Automatic cruise system of transformer substation electric insulator cleaning robot

Abstract

The invention discloses an automatic cruise system for a cleaning robot of a power insulator of a transformer substation, and relates to the technical field of intelligent operation and maintenance of the transformer substation, the automatic cruise system comprises a data extraction module, a processing module, a decision module and an execution module, and solves the technical problems of excessive shutdown when a wind field approaches danger and discontinuous operation when the wind field is safe. The decision-making module sets safety indexes according to the weights of the wind speed and the wind direction, divides the wind speed into four risk intervals, divides the wind direction into three grades, calculates the total safety score of the wind field through weighted summation, corresponds to three operation grades of safety, yellow early warning and red early warning, generates differential instructions according to different grades, and sends the differential instructions to the wind field. And real-time path updating during early warning level switching is supported, and safe operation under the dynamic change of the wind field is realized.

Description

Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance technology for substations, specifically to an automatic cruise system for a substation power insulator cleaning robot. Background Technology

[0002] With the widespread application of robotics technology in the power industry, insulator cleaning robots are gradually becoming a core tool to replace manual labor. However, the working environment of outdoor substations is complex, especially with variable wind conditions. Factors such as equipment cooling airflow and building obstructions can easily create localized unstable wind fields. The water mist generated by the cleaning robot is easily affected by wind and can drift. If the water mist drifts toward live busbars, switches, or other equipment, it will significantly increase the probability of phase-to-phase flashover or ground discharge accidents, posing a serious threat to the safe operation of the substation.

[0003] Existing cleaning robots mostly adopt a preset path cruising mode. When the wind field is in a safe range, there is a lack of a mechanism to ensure the continuity of the operation. Problems such as data acquisition deviation and path planning delay can easily lead to the interruption of the operation process, making it impossible to make full use of the safe window period to complete the cleaning task. At the same time, the perception dimension of wind field parameters is limited and the data processing lacks a standardized process, making it difficult to accurately predict the risk of water mist drift. They cannot efficiently advance the operation when the wind field is safe, nor can they avoid risks through fine-tuning when the wind field is close to danger. Instead, they fall into the dual dilemma of excessive downtime and discontinuous operation, which seriously restricts the practicality and safety of robots in outdoor substation cleaning operations. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an automatic cruise system for cleaning power insulators in substations, which solves the problems of excessive shutdown when the wind farm is approaching danger and discontinuous operation when the wind farm is safe.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an automatic cruise system for a substation power insulator cleaning robot, comprising: a data extraction module for collecting multi-source data related to the cruise and operation of the insulator cleaning robot within the substation; Decision module: Based on wind field data collected from multiple sources, wind speed and wind direction safety index weights are allocated according to a preset ratio. At the same time, wind speed intervals and wind direction status divisions are set to quantify wind speed safety score and wind direction safety score. The wind speed safety score is divided into four intervals based on the average wind speed on the robot side: no risk, low risk, high risk, and extremely high risk. Different score values ​​are assigned to the four intervals. The average wind speed corresponds to different scores in different value intervals and is marked as the corresponding risk status. The wind direction safety score is divided into three levels based on the real-time wind direction status and the deviation from the safe wind direction, and different score values ​​are assigned to the three levels. The real-time wind direction is divided into different areas within the three levels. Finally, the safety index weights of wind speed and wind direction are respectively incorporated into the wind speed safety score and wind direction safety score to obtain the wind field safety score. Based on the wind field safety score, three safety levels are divided, and corresponding graded warning instructions and appropriate cruise path strategies are issued according to the corresponding safety level.

[0006] As a further aspect of the present invention: the weights of the wind speed and wind direction safety indicators are specifically as follows: The safety index for wind speed has a weighting of 60%, while the safety index for wind direction has a weighting of 40%. The wind speed safety score is as follows: an average wind speed of ≤2m / s is considered no risk and scores 10 points; 2m / s < average wind speed ≤3m / s is considered low risk and scores 8 points; 3m / s < average wind speed ≤4m / s is considered high risk and scores 4 points; and an average wind speed >4m / s is considered extremely high risk and scores 0 points. The terms "no risk," "low risk," "high risk," and "extremely high risk" correspond to four states: light wind, low wind, high wind, and strong wind, respectively. The wind direction safety score is as follows: if the real-time wind direction is within the safe wind direction, 10 points are awarded. 5 points for real-time wind direction within ±30° of safe wind direction; If the real-time wind direction is within the danger zone, you will receive 0 points. The real-time wind direction is set with a wind direction risk coefficient based on the degree of deviation from the safe wind direction range. Different wind direction risk coefficients are corresponding to different angle ranges to the left and right of the boundary of the safe wind direction range.

[0007] As a further aspect of the present invention: the wind field safety score is used to classify safety levels, specifically as follows: Total safety score for wind field = wind speed safety score × wind speed safety index weight + wind direction safety score × wind direction safety index weight; The safety level of the cleaning operation is determined based on the calculated total safety score of the wind field. The operation is classified as safe if the total safety score of the wind field is greater than or equal to a certain score, indicating that the wind field is stable and the risk of water mist drift is extremely low. The wind field safety score is classified as critical level (yellow warning) within a specific score range, indicating that the wind field is close to danger and the water mist drift speed is within a specific range. A wind farm with a total safety score below a certain threshold is classified as a dangerous level, i.e., a red alert, indicating a high-risk wind farm and water mist drift speed within a specific range.

