inspection task, the three-dimensional model corresponding to the substation secondary equipment to be inspected on each inspection line and the three-dimensional model corresponding to the inspection equipment inspected on each inspection line, the position of the substation secondary equipment on the inspection line and the initial position of the inspection equipment, construct a simulated inspection scenario, and configure the state prediction equation of the corresponding three-dimensional model for the inspection equipment;

[0017] Step 4.2: Obtain the state value of the inspection device at the current moment, calculate the predicted value of the inspection device at the next moment according to the state value of the inspection device at the current moment and the state prediction equation of the inspection device, and obtain the predicted observation value at the next moment according to the predicted value; and calculate the error covariance matrix, which represents the error distribution between the predicted value of the state of the actual inspection device at the next moment and the true value of the state of the actual inspection device at the next moment;

[0018] Step 4.3: Measure the observation value of the inspection device in the real inspection scene at the current moment by measuring the device, the observation value includes the position at the current moment, the speed at the current moment, and the turning rate at the current moment, and obtain the difference between the observation value and the predicted observation value to obtain new information; and based on the environmental noise disturbance in the real inspection scene, initialize and set a unit matrix as the process noise covariance matrix, and based on the noise characteristics of the measuring device, obtain the observation noise covariance matrix;

[0019] Step 4.3: Determine whether the new information is greater than a preset threshold. If the new information is greater than the preset threshold, it indicates that the new information is abnormal, and adjust the observation noise covariance matrix to reduce the trust in the new information.

[0020] Step 4.4: According to the new information, the process noise covariance matrix and the observation noise covariance matrix are corrected through the adaptive filtering formula, and the error covariance matrix is ​​updated according to the state prediction equation and observation value of the inspection equipment;

[0021] Step 4.5: In the simulated inspection scenario, the inspection operation is performed according to the predicted value to realize the inspection of the three-dimensional model of the substation secondary equipment to be inspected by the three-dimensional model of the inspection equipment in the simulated inspection scenario.

[0022] The beneficial effects of adopting the above technical solution are:

[0023] 1. The present invention has the advantage of visualization. In a three-dimensional virtual simulation environment, the real inspection environment is restored one-to-one, and the system noise parameters are continuously corrected through real-time data feedback to correct the estimated value of the inspection equipment state, so as to achieve the purpose of synchronizing virtual and real inspection equipment and improve inspection efficiency;

[0024] 2. The present invention has the advantage of being reusable. Without being restricted by the environment and without human intervention on site, the system performs inspection tasks according to the intelligent inspection control plan, and adaptively corrects system noise during the inspection operation, thereby improving the accuracy of the inspection equipment status prediction;

[0025] 3. The intelligent inspection plan can be adjusted according to the requirements of different inspection tasks to complete the synchronization process of virtual and real, so as to meet the needs of different inspection scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A flowchart of a virtual-real synchronous inspection visualization method based on an improved extended Kalman filter in an embodiment of the present invention;

[0027] Figure 2 A schematic diagram of a flow chart of a multi-level scheduling task strategy in an embodiment of the present invention;

[0028] Figure 3 A schematic flow chart of another virtual-real synchronous inspection visualization method based on improved extended Kalman filtering in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0030] In view of the problems existing in the prior art, the present invention provides a virtual-real synchronous inspection visualization method based on an improved extended Kalman filter, which can simulate the inspection operation and synchronize the inspection equipment status, and realize the visualization and digitization of the real inspection process in a three-dimensional scene. Figure 1 , which may include the following steps:

[0031] Step 1: Convert the motion law of the inspection equipment in the real inspection scene into a motion formula model;

[0032] Among them, the real inspection scenario includes inspection equipment and inspection objects. The inspection object package refers to the combination of substation inspection point information (inspection point is the secondary equipment of the substation); the motion law can be the uniform motion of the inspection equipment maintaining the same speed, then the motion formula model is the uniform speed formula of the inspection equipment, and the motion law can be the accelerated motion of the inspection equipment maintaining a uniform acceleration, then the motion formula model is the acceleration formula of the inspection equipment.

