Multi-source verification method and device for switching operation of electrical switch cabinet robot

Abstract

The invention provides an electrical switch cabinet robot switching operation multi-source verification method and device, and the method comprises the steps: determining that a to-be-operated electrical cabinet is consistent with a background instruction through the triple verification of an Internet of Things identity tag, an electrical cabinet position coordinate and an image identifier; and fusing data three-dimensional reconstruction of an RGB camera, a depth camera and an infrared sensor, identifying operation target information by a deep learning algorithm, and comparing the operation target information with a digital monitoring system reading. A visual servo image, a six-dimensional force / torque sensor and other multi-source data are collected, and cross validation is carried out through a consistency judgment algorithm. And a recovery strategy is automatically executed in case of abnormal conditions such as jamming, slipping and tool falling, and a safety alarm is given if two times of recovery fails. Multi-level safety guarantee is constructed through physical sudden stop, software sudden stop and sensor over-limit triggering, and when communication is interrupted or main control fails, the equipment automatically returns to a safety position and keeps braking.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of robot operation control, and particularly relates to a robot switching operation multi-source verification method and device for an electrical switch cabinet. BACKGROUND

[0002] With the popularization of unattended substations, using robots to replace manual switch cabinet switching operation has become an industry development trend. However, existing robot systems still face severe reliability challenges in actual application.

[0003] Environmental perception is inaccurate, and is affected by factors such as light changes, device reflections, and dust shielding. Visual recognition is prone to misjudgment, leading to incorrect operation targets. During operation, the device position is inaccurate, and the operation process lacks real-time monitoring. Most systems only rely on pre-set trajectories for execution, and lack real-time feedback and regulation of key parameters such as operation torque and contact state, which can easily cause device damage or incomplete operation. The abnormal handling capability is weak, and when encountering abnormal situations such as jamming, excessive resistance, and tools not in place, the system often cannot identify and take correct measures in time, which may cause operation interruption or device damage. There is a lack of multi-source verification mechanism. After operation, only the robot itself is relied on for judgment, and there is a lack of state confirmation from the switch cabinet itself or the background system, which poses a risk of "false success". The fault tolerance and safety mechanism is insufficient, and the system lacks reliable degradation operation or safety shutdown strategies in the event of sensor failure or communication interruption. Therefore, there is an urgent need for a control method that can fully guarantee the reliable execution of switching operation, realizing the leap from "being able to operate" to "reliable operation". SUMMARY

[0004] The present application aims to at least partially solve one of the technical problems in the related art.

[0005] To this end, the first object of the present application is to propose a multi-source verification method for robot switching operation of an electrical switch cabinet.

[0006] The second object of the present application is to propose a multi-source verification device for robot switching operation of an electrical switch cabinet.

[0007] To achieve the above-mentioned objects, the first aspect of the present application proposes a multi-source verification method for robot switching operation of an electrical switch cabinet, comprising: S1, confirming the consistency of the electrical cabinet to be operated and the background instruction through three identification methods of Internet of Things identity tags, electrical cabinet position coordinates, and image identification; S2, performing three-dimensional reconstruction by fusing RGB camera, depth camera, and infrared sensor data, identifying the operation target type, position, and initial state using a deep learning algorithm, and comparing with the state reading of the digital monitoring system; S3 collects visual servo images, six-dimensional force / torque sensor data, encoder displacement data, and remote signaling data of switchgear protection and control devices. It then performs cross-validation of multi-source data within a preset time window using a consistency judgment algorithm. S4. When an anomaly is detected, a preset recovery strategy is automatically executed according to the anomaly type, which includes jamming, slippage, and tool falling off. If the recovery fails twice in a row, the system enters a safe state and reports an alarm. The S5 features a multi-level safety system with a physical emergency stop button, a software emergency stop function, and an automatic over-limit triggering mechanism for sensors. It automatically reverts to a safe position and maintains braking in the event of communication interruption or main control failure.