[0008] As a further aspect of the present invention: the graded instruction specifically means that, under the safety level, the robot stops at a distance of 1 meter directly in front of the target insulator as planned and moves sequentially; At the critical level: Calculate the fine-tuning distance using the following formula: Fine-tuning distance = average wind speed × yellow warning water mist drift time × wind direction risk coefficient; At the critical level, the fine-tuning direction is determined simultaneously, and the fine-tuning direction angle β is obtained. The calculated fine-tuning distance and β are substituted into the calculation of the X-axis and Y-axis adjustment coordinates at the critical level, and the adjustment coordinates are output. The calculation formula is as follows: X-axis adjustment coordinate = original X-axis coordinate + fine-tuning distance × cosβ; Y-axis adjustment coordinate = original Y-axis coordinate + fine-tuning distance × sinβ; At the danger level: The robot first extracts the coordinates of the originally planned stopping position, and at the same time reads the robot's moving speed, real-time wind direction and straight-line distance to the nearest powered equipment; When the straight-line distance to the nearest live equipment is greater than or equal to 5 meters, the safety coordinate is calculated by the formula: X-axis safety coordinate = original X-axis coordinate + robot moving speed × cosβ × red warning water mist drift time. Y-axis safe coordinate = original Y-axis coordinate + robot moving speed × sinβ × red warning water mist drift time, ensuring that the robot moves in the opposite direction of the water mist, and outputting the X-axis safe coordinate and Y-axis safe coordinate; When the straight-line distance to the nearest live equipment is less than 5 meters, a safety factor of 1.5 is introduced. The safety coordinates are calculated using the formula: X-axis safety coordinate = original X-axis coordinate + robot moving speed × cosβ × red warning water mist drift time × 1.5; Y-axis safety coordinate = original Y-axis coordinate + robot moving speed × sinβ × red warning water mist drift time × 1.5. The X-axis safety coordinates and Y-axis safety coordinates are then output.

[0009] As a further aspect of the present invention, it also includes: The processing module is used to receive multi-source data collected by the data extraction module, and to perform integrity verification, cleaning calculations, and interval definition on the data. Execution module: Used to receive instructions generated by the decision module and execute the corresponding movement control operations.

[0010] As a further aspect of the present invention: the data extraction module includes a robot mobile acquisition unit and a fixed environment monitoring node; The robot mobile data acquisition unit is integrated and installed on the top of the cleaning robot on a 360-degree rotating gimbal. It includes a high-precision wind direction sensor, a lidar and a 1080P low-light camera, which are used to collect wind field data, distance data between the robot and the electrical equipment and insulator status data, respectively. The fixed environmental monitoring nodes are evenly deployed in a 50-meter radius around the key live equipment of the substation. Each node includes a wind speed and direction sensor, a LoRa wireless communication module, a lithium battery and a solar charging panel, which are the same as the robot's mobile data acquisition unit. The node is installed at the same height as the robot's working height and avoids the equipment's heat dissipation vents and obstructions. When the data extraction module collects data, the robot mobile acquisition unit transmits the collected data to the robot main control unit through the internal bus. The fixed environmental monitoring nodes synchronously collect wind speed and wind direction data at their respective locations, calculate the average value, and then synchronize it to the robot main control unit through the LoRa network. If the robot's main control unit does not receive data from a certain acquisition device for more than 10 seconds, the device will trigger a local alarm and send a data acquisition interruption signal to the monitoring layer. When the processing module performs data integrity verification, it verifies the four elements of each data item—timestamp, parameter name, value, and unit—through the content detection unit. If any of the four elements is missing, the data is recorded as invalid data, and the system triggers a re-collection. If three consecutive collections are all invalid data, an invalid data alarm is immediately sent to the monitoring layer, and the current cleaning operation is suspended.

[0011] As a further aspect of the present invention: when the processing module cleans the wind speed data, it uses a multi-group averaging method to calculate the cleaned robot-side wind speed value, collects at least 10 sets of continuous robot-side wind speed data, and calculates the average value as the core wind speed indicator for subsequent risk assessment. Simultaneously, a 1-second data acquisition interval is set to obtain at least 3 consecutive wind speed data from the robot side. The wind speed trend change rate is calculated to reflect the wind speed change trend. The wind speed trend change rate is obtained by subtracting the wind speed data from the previous acquisition from the wind speed data acquired in the later acquisition, and then dividing by the difference in the number of acquisitions. A positive value indicates that the wind speed is increasing, and a negative value indicates that the wind speed is decreasing.

[0012] As a further aspect of the present invention: when the processing module defines the wind direction range, it first reads the relative position of the target insulator and the nearest energized equipment, and sets the 180-degree range away from the equipment as the safe wind direction range. Real-time wind direction within the safe range is marked as a safe state; within ±30 degrees of the range boundary is marked as a critical state; otherwise, it is marked as a dangerous state. Core data is pushed to the decision-making module via the MQTT protocol. The core data includes average wind speed, real-time wind direction, wind speed trend change rate, wind direction status, and equipment distance.

[0013] As a further aspect of the present invention: the decision module monitors the total safety score of the wind field in real time, and updates the path strategy immediately when the warning level changes: when the yellow warning changes to the red warning, the current cleaning operation is stopped, and the adjustment coordinates under the critical level are replanned to the safe coordinates under the danger level, and obstacle avoidance is carried out in real time through lidar during the planning process. When a red alert is switched to a yellow alert, the safe coordinates under the danger level are replanned to the adjustment coordinates under the critical level, and the current insulator cleaning operation is resumed. When the yellow alert is switched to a safe level and the total safety score of the wind farm is stable, the robot will resume docking operations as originally planned; when the red alert is switched to a safe level and the total safety score of the wind farm is stable, the robot will automatically return to the originally planned docking position and operate according to the original plan.

[0014] As a further aspect of the present invention: the execution operation module implements movement control according to different warning instructions: when a yellow warning is issued, it moves to the fine-tuning coordinates and ensures a safe distance; When a red alert is issued, a specific algorithm is used to plan a path, and if a nearby obstacle is encountered, the path is stopped and replanned; when there is no alert, the cleaning system is activated. At the same time, operation information is pushed at fixed intervals, and corresponding response mechanisms are triggered when there is a mobile failure, insufficient water pressure, or communication interruption.