[0033] Step 2: According to the motion formula model of each inspection device, formulate the state prediction equation required by the extended Kalman filter algorithm for each inspection device;

[0034] Step 3: According to the inspection plan, multiple inspection tasks and inspection point information resources are obtained, wherein the inspection plan is a plan for the inspection equipment to inspect the substation secondary equipment to be inspected; and the multiple inspection tasks are scheduled to obtain an inspection job, wherein the inspection job includes all the inspection tasks obtained according to the inspection plan, and all the inspection tasks are arranged according to priority in the inspection job;

[0035] Among them, the inspection task includes multiple inspection lines, and the substation secondary equipment to be inspected on each inspection line, and the position of the substation secondary equipment on the inspection line, and the inspection equipment for inspection on each inspection line, and the initial position of the inspection equipment. For example, the inspection plan may be inspection line No. 1, the substation secondary equipment inspected is equipment 001 to equipment 003, the inspection equipment is equipment 101, and the initial position of the inspection equipment 101 is the position directly in front of equipment 001.

[0036] The inspection point information resources include a three-dimensional model of the substation secondary equipment to be inspected, a three-dimensional model of the inspection equipment, and a state prediction equation of the inspection equipment.

[0037] Specifically, this can be achieved through the following steps:

[0038] Step 3.1: Load the inspection plan, extract the inspection plan, obtain multiple inspection tasks, and obtain inspection point information resources according to the inspection plan;

[0039] Among them, obtaining inspection point information resources according to the inspection plan specifically includes: the present application pre-constructs an inspection information set, the inspection information set includes three-dimensional models of multiple substation secondary equipment and three-dimensional models of multiple inspection equipment, and state prediction equations of multiple inspection equipment. Based on this, the model of the substation secondary equipment and the model of the inspection equipment are obtained in the inspection plan. According to the model of the substation secondary equipment and the model of the inspection equipment, the corresponding three-dimensional model of the substation secondary equipment and the three-dimensional model of the inspection equipment are obtained in the inspection information set. Based on the model of the inspection equipment, the corresponding state prediction equation is obtained in the inspection information set. Therefore, the three-dimensional model of the substation secondary equipment, the three-dimensional model of the inspection equipment and the state prediction equation of the inspection equipment constitute the inspection point information resources.

[0040] Specific, combined Figure 2 , the inspection plan is extracted. During the specific implementation, the inspection plan can be digitized to obtain an inspection task set. The inspection task set includes multiple inspection tasks, for example, inspection task 1, inspection task 2, and inspection task N in the figure.

[0041] Step 3.2: Perform multi-level feedback queue scheduling on the multiple inspection tasks to obtain inspection jobs.

[0042] Specific, combined Figure 2 , a multi-level scheduling task strategy is performed on multiple inspection tasks in the inspection task set to obtain inspection operations, so that multiple inspection tasks are executed according to priority.

[0043] Among them, multi-level feedback queue scheduling is a scheduling algorithm. The multi-level feedback queue scheduling algorithm dynamically adjusts the priority according to the characteristics of the process through multiple priority queues and time slice mechanisms to improve system throughput, reduce average waiting time and improve response time, while avoiding starvation.

[0044] Step 4: According to the state value of the inspection equipment at the current moment and the state prediction equation of the inspection equipment, the predicted value of the inspection equipment at the next moment and the predicted observation value of the inspection equipment at the next moment are calculated, the state value includes at least speed, acceleration, position, and turning rate, the predicted value includes speed, acceleration, position, and turning rate, the predicted observation value includes the position at the next moment, the speed at the next moment, and the turning rate at the next moment, according to the predicted value, the predicted observation value, the observation value measured by the measuring equipment and the state prediction equation, the process noise covariance matrix, the observation noise covariance matrix and the error covariance matrix are updated in real time, and the inspection operation is analyzed to construct a simulated inspection scene, and in the simulated inspection scene, the inspection operation is performed according to the predicted value;

[0045] The specific steps may include:

[0046] Step 4.1: Constructing a simulation inspection scenario, specifically including: according to the substation secondary equipment to be inspected on each inspection line in the inspection task, and the inspection equipment inspected on each inspection line, obtaining the three-dimensional model corresponding to the substation secondary equipment to be inspected on each inspection line and the three-dimensional model corresponding to the inspection equipment inspected on each inspection line in the inspection point information resource; according to multiple inspection lines in the inspection task, the three-dimensional model corresponding to the substation secondary equipment to be inspected on each inspection line, and the three-dimensional model corresponding to the inspection equipment inspected on each inspection line, the position of the substation secondary equipment on the inspection line and the initial position of the inspection equipment, constructing a simulation inspection scenario, and configuring the state prediction equation of the corresponding three-dimensional model for the inspection equipment;