[0008] In one embodiment of the present invention, S1 includes: S11 reads the encrypted information of the IoT identity tag through radio frequency identification technology and performs hash verification with the device's unique identifier in the background database; S12, an image recognition algorithm based on feature point matching is used to compare the electrical cabinet image labels, and the number of feature points must meet the matching threshold.

[0009] In one embodiment of the present invention, S2 includes: S21 constructs a 3D feature matrix that fuses point clouds and textures through spatiotemporal synchronous acquisition by an RGB camera and a depth camera. S22, when the confidence level of the YOLOv8 algorithm is less than the preset minimum threshold, the multispectral supplementary lighting module is activated and the sampling frequency of the infrared sensor is adjusted.

[0010] In one embodiment of the present invention, S3 includes: S31: Construct an environmental obstacle probability map based on LiDAR point cloud data. When the obstacle probability is greater than the system's preset safety threshold, trigger path replanning. S32 calls up the historical operation torque distribution in the operation knowledge base and dynamically adjusts the current operation torque upper limit.

[0011] In one embodiment of the present invention, S4 includes: S41, when using the admittance control algorithm, set the end impedance parameter to enhance vertical compliance; S42, when the six-dimensional force sensor detects that the torque component exceeds the safety threshold and the duration is greater than the critical time, the self-locking mechanism at the end of the robotic arm is triggered and the power source is cut off.

[0012] To achieve the above objectives, a second aspect of the present invention provides a multi-source verification device for switching operations of an electrical switchgear robot, comprising: The triple authentication module is used to confirm the consistency between the electrical cabinet to be operated and the backend command through three methods: IoT identity tag, electrical cabinet location coordinates and image identification. The 3D data fusion and recognition module is used to fuse data from RGB cameras, depth cameras, and infrared sensors for 3D reconstruction. It uses deep learning algorithms to identify the type, location, and initial state of the target and compares it with the status readings of the digital monitoring system. The multi-source data cross-validation module is used to collect visual servo images, six-dimensional force / torque sensor data, encoder displacement data and remote signaling data of switchgear protection and control devices, and performs multi-source data cross-validation within a preset time window through a consistency judgment algorithm. The anomaly handling and recovery strategy module is used to automatically execute a preset recovery strategy according to the anomaly type when an anomaly is detected. The anomaly types include jamming, slipping, and tool falling off. If the recovery fails twice in a row, the system will enter a safe state and report an alarm. The multi-level safety protection module is used to build a multi-level safety protection through physical emergency stop buttons, software emergency stop function and sensor over-limit automatic triggering mechanism. It automatically reverts to the safe position and maintains braking when communication is interrupted or the main control fails.

[0013] This invention discloses a multi-source verification method and device for switching operations of an electrical switchgear robot. This method improves operational reliability by reducing operational misjudgment rates through multi-source sensing and cross-verification. It enhances safety by employing force-position hybrid control and a multi-level safety fallback mechanism to effectively prevent equipment damage and personal injury risks. The introduction of background remote signal verification eliminates "false success" phenomena and ensures auditable operation results. The system is robust, supports degraded operation, and adapts to complex field environments and occasional failures. Continuous self-optimization and a self-learning mechanism ensure that system reliability continuously improves over time.

[0014] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0015] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of a multi-source verification method for switching operation of an electrical switchgear robot according to an embodiment of the present invention; Figure 2 This is an architecture diagram of a multi-source verification method for switching operation of an electrical switchgear robot according to an embodiment of the present invention; Figure 3 This is a structural diagram of a multi-source verification device for switching operation of an electrical switchgear robot according to an embodiment of the present invention. Detailed Implementation

[0016] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0017] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. 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 scope of protection of the present invention.

[0018] The following description, with reference to the accompanying drawings, describes a multi-source verification method and apparatus for switching operations of an electrical switchgear robot according to an embodiment of the present invention.