[0015] This invention provides an automated cruise system for cleaning power insulators in substations. Compared with existing technologies, it has the following advantages: (1) The present invention ensures data integrity by verifying four elements through the processing module. Operation is suspended after three consecutive invalid data. The wind speed data is cleaned by multiple average methods and the wind speed change is predicted by the trend change rate. The 180° safe wind direction range is defined by the location of the live equipment. The wind direction is divided into three states: safe, critical, and dangerous. The structured core data is output to provide accurate input for risk assessment. (2) The decision module of this invention sets safety indicators according to the weight of wind speed and wind direction, divides wind speed into 4 risk ranges and wind direction into 3 levels, calculates the total safety score of the wind field by weighted summation, and generates differentiated instructions for different levels, and supports real-time path updates when switching warning levels, so as to realize safe operation under dynamic changes in the wind field. Attached Figure Description

[0016] Figure 1 This is a system flowchart of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0018] Please see Figure 1 This application provides an automatic cruise system for cleaning substation power insulators, and the execution steps are as follows: I. Data Extraction Module: Accurate Collection and Anomaly Monitoring of Multi-Source Data 1.1 Deployment of Data Acquisition Equipment 1.11 Robotic Mobile Data Acquisition Unit: Three types of core sensors are integrated and installed on the 360° rotating gimbal on top of the cleaning robot: High-precision wind direction sensor: measures wind speed and 360° wind direction; LiDAR: used to measure the distance between robots and energized equipment, and to scan the surface condition of insulators; 1080P low-light camera: Supports shooting in low-light environments, assists in identifying the degree of insulator contamination and surface debris, and helps identify the degree of insulator contamination (light / medium / heavy pollution areas) and surface debris (such as branches and bird droppings) through image details, providing a basis for subsequent cleaning parameter adjustments; 1.12 Fixed Environmental Monitoring Nodes: Three fixed nodes are evenly deployed within a 50m radius of key live equipment in the substation to complement and back up data from the mobile units. The node deployment must meet the following requirements: Each node includes wind speed and direction sensors consistent with the robot, a LoRa wireless communication module, a lithium battery, and a solar charging panel. Furthermore, the node installation height is consistent with the robot's working height, while avoiding equipment heat dissipation vents and obstructions to prevent local airflow from interfering with data acquisition.

[0019] 1.2 Raw Data Acquisition and Transmission 1.21 Robot Data Acquisition: The robot collects the following data in real time using three types of core sensors: Wind field data: real-time wind speed, denoted as V_machine, unit: m / s; real-time wind direction, denoted as α_machine, unit: °; Distance data: The straight-line distance between the robot and the nearest powered device, denoted as D, in meters; Insulator condition data: surface images (including information on dirt and debris); The collected data is transmitted to the main control unit via the robot's internal bus (such as CAN bus).

[0020] 1.22 Data Acquisition at Fixed Nodes: Three fixed nodes synchronously collect real-time wind speed (Vfixed1, Vfixed2, Vfixed3) and wind direction (αfixed1, αfixed2, αfixed3) at their respective locations, calculate their average values, and then synchronize them to the robot's main control unit via the LoRa network to achieve multi-source data backup.

[0021] 1.3 Data Acquisition Anomaly Alarm If the robot's main control unit does not receive data from a certain data acquisition device for more than 10 seconds (e.g., sensor disconnection or communication interruption), the device will immediately trigger a local alarm (indicator light flashing) and send a data acquisition interruption signal to the monitoring layer to ensure that abnormalities in the acquisition process can be detected in a timely manner.

[0022] II. Processing Modules: Data Cleaning and Standardization, and Decision-Making Basis The processing module performs integrity verification, cleaning calculations, and interval definition on the collected raw data, transforming unordered data into structured and quantifiable indicators, providing accurate input for subsequent risk assessment.

[0023] 2.1 Multi-source data reception and integrity verification The robot's main control unit has a built-in multi-source data receiving module that simultaneously receives data from both the robot and fixed nodes. A content inspection unit verifies each data entry for four key elements: timestamp, parameter name, value, and unit. If any element is missing, the data is invalid, and the system will trigger a re-collection until complete data is obtained. If three consecutive data collections are invalid, an invalid data alarm will be sent to the monitoring layer immediately, and the current cleaning operation will be suspended to avoid the risk of making decisions based on erroneous data.

[0024] 2.2 Wind speed data cleaning and trend calculation 2.21 Robot wind speed data cleaning To eliminate the interference of instantaneous wind speed fluctuations, the wind speed value after cleaning is calculated using a multi-set averaging method. Ten sets of continuous V-machine data (such as V-machine 1 to V-machine 10) are collected. The average value of the 10 sets of V-machine data is calculated as (V-machine 1 + V-machine 2 ... + V-machine 10) ÷ 10 to obtain the average value of the 10 sets of wind speed on the robot side, V-machine average, which serves as the core wind speed indicator for subsequent risk assessment. 2.22 Wind speed trend calculation To predict wind field changes, a 1-second data acquisition interval was set to obtain three consecutive wind speed data points from the robot (V-machine 1, V-machine 2, V-machine 3). The wind speed trend change rate was calculated using the formula: Wind speed trend change rate = V-machine 3 - V-machine 1 / 3 - 1, to reflect the wind speed change trend. A positive value indicates an increase in wind speed, and a negative value indicates a decrease in wind speed.