[0047] Step 4.2: Obtain the state value of the inspection device at the current moment, calculate the predicted value of the inspection device at the next moment according to the state value of the inspection device at the current moment and the state prediction equation of the inspection device, and obtain the predicted observation value at the next moment according to the predicted value; and calculate the error covariance matrix, which represents the error distribution between the predicted value of the state of the actual inspection device at the next moment and the true value of the state of the actual inspection device at the next moment;

[0048] Step 4.3: Measure the observation value of the inspection device in the real inspection scene at the current moment by measuring the device, the observation value includes the position at the current moment, the speed at the current moment, and the turning rate at the current moment, and obtain the difference between the observation value and the predicted observation value to obtain new information; and based on the environmental noise disturbance in the real inspection scene, initialize and set a unit matrix as the process noise covariance matrix, and based on the noise characteristics of the measuring device, obtain the observation noise covariance matrix;

[0049] Among them, the measuring equipment may include an accelerometer, an odometer, a gyroscope and a linear radar, wherein the accelerometer is used to measure the acceleration in the observation value, the odometer is used to measure the speed in the observation value, the gyroscope is used to measure the turning rate in the observation value, and the linear radar is used to measure the position in the observation value.

[0050] Step 4.3: Determine whether the new information is greater than a preset threshold. If the new information is greater than the preset threshold, it indicates that the new information is abnormal, and adjust the observation noise covariance matrix to reduce the trust in the new information.

[0051] Step 4.4: According to the new information, the process noise covariance matrix and the observation noise covariance matrix are corrected through the adaptive filtering formula, and the error covariance matrix is ​​updated according to the state prediction equation and observation value of the inspection equipment;

[0052] Step 4.5: In the simulated inspection scenario, the inspection operation is performed according to the predicted value to realize the inspection of the three-dimensional model of the substation secondary equipment to be inspected by the three-dimensional model of the inspection equipment in the simulated inspection scenario.

[0053] Combination Figure 3 , the above steps are illustrated by examples, and the specific steps include:

[0054] In the real inspection scene, the present invention measures acceleration by an accelerometer, measures speed by an odometer, measures turning rate by a gyroscope, and measures position by a linear radar to obtain an observation value, and combines the control command in the simulated inspection scene as the input of the system at time t, that is, the state value (speed, acceleration, position and turning rate) in the previous text, and combines the state prediction equation based on the extended Kalman filter (EKF) to calculate the predicted value of the inspection equipment at time t+1, that is, the speed, acceleration, position and turning rate of the inspection equipment at time t+1, and then extract the predicted observation value according to the predicted value, the predicted observation value includes the speed, position and turning rate of the inspection equipment at time t+1, and then calculate the difference between the predicted observation value and the observed value to obtain new information, and then judge whether the observation value is abnormal by the Mahalanobis distance. If the observation value is judged to be abnormal, the observation noise covariance matrix R is adjusted to reduce the trust in the abnormal observation value. A threshold γ is set. If the Mahalanobis distance exceeds this threshold, the observation noise covariance matrix R is increased, otherwise it is updated normally. According to the new information, the process noise covariance matrix Q and the observation noise covariance matrix R are adjusted in real time through the sage-Husa adaptive filter. According to the state prediction equation and the observation value of the inspection equipment, the error covariance matrix P is updated, and the best state estimate at time t+1 is recorded. The definition of the best estimate is: the prediction value plus the product of the Kalman gain (K) and the information, that is, the final prediction value of the inspection equipment at time t+1 obtained by the extended Kalman filter algorithm.

[0055] Among them, the process noise covariance matrix Q is obtained based on the environmental noise disturbance. It is initialized to a unit matrix, which describes the uncertainty caused by external interference in the process of predicting the state of the inspection equipment. For example, the speed of the inspection equipment may change due to factors such as traffic and wind speed. These changes are process noise, and the process noise covariance matrix Q is obtained. Q reflects the size of these uncertainties. If the road conditions change greatly, Q will be larger, and if the speed of the optional equipment is relatively stable, Q will be smaller.