[0019] Figure 1 This is a flowchart of a multi-source verification method for switching operation of an electrical switchgear robot according to an embodiment of the present invention, such as... Figure 1 As shown, it includes: S1 confirms the consistency between the electrical cabinet to be operated and the backend command through three methods: IoT identity tag, electrical cabinet location coordinates and image identification; S2 integrates data from RGB cameras, depth cameras, and infrared sensors for 3D reconstruction, uses deep learning algorithms to identify the type, location, and initial state of the target, and compares it with the status readings of the digital monitoring system. S3 collects visual servo images, six-dimensional force / torque sensor data, encoder displacement data, and remote signaling data of switchgear protection and control devices. It then performs cross-validation of multi-source data within a preset time window using a consistency judgment algorithm. S4. When an anomaly is detected, a preset recovery strategy is automatically executed according to the anomaly type, which includes jamming, slippage, and tool falling off. If the recovery fails twice in a row, the system enters a safe state and reports an alarm. The S5 features a multi-level safety system with a physical emergency stop button, a software emergency stop function, and an automatic over-limit triggering mechanism for sensors. It automatically reverts to a safe position and maintains braking in the event of communication interruption or main control failure.

[0020] This invention discloses a multi-source verification method for switching operations of an electrical switchgear robot. Through multi-source data fusion verification and consistency judgment mechanism, it effectively eliminates the risk of "false success" in switching operations, ensures that the operation results are true, reliable and auditable, and significantly improves the reliability and safety of the robot performing electrical switchgear operations.

[0021] The following describes in detail, with reference to the accompanying drawings, a multi-source verification method for switching operation of an electrical switchgear robot according to an embodiment of the present invention.

[0022] The purpose of this invention is to provide a reliable execution control method for switching operations of electrical switchgear robots, significantly improving the reliability and safety of robot switching operations. For example... Figure 2 As shown, it includes the following steps: S10. Secondary verification of operator identity for electrical cabinet: The encrypted information of the IoT identity tag is read by radio frequency identification technology and hashed and verified with the unique device identifier in the background database; the image identification of the electrical cabinet is compared by an image recognition algorithm based on feature point matching, and the number of feature points must meet the matching threshold.

[0023] Specifically, when the robot arrives at the location of the electrical cabinet to be operated, it uses three methods—the electrical cabinet's IoT identification tag, the electrical cabinet's location coordinates, and the electrical cabinet's image identification—to jointly confirm whether the electrical cabinet to be operated is accurately consistent with the electrical cabinet specified in the operation instructions issued by the backend, as well as the operation ticket and work order. If they are consistent, the operation continues; if they are inconsistent, the maintenance personnel are alerted to intervene.

[0024] S20. Operation target identification and content confirmation: By acquiring data spatiotemporally through RGB and depth cameras, a 3D feature matrix is ​​constructed that fuses point clouds and textures. When the confidence level of the YOLOv8 algorithm is less than the preset minimum threshold, the multispectral illumination module is activated and the sampling frequency of the infrared sensor is adjusted.

[0025] Specifically, the robot performs 3D reconstruction and feature recognition of the target switch cabinet operation panel by fusing data from RGB cameras, depth cameras, and infrared sensors. A deep learning-based target detection algorithm (such as YOLOv8+PointNet++) is used to identify the type, location, and initial state of the target, and a confidence assessment module determines the reliability of the recognition results. If the confidence level is below a threshold, supplementary lighting, rescanning, or manual verification is initiated. Simultaneously, the status readings are compared with those from the backend digital monitoring system. If they match, operation continues; otherwise, maintenance personnel are alerted for intervention.

[0026] S30, Operational Risk Assessment and Strategy Generation: An environmental obstacle probability map is constructed based on LiDAR point cloud data. When the obstacle probability exceeds the system's preset safety threshold, path replanning is triggered. The historical operation torque distribution in the operation knowledge base is called to dynamically adjust the current operation torque upper limit.