[0025] 2.3 Definition of wind direction range By combining the electronic map of the substation with the location of energized equipment, the safe wind direction range is clearly defined, providing a standard for wind direction risk assessment. Obtain relative orientation: Read the relative orientation of the target insulator and the nearest energized equipment from the substation electronic map database (e.g., the energized equipment is located to the east of the insulator). Defining the safe wind direction range: The 180° range away from the live equipment is defined as the safe wind direction range and denoted as α ampere. For example, if the nearest live equipment is located on the east side of the insulator, then the 180° range away from the live equipment, i.e., the opposite direction to the east, is α ampere = 180° - 360°, which is the safe wind direction range. Wind direction status marking: Based on the real-time relationship between wind direction α_machine and α_safety, the wind direction status is marked: Safety Status: If the wind direction α machine is detected to be exactly within the safe wind direction range αan, it is marked as safe. At this time, the wind direction α machine and the safe wind direction range αan form an inclusive relationship, indicating that the wind direction is within a suitable range, the water mist is away from the live equipment, and the risk of cleaning operation is low. Critical state: When the wind direction α machine is not within the safe wind direction range αan, but within 30 degrees to the left and right of the boundary of the safe wind direction range (that is, ∈αan ± 30°), it is marked as critical, indicating that the water mist is close to the live equipment and has approached the danger range, and the risk needs to be monitored. Hazardous condition: When the wind direction α is neither within the safe range α nor within the critical ±30° range, it is marked as hazardous, indicating that the water mist is directed towards the live equipment and the operation is highly risky.

[0026] 2.4 Data Output The system uses the MQTT protocol (a lightweight, highly reliable IoT communication protocol) to push the processed core data to the risk assessment module of the decision-making layer. The output data includes: V-machine level, wind direction α-machine, wind speed trend change rate, wind direction status (safe / critical / dangerous), and equipment distance D.

[0027] III. Decision Module: Wind Farm Risk Classification and Cruise Strategy Generation Based on the standardized data output by the processing module, the decision-making module divides security levels by weight setting and score calculation, generates corresponding early warning instructions and path strategies, and adapts to early warning level switching scenarios to ensure the real-time nature and security of decision-making.

[0028] 3.1 Setting Parameter Indicators 3.11 Indicator Weight Setting Based on the influence of comprehensive wind field data on water mist drift, two types of core indicators were assigned weights: Wind speed safety index: weight 60%; Wind direction safety index: weight 40%; 3.12 Setting of wind speed safety score Divide the score range according to the average wind speed of the V machine to quantify the wind speed risk: When the average wind speed of the V machine ≤ 2 m / s, get 10 points, marked as risk-free state, that is, gentle breeze state; When 2 m / s < average wind speed of the V machine ≤ 3 m / s, get 8 points, marked as low-risk state, that is, low-wind state; When 3 m / s < average wind speed of the V machine ≤ 4 m / s, get 4 points, marked as high-risk state, that is, high-wind state; When the average wind speed of the V machine > 4 m / s, get 0 points, marked as extremely high wind direction state, that is, strong wind state.

[0029] 3.13 Setting of wind direction safety score Divide the score according to the state of the α machine to quantify the wind direction risk: When the α machine ∈ α safe, get 10 points, and at this time the water mist deviates from the live equipment; When the α machine ∈ within α ± 30°, get 5 points, and at this time the water mist approaches the live equipment; When the α machine ∈ the dangerous range, get 0 points, and at this time the water mist faces the live equipment.

[0030] 3.14 Setting of wind direction risk coefficient To accurately calculate the path fine-tuning distance under the yellow warning, set the wind direction risk coefficient (K adjustment) according to the deviation degree between the α machine and α safe: When the α machine ∈ within α safe ± 10°, then K adjustment = 1.0; When within α safe ± 20°, then K adjustment = 1.2; When within α safe ± 30°, then K adjustment = 1.4.

[0031] For specific operations, when the safe wind direction angle range of a certain wind turbine is α safe = 180°, if the actual wind direction α machine is between 170° - 190° (that is, the α machine ∈ within α safe ± 10°), at this time the value of K adjustment is 1.0; if the actual wind direction α machine is between 160° - 200° (that is, within α safe ± 20°), K adjustment will be increased to 1.2; when the wind direction reaches 150° - 210° (that is, within α safe ± 30°), K adjustment is 1.4.

[0032] 3.2 Wind farm safety score 3.21 Calculation of the total wind farm safety score Calculate the total wind farm safety score through the weighted summation formula. Specifically, through the formula: Total wind farm safety score (S total) = Wind speed safety score (S speed) × Wind speed safety index (0.4) + Wind direction safety score (S direction) × Wind direction safety index (0.6).

[0033] 3.22 Classification of cleaning operation safety levels Based on the S-level classification, risks are divided into three levels, with clear characteristics for each level: Safety level: S total score ≥ 6 points, wind field is stable, water mist drift risk is extremely low, and cleaning and patrol operations can be performed normally; Critical level (yellow alert): 4 points ≤ total S < 6 points, the wind field is approaching danger, the water mist drift speed is about 2m / s, and the risk needs to be avoided by fine-tuning the path; Hazard Level (Red Alert): Total S < 4 points, high risk in wind field, water mist drift speed approximately 5 m / s, operations must be stopped immediately and personnel evacuated to a safe location.