[0056] Among them, the observation noise covariance matrix R is obtained based on the noise characteristics of the measurement equipment and obtained through data provided by the sensor manufacturer. It describes the uncertainty of the observation instrument. The accuracy index of the observation instrument can be directly converted into the elements of the R matrix. Assuming that the location of the inspection equipment is observed by GPS, since GPS is not perfect, it has certain observation errors. These errors are observation noise, and R reflects the size of these observation errors. If the GPS device is of good quality and the signal is strong, R will be smaller. If the GPS device is poor or the signal is unstable, R will be larger.

[0057] Among them, the error covariance matrix P is initialized based on the prior (uncertainty estimate of the initial state of the system), and then dynamically updated through the state prediction equation and observation values ​​to predict the uncertainty of the current state of the system. By continuously updating the P matrix, the Kalman filter can gradually reduce this uncertainty and thus more accurately estimate the state of the system.

[0058] Among them, the Extended Kalman Filter (Extended Kalman Filter, EKF) is improved based on the Kalman filter (KF), which solves the problem that the Kalman filter cannot be applied in a nonlinear Gaussian system. It solves the problem of nonlinearity through local linearity. The nonlinear prediction equation and observation equation are derived and linearized in the form of tangent replacement. In essence, the nonlinear function is Taylor expanded at the current estimated state, and only the linear term (i.e., the first-order approximation) is retained, thereby converting it into a linear model. The present invention uses the sage-Husa adaptive filter to enhance the prediction accuracy of the extended Kalman filter in a dynamic system by real-time estimation and adjustment of the process noise and observation noise covariance matrix, and improves the robustness and adaptability of the filter in practical applications. Through the above-mentioned adaptively improved extended Kalman filter, the prediction effect is optimized, but the observed values ​​in the project are often affected by factors such as the environment and have large deviations. Therefore, in the above-mentioned update step, the Mahalanobis distance is used to judge whether the observed value is abnormal. If the observed value is judged to be abnormal, the noise covariance matrix is ​​adjusted to reduce the trust in the abnormal observed value.

[0059] Step 5: The execution results of the inspection operation are presented in real time in the simulated inspection scene through a three-dimensional rendering mechanism.

[0060] Specifically, a two-way information transmission communication is established through Socket, with the unity program as the client and the python project as the server, and the predicted value is passed into the simulation inspection scene. Specifically, the speed, turning rate and position of the inspection equipment act on the CTRV motion model, and the position of the inspection equipment in the simulation inspection scene is updated, so that the three-dimensional model of the inspection equipment can be inspected by the three-dimensional model of the substation secondary equipment in the simulation inspection scene. Figure 3, and also includes judging whether the time t+1 is greater than or equal to the final time. If so, the process ends. If not, t=t+1 is set to return and re-execute to achieve real-time update. Among them, the best state estimation value at time t+1 is recorded, Socket transmits data, unity provides an interface, and the speed, turning rate, and position of the inspection equipment act on the CTRV motion model, and the inspection equipment position in the simulated inspection scene is updated as the system prediction data acting on the three-dimensional scene simulation process.

[0061] Among them, the improved extended Kalman filter algorithm is calculated in the Python environment, and the calculation results are transmitted to the Unity 3D simulation virtual scene) to update the motion parameters of the inspection equipment in the virtual scene and the position information in the inspection scene, so as to achieve the purpose of patrol synchronization.

[0062] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) to form a technical solution.