[0027] Specifically, the robot uses LiDAR combined with cameras to model the external environment. Within this modeled environment, the robot simulates and plans its robotic arm's movement path, ensuring there are no obstructions or objects that could affect operational safety. Simultaneously, it accesses a built-in "operation knowledge base" to obtain prior information such as standard procedures, permissible force ranges, and typical anomaly patterns, as well as legal and regulatory information, in accordance with the regulations of the relevant production authorities, to conduct a pre-assessment of operational risks. Furthermore, based on IoT tags and historical operation records stored in the background, it retrieves previous operation history for this equipment, including the range of force applied and any special operational precautions for equipment malfunctions. Finally, based on this information, it dynamically generates an operation strategy with low risk, high reliability, and safety, including the propulsion speed, turning speed and torque, and key insertion depth.

[0028] S40, Force-Position Hybrid Compliant Control Execution: When using the admittance control algorithm, the end impedance parameter is set to enhance vertical compliance; when the six-dimensional force sensor detects that the torque component exceeds the safety threshold and the duration is greater than the critical time, the self-locking mechanism at the end of the robotic arm is triggered and the power source is cut off.

[0029] Specifically, when the robot performs operations, it adopts impedance control or admittance control algorithms to achieve force-position hybrid control: during the approach phase, position control is the main method for fast and accurate positioning; during the contact phase, it switches to force / torque control, dynamically adjusting the action according to the preset force window to prevent overload; and it monitors the data of the six-dimensional force / torque sensor at the end of the robot arm in real time. If the data exceeds the safety threshold, it immediately pauses and alarms.

[0030] S50, Multi-source fusion verification of operational status: During and after the operation, the operation status is verified through cross-validation of multi-source information. By combining visual servo images, torque change curves, and encoder displacement data, it is determined whether the operation is in place. Remote signaling quantities such as the actual opening and closing signals of the circuit breaker and the knob position signals are queried from the switchgear protection and control device through the station's communication network. The operation is considered successful only when the local and remote verification results are consistent.

[0031] S60, Anomaly Detection and Adaptive Recovery: The switchgear's back-end operating system and the robot both have built-in anomaly detection models that analyze their respective sensor data streams in real time. If anomalies such as jamming, slippage, or tool detachment are detected, the operation is stopped immediately. Pre-set recovery strategies are automatically executed according to the anomaly type. For example, if there is slight jamming, a slight oscillation is attempted to loosen it; if the tool is not in place, it is reconnected. If two consecutive recovery attempts fail, the system enters a safe state and reports an alarm.

[0032] S70, Establish a multi-level security mechanism: During robot operation, the entire system is equipped with a multi-level safety mechanism, supporting remote braking intervention and safety assurance in various ways. The robot body and remote controller provide emergency stop buttons, and a software-based emergency stop function is available remotely, enabling physical power cut-off. The robot's end effector provides automatic stop operation based on sensor over-limit triggering; operation automatically stops when the feedback torque exceeds a set danger threshold. In the event of communication interruption or main control failure, the robotic arm automatically retracts to a safe position and maintains braking. A "remote control" degradation mode is supported, allowing manual intervention when fully automatic mode is unavailable.

[0033] S80, adaptive learning and operation optimization: The system records data for each operation, including recognition accuracy, success rate, operation feedback torque curve, and operation time. It continuously optimizes the recognition model and operation strategy using machine learning algorithms, as well as force control parameters, recognition models, and anomaly criteria, thereby improving long-term operational reliability and enabling the system to self-improve its performance.

[0034] To achieve the above embodiments, such as Figure 3 As shown, this embodiment also provides a multi-source verification device 10 for switching operation of electrical switchgear robot. The device 10 includes a triple identity verification module 100, a three-dimensional data fusion recognition module 200, a multi-source data cross-verification module 300, an anomaly handling and recovery strategy module 400, and a multi-level security module 500.