[0034] 3.3 Generation of Tiered Early Warning Instructions Based on the three-level classification of cleaning operations, the following corresponding instructions are set: Safety level (total S score ≥ 6): No warning command is generated. The robot does not need to adjust its path and stops at coordinates (X0, Y0) 1m in front of the target insulator as originally planned. It then performs the cleaning operation and moves to the subsequent insulators in sequence. If an unannounced command is triggered, the robot will continue to execute the cleaning command without changing its cruise direction. Critical Level (Yellow Alert): Generate a yellow alert command; If a yellow warning is triggered, the robot retrieves the original planned stopping position (X0, Y0) from the path planning database, which is the ground coordinate 1m in front of the insulator. It also reads the data values ​​of the straight-line distance (D) between the V-machine level, the α-machine, the robot and the nearest live equipment. Based on the wind speed and wind direction risk, it determines the fine-tuning distance (L-adjustment). The formula is: L-adjustment = V-machine level × yellow warning water mist drift time × K-adjustment. At the same time, determine the fine-tuning direction (the direction away from the electrical equipment and opposite to the wind direction). For example, if the wind direction is southeast and the electrical equipment is in the east, then the fine-tuning direction is northwest. The fine-tuning direction angle β is obtained by determining the angle between the fine-tuning direction and the due east direction in the geographic coordinate system.

[0035] For example, if the fine-tuning direction is northwest, in the geographic coordinate system, the angle between this direction and the due east direction is 135°, and this angle is the β value; The calculated L-adjustment and β values ​​are input into the fine-tuning coordinate formula: X-axis adjustment coordinate (X-adjustment) X = X0 + L-adjustment × cosβ, Y-axis adjustment coordinate (Y-adjustment) = Y0 + L-adjustment × sinβ. After obtaining the X-adjustment and Y-adjustment values, the fine-tuning coordinates (X-adjustment and Y-adjustment) are output and the system waits for the execution command. At the same time, the cleaning range is limited: after fine-tuning, only the current insulator is cleaned, and the system does not move to the next insulator in the direction of live equipment to avoid large-scale risks.

[0036] Danger Level (Red Alert): Generate a red alert command; If the red warning instruction is triggered, the acquisition module reads and extracts the original planned position (X0, Y0), and at the same time reads the numerical values of the robot's moving speed, α machine, and the straight-line distance (D) to the nearest live equipment, and makes a judgment according to the numerical value of D; When D≥5m, through the formula: Xsafe = X0 + robot moving speed × cosα machine × red warning water mist drift time, Ysafe = Y0 + robot moving speed × sinα machine × red warning water mist drift time, ensure that the robot moves in the opposite direction of the water mist, obtain the safe coordinates (Xsafe, Ysafe), and output the safe coordinates (Xsafe, Ysafe) waiting for the execution instruction; When D<5m, according to the safety distance specification of live equipment, in the case of a short-distance risk scenario, in order to fully ensure the safety of personnel and equipment, considering multiple conditions such as equipment operation characteristics, environmental factors, and potential risks, finally determine to introduce a safety factor K = 1.5 to enhance the safety protection level, through the formula: Xsafe = X0 + robot moving speed × cosα machine × red warning water mist drift time × K, Ysafe = Y0 + robot moving speed × sinα machine × red warning water mist drift time × K, and output the safe coordinates (Xsafe, Ysafe) waiting for the execution instruction.

[0037] 3.4 Adaptation of warning level switching path The system continuously monitors the total wind field safety score (S total). If a level switch occurs, immediately update the path strategy: If the yellow warning changes to a red warning: then abort the current robot cleaning operation, re-plan from the fine-tuning position (Xadj, Yadj) to the red warning safe coordinates (Xsafe, Ysafe), and avoid obstacles in real time through the lidar; If the red warning changes to a yellow warning, then re-plan from the avoidance position (Xsafe, Ysafe) to the yellow warning fine-tuning coordinates (Xadj, Yadj), and resume cleaning within the effective range; If the yellow warning changes to no warning and ensure that the S total value is stable, then resume the normal cleaning state; If the red warning changes to no warning and ensure that the S total value is stable, the robot automatically returns to the original planned position (X0, Y0) coordinate point and resumes the normal cleaning operation.

[0038] IV. Execution module 4.1 Movement control: Precise positioning and path execution 4.11 Execution of yellow warning fine-tuning movement After receiving the (Xadj, Yadj) instruction, the robot moves towards the target through the walking drive system (crawler type). During the movement: the lidar continuously checks the distance to the live equipment to ensure that the distance ≥ the safety threshold, and at the same time only cleans the current insulator and does not move towards the live equipment to the next insulator.

[0039] 4.12 Execution of red warning avoidance movement Upon receiving the (X Security, Y Security) instruction, the following logic will be followed: Using the shortest path algorithm, A* algorithm, plan the route from the current location to (X An, Y An), and avoid obstacles using LiDAR; It should be noted that the A* (A-Star) algorithm is a heuristic path search algorithm proposed in 1968. Essentially, it is a balance optimization algorithm between Dijkstra's algorithm (globally optimal but inefficient) and greedy best-first search (efficient but possibly locally optimal). By introducing a heuristic function to guide the search direction, it can both guarantee finding the global shortest path (provided that the heuristic function satisfies the acceptability) and significantly reduce invalid search nodes.

[0040] If the lidar detects a distance of ≤3 meters from the powered equipment during the movement, the movement will stop and the path will be replanned. Once the (X-safety, Y-safety) coordinates are reached, a signal indicating that a safe position has been reached is sent to the decision-making module.

[0041] 4.13 Execution of Safe Operations Without Warning If the robot is in the no-warning zone, it will start the cleaning system (high-pressure water mist nozzles in conjunction with cleaning brushes) after reaching the original planned position (X0, Y0) to clean the surface of the insulator.

[0042] 4.2 Status Feedback 4.21 Real-time feedback The execution module pushes abnormal status data to the backend monitoring layer every second, including: The current location coordinates and distance D from the energized equipment are used to ensure that the operation process complies with safety regulations. The system records the cleaning operation status in real time, including whether it is in cleaning mode, the current water pressure parameters, and the remaining cleaning time estimated based on the task progress, providing data support for dynamically adjusting the cleaning strategy. Continuously monitor key hardware indicators, including remaining battery power (percentage), drive system temperature (°C), and the operating status of various sensors, to promptly identify potential equipment malfunctions.