Claims

1. A virtual and real synchronous inspection visualization method based on improved extended Kalman filter, It is characterized in that include: Step 1: Convert the motion law of the inspection equipment in the real inspection scene into a motion formula model; Step 2: According to the motion formula model of each inspection device, formulate the state prediction equation required by the extended Kalman filter algorithm for each inspection device; Step 3: According to the inspection plan, multiple inspection tasks and inspection point information resources are obtained, wherein the inspection plan is a plan for the inspection equipment to inspect the substation secondary equipment to be inspected; and the multiple inspection tasks are scheduled to obtain an inspection job, wherein the inspection job includes all the inspection tasks obtained according to the inspection plan, and all the inspection tasks are arranged according to priority in the inspection job; Step 4: According to the state value of the inspection equipment at the current moment and the state prediction equation of the inspection equipment, the predicted value of the inspection equipment at the next moment and the predicted observation value of the inspection equipment at the next moment are calculated, the state value includes at least speed, acceleration, position, and turning rate, the predicted value includes speed, acceleration, position, and turning rate, the predicted observation value includes the position at the next moment, the speed at the next moment, and the turning rate at the next moment, according to the predicted value, the predicted observation value, the observation value measured by the measuring equipment and the state prediction equation, the process noise covariance matrix, the observation noise covariance matrix and the error covariance matrix are updated in real time, and the inspection operation is analyzed to construct a simulated inspection scene, and in the simulated inspection scene, the inspection operation is performed according to the predicted value; Step 5: The execution results of the inspection operation are presented in real time in the simulated inspection scene through a three-dimensional rendering mechanism.

2. According to claim 1, the virtual-real synchronous inspection visualization method based on the improved extended Kalman filter, It is characterized in that The inspection task includes multiple inspection routes, substation secondary equipment to be inspected on each inspection route, the position of the substation secondary equipment on the inspection route, the inspection equipment for inspection on each inspection route, and the initial position of the inspection equipment. The inspection point information resources include a three-dimensional model of the substation secondary equipment to be inspected, a three-dimensional model of the inspection equipment, and a state prediction equation of the inspection equipment.

3. The virtual-real synchronous inspection visualization method based on improved extended Kalman filter according to claim 1, It is characterized in that Step 3 specifically includes: Step 3.1: Load the inspection plan, extract the inspection plan, obtain multiple inspection tasks, and obtain inspection point information resources according to the inspection plan; Step 3.2: Perform multi-level feedback queue scheduling on the multiple inspection tasks to obtain inspection jobs.

4. The virtual-real synchronous inspection visualization method based on the improved extended Kalman filter according to claim 2, It is characterized in that Step 4 specifically includes: Step 4.1: Analyze the inspection operation and construct a simulated inspection scenario, specifically including: according to the substation secondary equipment to be inspected on each inspection line in the inspection task and the inspection equipment inspected on each inspection line, obtain the three-dimensional model corresponding to the substation secondary equipment to be inspected on each inspection line and the three-dimensional model corresponding to the inspection equipment inspected on each inspection line in the inspection point information resource; according to multiple inspection lines in the inspection task, the three-dimensional model corresponding to the substation secondary equipment to be inspected on each inspection line and the three-dimensional model corresponding to the inspection equipment inspected on each inspection line, the position of the substation secondary equipment on the inspection line and the initial position of the inspection equipment, construct a simulated inspection scenario, and configure the state prediction equation of the corresponding three-dimensional model for the inspection equipment; Step 4.2: Obtain the state value of the inspection device at the current moment, calculate the predicted value of the inspection device at the next moment according to the state value of the inspection device at the current moment and the state prediction equation of the inspection device, and obtain the predicted observation value at the next moment according to the predicted value; and calculate the error covariance matrix, which represents the error distribution between the predicted value of the state of the actual inspection device at the next moment and the true value of the state of the actual inspection device at the next moment; Step 4.3: Measure the observation value of the inspection device in the real inspection scene at the current moment by measuring the device, the observation value includes the position at the current moment, the speed at the current moment, and the turning rate at the current moment, and obtain the difference between the observation value and the predicted observation value to obtain new information; and based on the environmental noise disturbance in the real inspection scene, initialize and set a unit matrix as the process noise covariance matrix, and based on the noise characteristics of the measuring device, obtain the observation noise covariance matrix; Step 4.3: Determine whether the new information is greater than a preset threshold. If the new information is greater than the preset threshold, it indicates that the new information is abnormal, and adjust the observation noise covariance matrix to reduce the trust in the new information. Step 4.4: According to the new information, the process noise covariance matrix and the observation noise covariance matrix are corrected through the adaptive filtering formula, and the error covariance matrix is ​​updated according to the state prediction equation and observation value of the inspection equipment; Step 4.5: In the simulated inspection scenario, the inspection operation is performed according to the predicted value to realize the inspection of the three-dimensional model of the substation secondary equipment to be inspected by the three-dimensional model of the inspection equipment in the simulated inspection scenario.

Images