[0035] The triple authentication module 100 is used to confirm the consistency between the electrical cabinet to be operated and the backend command through three methods: IoT identity tag, electrical cabinet location coordinates and image identification. The 3D data fusion and recognition module 200 is used to fuse data from RGB cameras, depth cameras, and infrared sensors for 3D reconstruction. It uses deep learning algorithms to identify the type, location, and initial state of the target and compares it with the status readings of the digital monitoring system. The multi-source data cross-validation module 300 is used to collect visual servo images, six-dimensional force / torque sensor data, encoder displacement data and remote signaling of switch cabinet protection and control devices, and to perform multi-source data cross-validation within a preset time window through a consistency judgment algorithm. The anomaly handling and recovery strategy module 400 is used to automatically execute a preset recovery strategy according to the anomaly type when an anomaly is detected. The anomaly types include jamming, slipping, and tool falling off. If the recovery fails twice in a row, it enters a safe state and reports an alarm. The multi-level safety module 500 is used to build a multi-level safety system through a physical emergency stop button, a software emergency stop function, and an automatic triggering mechanism for sensor over-limit. It automatically reverts to a safe position and maintains braking when communication is interrupted or the main control fails.

[0036] Furthermore, the aforementioned triple authentication module 100 is also used for: The encrypted information of the IoT identity tag is read using radio frequency identification technology and hash-verified with the unique device identifier in the background database. An image recognition algorithm based on feature point matching is used to compare the identification marks of electrical cabinet images, and the number of feature points must meet the matching threshold.

[0037] Furthermore, the aforementioned three-dimensional data fusion and recognition module 200 is also used for: A 3D feature matrix fused with point cloud and texture is constructed by spatiotemporally acquiring data from an RGB camera and a depth camera. When the YOLOv8 algorithm detects a confidence level less than the preset minimum threshold, the multispectral illumination module is activated and the sampling frequency of the infrared sensor is adjusted.

[0038] Furthermore, the aforementioned multi-source data cross-validation module 300 is also used for: An environmental obstacle probability map is constructed based on LiDAR point cloud data. When the obstacle probability is greater than the system's preset safety threshold, path replanning is triggered. The historical operation torque distribution in the operation knowledge base is invoked to dynamically adjust the current operation torque upper limit.

[0039] Furthermore, the aforementioned exception handling and recovery strategy module 400 is also used for: When using the admittance control algorithm, the end impedance parameter is set to enhance vertical compliance; When the six-dimensional force sensor detects that the torque component exceeds the safety threshold and the duration is greater than the critical time, it triggers the self-locking mechanism at the end of the robotic arm and cuts off the power source.

[0040] This invention discloses a multi-source verification device for switching operations of an electrical switchgear robot. Through multi-source data fusion verification and consistency judgment mechanism, it effectively eliminates the risk of "false success" in switching operations, ensures that the operation results are true, reliable and auditable, and significantly improves the reliability and safety of the robot performing electrical switchgear operations.

[0041] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0042] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A multi-source verification method for switching operations of an electrical switchgear robot, characterized in that, include: S1 confirms the consistency between the electrical cabinet to be operated and the backend command through three methods: IoT identity tag, electrical cabinet location coordinates and image identification; S2 integrates data from RGB cameras, depth cameras, and infrared sensors for 3D reconstruction, uses deep learning algorithms to identify the type, location, and initial state of the target, and compares it with the status readings of the digital monitoring system. S3 collects visual servo images, six-dimensional force / torque sensor data, encoder displacement data, and remote signaling data of switchgear protection and control devices. It then performs cross-validation of multi-source data within a preset time window using a consistency judgment algorithm. S4. When an anomaly is detected, a preset recovery strategy is automatically executed according to the anomaly type, which includes jamming, slippage, and tool falling off. If the recovery fails twice in a row, the system enters a safe state and reports an alarm. The S5 features a multi-level safety system with a physical emergency stop button, a software emergency stop function, and an automatic over-limit triggering mechanism for sensors. It automatically reverts to a safe position and maintains braking in the event of communication interruption or main control failure.