[0043] 4.22 Anomaly Feedback and Handling If the robot cannot reach the designated location during movement (e.g., due to ground obstacles), an audible and visual alarm will be triggered immediately (buzzer sounds + red light flashes), and a "movement fault" signal will be sent to the monitoring layer. This mechanism can quickly attract the attention of on-site personnel and transmit fault information to monitoring personnel in a timely manner, so as to facilitate rapid intervention to remove obstacles and restore robot operation. If the water pressure of the cleaning system is 10% lower than the set value, the cleaning will be paused and a "low water pressure" signal will be sent to prompt the replenishment of cleaning fluid. This can prevent poor cleaning effect due to insufficient water pressure and avoid damage caused by continuous operation of the equipment under abnormal conditions, thus ensuring the quality of the cleaning operation and the safety of the equipment. If communication with the decision-making module is interrupted, the robot immediately stops operation and docks at its current position, saving the state data before the failure locally. This data will be uploaded after communication is restored. This operation prevents the robot from running blindly without control commands, which could lead to danger. At the same time, it preserves key data to provide a basis for subsequent fault analysis and recovery operations, ensuring that the system can quickly and accurately continue the task after communication is restored.

[0044] Example 2 This illustrates another embodiment of the present invention, which is largely the same as the technical solution of Embodiment 1, so only the differences are described.

[0045] Taking the cleaning of P1 insulator string as an example: 1.11: The robot moves to a position 5m away from the P1 insulator string, activates three types of sensors, and collects data in real time: wind speed V_machine = 2.5m / s, wind direction α_machine = 270° (due west); the lidar measures the distance between the robot and the busbar as D = 5.2m; the camera captures an image of the insulator surface, which is identified as a medium-pollution area, free of branches, bird droppings, and other debris. The collected data is transmitted to the main control unit via the CAN bus.

[0046] 1.12: The main control unit receives multi-source data once per second and receives data from the robot and fixed nodes normally for 10 consecutive seconds without alarm triggering. If a fixed node interrupts data transmission due to signal obstruction, the indicator light of that node will flash after 10 seconds, and a data acquisition interruption signal will be sent to the monitoring layer. The main control unit will then switch to using only robot-side data to continue operation.

[0047] 2.1: The main control unit verifies the received data from the robot side and the fixed node side to confirm that the timestamp (accurate to milliseconds), parameter names (wind speed, wind direction, distance, image), values, and units are all complete, there is no invalid data, and there is no need to re-collect data.

[0048] 2.21: Collect 10 consecutive V-machine data (2.5, 2.6, 2.4, 2.5, 2.7, 2.3, 2.5, 2.6, 2.4, 2.5 m / s), calculate the average V-machine average (2.5+2.6+.....+2.5) / 10=2.5 ​​m / s, and use it as the core wind speed indicator.

[0049] 2.22: At 1-second intervals, three V-machine data points were acquired (2.5, 2.6, and 2.7 m / s). Using the formula: Wind speed trend change rate = (V-machine 3 - V-machine 1) / (3 - 1) = (2.7 - 2.5) / 2 = 0.1 m / s2 , it is determined that the wind speed shows an increasing trend.

[0050] 2.3: According to the electronic map, the nearest live equipment to the P1 insulator string is on the west side. Therefore, the 180° range背离带电设备 (east 180°) is defined as the safe wind direction interval α安 = 90° - 270° (due east - due west direction); currently, α机 = 270°, which belongs to the α安全 range, and the wind direction status is marked as: safe.

[0051] 2.4: Push the processed data to the decision-making module through the MQTT protocol. The output content includes: V机平 = 2.5 m / s, α机 = 270°, wind speed trend change rate = 0.1 m / s 2 , wind direction status = safe, D = 5.2 m.

[0052] 3.11: The weight of the wind speed safety index is 60%, and the weight of the wind direction safety index is 40%.

[0053] 3.12: V机平 = 2.5 m / s, which is in the range of 2 m / s < V机平 ≤ 3 m / s, and gets 8 points, that is, S速 = 8.

[0054] 3.13: α机 = 270° ∈ α (90° - 270°), and gets 10 points, that is, S向 = 10.

[0055] 3.21: S总 = S速 × 60% + S向 × 40% = 8 × 0.6 + 10 × 0.4 = 8.8 points.

[0056] 3.3: S总 = 8.8 points > 6 points, it is determined as the safe level, generate a no-warning instruction, and the robot docks 1 m in front of the P1 insulator string (X1, Y1) according to the original plan, performs the cleaning operation, and plans to move to the P2 insulator string after completion.

[0057] 3.4: If during the cleaning of P1, the wind speed suddenly increases to 3.5 m / s, recalculate V机平 = 3.5 m / s, S速 = 4 points; the wind direction shifts to α机 = 300° (beyond α安 = 90° - 270°, and not within the range of α安 ± 30°), S向 = 0 points, S总 = 4 × 0.6 + 0 × 0.4 = 2.4 < 4 points, switch to the dangerous level, immediately generate a red warning instruction, abort the cleaning operation, and plan the evacuation path.

[0058] 4.1 The robot receives the (X1, YI) instruction and moves towards the target position through the crawler walking drive system. The lidar monitors the distance from the bus in real time, keeps D ≥ 5 m, and finally accurately docks at (X1, Y1).

[0059] The cleaning system is activated, and the high-pressure water mist nozzles (water pressure set to 8 MPa) work in conjunction with the cleaning brushes to scrub the surface of the P1 insulator string. A 1080P camera monitors the cleaning results in real time until the image is recognized as clean.

[0060] 4.2: After receiving the red warning command, read the current position (X1, Y1), robot speed (0.5m / s), α_machine = 300°, D = 5.2m ≥ 5m, and calculate the safety coordinates according to the formula: Assuming the red warning water mist drift time = 10s, X_safe = X1 + 0.5 × cos300° = X1 + 2.5, Y_safe = Y1 + 0.5 + sin300° × 10 = Y1 - 4.33. Use the A* algorithm to plan the shortest path from (X1, Y1) to (X_safe, Y_safe), and use the lidar for real-time obstacle avoidance.

[0061] During the movement, the lidar detected that the distance to a certain device had dropped to 2.8m, which is less than the safety threshold of 3m. It immediately stopped moving, replanned the path, and finally arrived at (X safety, Y safety) and sent a signal to the processing module that it had reached the safe position.

[0062] 4.3: During the execution phase, data is pushed to the monitoring layer every second: If there is no warning operation, the current coordinates (X1, Y1), D=5.2m, cleaning mode=cleaning, water pressure=8MPa, remaining cleaning time=3min, battery power=85%, and drive system temperature=42℃; when a red warning is issued for evacuation, the current coordinates, D value, movement speed, and remaining evacuation distance are pushed.

[0063] If the robot encounters obstacles such as stones on the ground during movement and cannot reach (X1, Y1), a buzzer will sound and a red light will flash, sending a movement fault signal to the monitoring layer. After the on-site maintenance personnel remove the stones, the robot will restart its movement. If the cleaning water pressure drops to 7.2MPa (10% lower than the set value of 8MPa), cleaning will be paused and a water pressure deficiency signal will be sent. After the maintenance personnel replenish the cleaning fluid, the operation will resume.

[0064] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0065] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended 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 methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. An automatic cruise system for a substation power insulator cleaning robot, characterized in that, include: The data extraction module is used to collect multi-source data related to the patrol and operation of the insulator cleaning robot in the substation; Decision module: Based on wind field data collected from multiple sources, wind speed and wind direction safety index weights are allocated according to a preset ratio. At the same time, wind speed intervals and wind direction status divisions are set to quantify wind speed safety score and wind direction safety score. The wind speed safety score is divided into four intervals based on the average wind speed on the robot side: no risk, low risk, high risk, and extremely high risk. Different score values ​​are assigned to the four intervals. The average wind speed corresponds to different scores in different value intervals and is marked as the corresponding risk status. The wind direction safety score is divided into three levels based on the real-time wind direction status and the deviation from the safe wind direction, and different score values ​​are assigned to the three levels. The real-time wind direction is divided into different areas within the three levels. Finally, the safety index weights of wind speed and wind direction are respectively incorporated into the wind speed safety score and wind direction safety score to obtain the wind field safety score. Based on the wind field safety score, three safety levels are divided, and corresponding graded warning instructions and appropriate cruise path strategies are issued according to the corresponding safety level.

2. The automatic cruise system for cleaning substation power insulators according to claim 1, characterized in that, The specific weights of the wind speed and wind direction safety indicators are as follows: The safety index for wind speed has a weighting of 60%, while the safety index for wind direction has a weighting of 40%. The wind speed safety score is as follows: an average wind speed of ≤2m / s is considered no risk and scores 10 points; 2m / s < average wind speed ≤3m / s is considered low risk and scores 8 points; 3m / s < average wind speed ≤4m / s is considered high risk and scores 4 points; and an average wind speed >4m / s is considered extremely high risk and scores 0 points. The terms "no risk," "low risk," "high risk," and "extremely high risk" correspond to four states: light wind, low wind, high wind, and strong wind, respectively. The wind direction safety score is as follows: if the real-time wind direction is within the safe wind direction, 10 points are awarded. 5 points for real-time wind direction within ±30° of safe wind direction; If the real-time wind direction is within the danger zone, you will receive 0 points. The real-time wind direction is set with a wind direction risk coefficient based on the degree of deviation from the safe wind direction range. Different wind direction risk coefficients are corresponding to different angle ranges to the left and right of the boundary of the safe wind direction range.

3. The automatic cruise system for a substation power insulator cleaning robot according to claim 1, characterized in that, The wind farm safety score is used to classify safety levels, specifically as follows: Total safety score for wind field = wind speed safety score × wind speed safety index weight + wind direction safety score × wind direction safety index weight; The safety level of the cleaning operation is determined based on the calculated total safety score of the wind field. The operation is classified as safe if the total safety score of the wind field is greater than or equal to a certain score, indicating that the wind field is stable and the risk of water mist drift is extremely low. The wind field safety score is classified as critical level (yellow warning) within a specific score range, indicating that the wind field is close to danger and the water mist drift speed is within a specific range. A wind farm with a total safety score below a certain threshold is classified as a dangerous level, i.e., a red alert, indicating a high-risk wind farm and water mist drift speed within a specific range.