2. The method as described in claim 1, characterized in that, S1 includes: S11 reads the encrypted information of the IoT identity tag through radio frequency identification technology and performs hash verification with the device's unique identifier in the background database; S12, an image recognition algorithm based on feature point matching is used to compare the electrical cabinet image labels, and the number of feature points must meet the matching threshold.

3. The method as described in claim 1, characterized in that, S2 includes: S21 constructs a 3D feature matrix that fuses point clouds and textures through spatiotemporal synchronous acquisition by an RGB camera and a depth camera. S22, when the confidence level of the YOLOv8 algorithm is less than the preset minimum threshold, the multispectral supplementary lighting module is activated and the sampling frequency of the infrared sensor is adjusted.

4. The method as described in claim 1, characterized in that, The S3 further includes: S31: Construct an environmental obstacle probability map based on LiDAR point cloud data. When the obstacle probability is greater than the system's preset safety threshold, trigger path replanning. S32 calls up the historical operation torque distribution in the operation knowledge base and dynamically adjusts the current operation torque upper limit.

5. The method as described in claim 1, characterized in that, The S4 includes: S41, when using the admittance control algorithm, set the end impedance parameter to enhance vertical compliance; S42, when the six-dimensional force sensor detects that the torque component exceeds the safety threshold and the duration is greater than the critical time, the self-locking mechanism at the end of the robotic arm is triggered and the power source is cut off.

6. A multi-source verification device for switching operation of an electrical switchgear robot, characterized in that, include: The triple authentication module is used to confirm the consistency between the electrical cabinet to be operated and the backend command through three methods: IoT identity tag, electrical cabinet location coordinates and image identification. The 3D data fusion and recognition module is used to fuse data from RGB cameras, depth cameras, and infrared sensors for 3D reconstruction. It uses deep learning algorithms to identify the type, location, and initial state of the target and compares it with the status readings of the digital monitoring system. The multi-source data cross-validation module is used to collect visual servo images, six-dimensional force / torque sensor data, encoder displacement data and remote signaling data of switchgear protection and control devices, and performs multi-source data cross-validation within a preset time window through a consistency judgment algorithm. The anomaly handling and recovery strategy module is used to automatically execute a preset recovery strategy according to the anomaly type when an anomaly is detected. The anomaly types include jamming, slipping, and tool falling off. If the recovery fails twice in a row, the system will enter a safe state and report an alarm. The multi-level safety protection module is used to build a multi-level safety protection through physical emergency stop buttons, software emergency stop function and sensor over-limit automatic triggering mechanism. It automatically reverts to the safe position and maintains braking when communication is interrupted or the main control fails.

7. The apparatus as claimed in claim 6, characterized in that, The triple authentication module is also used for: The encrypted information of the IoT identity tag is read using radio frequency identification technology and hash-verified with the unique device identifier in the background database. An image recognition algorithm based on feature point matching is used to compare the identification marks of electrical cabinet images, and the number of feature points must meet the matching threshold.

8. The apparatus as claimed in claim 6, characterized in that, The three-dimensional data fusion and recognition module is also used for: A 3D feature matrix fused with point cloud and texture is constructed by spatiotemporally acquiring data from an RGB camera and a depth camera. When the YOLOv8 algorithm detects a confidence level less than the preset minimum threshold, the multispectral illumination module is activated and the sampling frequency of the infrared sensor is adjusted.

9. The apparatus as claimed in claim 6, characterized in that, The multi-source data cross-validation module is also used for: An environmental obstacle probability map is constructed based on LiDAR point cloud data. When the obstacle probability is greater than the system's preset safety threshold, path replanning is triggered. The historical operation torque distribution in the operation knowledge base is invoked to dynamically adjust the current operation torque upper limit.

10. The apparatus as claimed in claim 6, characterized in that, The exception handling and recovery strategy module is also used for: When using the admittance control algorithm, the end impedance parameter is set to enhance vertical compliance; When the six-dimensional force sensor detects that the torque component exceeds the safety threshold and the duration is greater than the critical time, it triggers the self-locking mechanism at the end of the robotic arm and cuts off the power source.

Images