4. The automatic cruise system for a substation power insulator cleaning robot according to claim 1, characterized in that, The specific graded instruction is as follows: Under the safety level, the robot stops 1 meter in front of the target insulator as planned and moves sequentially. At the critical level: Calculate the fine-tuning distance using the following formula: Fine-tuning distance = average wind speed × yellow warning water mist drift time × wind direction risk coefficient; At the critical level, the fine-tuning direction is determined simultaneously, and the fine-tuning direction angle β is obtained. The calculated fine-tuning distance and β are substituted into the calculation of the X-axis and Y-axis adjustment coordinates at the critical level, and the adjustment coordinates are output. The calculation formula is as follows: X-axis adjustment coordinate = original X-axis coordinate + fine-tuning distance × cosβ; Y-axis adjustment coordinate = original Y-axis coordinate + fine-tuning distance × sinβ; At the danger level: The robot first extracts the coordinates of the originally planned stopping position, and at the same time reads the robot's moving speed, real-time wind direction and straight-line distance to the nearest powered equipment; When the straight-line distance to the nearest live equipment is greater than or equal to 5 meters, the safety coordinate is calculated by the formula: X-axis safety coordinate = original X-axis coordinate + robot moving speed × cosβ × red warning water mist drift time. Y-axis safe coordinate = original Y-axis coordinate + robot moving speed × sinβ × red warning water mist drift time, ensuring that the robot moves in the opposite direction of the water mist, and outputting the X-axis safe coordinate and Y-axis safe coordinate; When the straight-line distance to the nearest live equipment is less than 5 meters, a safety factor of 1.5 is introduced. The safety coordinates are calculated using the formula: X-axis safety coordinate = original X-axis coordinate + robot moving speed × cosβ × red warning water mist drift time × 1.5; Y-axis safety coordinate = original Y-axis coordinate + robot moving speed × sinβ × red warning water mist drift time × 1.

5. The X-axis safety coordinates and Y-axis safety coordinates are then output.

5. The automatic cruise system for a substation power insulator cleaning robot according to claim 1, characterized in that, Also includes: The processing module is used to receive multi-source data collected by the data extraction module, and to perform integrity verification, cleaning calculations, and interval definition on the data. Execution module: Used to receive instructions generated by the decision module and execute the corresponding movement control operations.

6. The automatic cruise system for a substation power insulator cleaning robot according to claim 5, characterized in that, The data extraction module includes a robot mobile data acquisition unit and a fixed environmental monitoring node; The robot mobile data acquisition unit is integrated and installed on the top of the cleaning robot on a 360-degree rotating gimbal. It includes a high-precision wind direction sensor, a lidar and a 1080P low-light camera, which are used to collect wind field data, distance data between the robot and the electrical equipment and insulator status data, respectively. The fixed environmental monitoring nodes are evenly deployed in a 50-meter radius around the key live equipment of the substation. Each node includes a wind speed and direction sensor, a LoRa wireless communication module, a lithium battery and a solar charging panel, which are the same as the robot's mobile data acquisition unit. The node is installed at the same height as the robot's working height and avoids the equipment's heat dissipation vents and obstructions. When the data extraction module collects data, the robot mobile acquisition unit transmits the collected data to the robot main control unit through the internal bus. The fixed environmental monitoring nodes synchronously collect wind speed and wind direction data at their respective locations, calculate the average value, and then synchronize it to the robot main control unit through the LoRa network. If the robot's main control unit does not receive data from a certain acquisition device for more than 10 seconds, the device will trigger a local alarm and send a data acquisition interruption signal to the monitoring layer. When the processing module performs data integrity verification, it verifies the four elements of each data item—timestamp, parameter name, value, and unit—through the content detection unit. If any of the four elements is missing, the data is recorded as invalid data, and the system will trigger a re-collection. If three consecutive data collections are invalid, an invalid data alarm will be sent to the monitoring layer immediately, and the current cleaning operation will be suspended.

7. The automatic cruise system for a substation power insulator cleaning robot according to claim 5, characterized in that, When cleaning the wind speed data, the processing module uses a multi-group averaging method to calculate the cleaned robot-side wind speed value, collects at least 10 sets of continuous robot-side wind speed data, and calculates the average value as the core wind speed indicator for subsequent risk assessment. Simultaneously, a 1-second data acquisition interval is set to obtain at least 3 consecutive wind speed data from the robot side. The wind speed trend change rate is calculated to reflect the wind speed change trend. The wind speed trend change rate is obtained by subtracting the wind speed data from the previous acquisition from the wind speed data acquired in the later acquisition, and then dividing by the difference in the number of acquisitions. A positive value indicates that the wind speed is increasing, and a negative value indicates that the wind speed is decreasing.

8. The automatic cruise system for a substation power insulator cleaning robot according to claim 5, characterized in that, When defining a wind direction range, the processing module first reads the relative position of the target insulator and the nearest energized equipment, and sets the 180-degree range away from the equipment as the safe wind direction range. Real-time wind direction within the safe range is marked as a safe state; within ±30 degrees of the range boundary is marked as a critical state; otherwise, it is marked as a dangerous state. Core data is pushed to the decision-making module via the MQTT protocol. The core data includes average wind speed, real-time wind direction, wind speed trend change rate, wind direction status, and equipment distance.

9. The automatic cruise system for a substation power insulator cleaning robot according to claim 1, characterized in that, The decision module monitors the total safety score of the wind field in real time. When the warning level changes, it immediately updates the path strategy: when the yellow warning changes to the red warning, the current cleaning operation is stopped, and the adjustment coordinates under the critical level are replanned to the safe coordinates under the dangerous level. During the planning process, obstacle avoidance is carried out in real time through lidar. When a red alert is switched to a yellow alert, the safe coordinates under the danger level are replanned to the adjustment coordinates under the critical level, and the current insulator cleaning operation is resumed. When the yellow alert is switched to a safe level and the total safety score of the wind farm is stable, the robot will resume docking operations as originally planned; when the red alert is switched to a safe level and the total safety score of the wind farm is stable, the robot will automatically return to the originally planned docking position and operate according to the original plan.

10. The automatic cruise system for a substation power insulator cleaning robot according to claim 5, characterized in that, The execution module implements movement control according to different warning instructions: when a yellow warning is issued, it moves to the fine-tuning coordinates and ensures a safe distance; When a red alert is issued, a specific algorithm is used to plan a path, and if a nearby obstacle is encountered, the path is stopped and replanned; when there is no alert, the cleaning system is activated. At the same time, operation information is pushed at fixed intervals, and corresponding response mechanisms are triggered when there is a mobile failure, insufficient water pressure, or communication interruption.

Images