Inspection robot centralized control method and system with quick response

Abstract

The invention discloses a quick-response centralized control method and system for an inspection robot, and the method comprises the steps: enabling the inspection robot to collect target data, and enabling a robot edge end to carry out the preprocessing of the collected data; after the inspection robot reaches a predetermined starting point, the central end plans an inspection path according to the inspection range and sends the inspection path to the inspection robot; the robot edge end uploads the preprocessed data to the central end, a data model is established for data analysis and real-time early warning, and centralized control over the inspection robot is achieved. According to the method, the robot plans the inspection path through the central system based on the preprocessed data, the inspection efficiency and the energy utilization rate are improved to the maximum extent, and comprehensive and efficient inspection coverage is guaranteed. Meanwhile, real-time analysis and early warning are carried out by establishing a data model, abnormal conditions can be found and responded in time, operation safety is improved, downtime caused by sudden faults is shortened, timely repair and maintenance are facilitated, and long-term stable operation of equipment is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of inspection robots, and specifically to a centralized control method and system for inspection robots with rapid response. Background Art

[0002] The emergence of inspection robots has brought unprecedented possibilities and changes to us. With their unique advantages, they are gradually changing our working and living ways. From industrial production to energy facilities, from urban management to public safety, the figures of inspection robots can be found everywhere. Automatic inspection: They can autonomously carry out inspection work according to preset routes and tasks, whether in factory workshops, power facilities or petrochemical industries. Data collection and analysis: By being equipped with various sensors and monitoring devices, they can collect a large amount of data in real time, such as temperature, humidity, pressure, etc. This data can help staff promptly discover potential problems and conduct targeted maintenance and repair. Safety monitoring: Real-time monitoring of abnormal situations in the environment, such as fires, leaks, etc., and timely issuing of alarms to ensure the safety of personnel and equipment.

[0003] The substation inspection robot system is modularly divided according to functions, including a walking drive module, a power management module, a navigation module, a detection module, a safety protection module, etc. These modules are independent of each other but interrelated. The existing centralized control systems for inspection robots generally have the same architecture in the centralized control system. However, when some emergencies occur, such as relatively large changes in the environmental state and the robot parameter state, they cannot carry out early warning processing according to the actual situation, resulting in not very good overall use effects. Summary of the Invention

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

[0005] Therefore, the technical problem solved by the present invention is that the existing centralized control system for inspection robots is difficult to respond quickly to emergencies and cannot give early warnings for emergencies.

[0006] To solve the above technical problem, the present invention provides the following technical solution: A centralized control method for inspection robots with rapid response, including: the inspection robot collects target data, and the robot edge end preprocesses the collected data; after the inspection robot reaches the predetermined starting point, the central end plans an inspection path according to the inspection range and sends it to the inspection robot; the robot edge end uploads the preprocessed data to the central end, establishes a data model for data analysis and gives real-time warnings, so as to realize the centralized control of the inspection robot.

[0007] As a preferred scheme of the centralized control method of the fast-response inspection robot described in the present invention, the target data is divided into environmental data and robot data, the environmental data includes temperature, humidity, wind speed and road condition data, and the robot data includes the temperature, current, voltage and running time data of the inspection robot itself.

[0008] As a preferred solution of the centralized control method of the fast-response inspection robot described in the present invention, the preprocessing includes using a low-pass filter to remove high-frequency noise, retain low-frequency signals, and smooth the collected target data, which is expressed as:

[0009]

[0010] Where x(t) represents the filtered data; x(τ) represents the target data collected, and T represents the time window; the Z-score of each data point is calculated, and the data exceeding the Z-score threshold is judged as an outlier and removed; for each type of data collected, a risk threshold δ is set based on historical normal data l and the warning threshold δ h , the robot edge compares the collected real-time data with the set threshold. When x(t)<δ l When δ l ≤x(t)<δ h When x(t)≥δ h When the data exceeds the safety range, the inspection robot will sound an alarm and send a warning signal to the central end, which will also issue a warning at the same time.

[0011] As a preferred solution of the centralized control method of the fast-response inspection robot described in the present invention, wherein: the central end plans the inspection path according to the inspection range, including: the central end allocates an inspection area to the inspection robot, divides the inspection area into irregular grids, the edges of the grids represent the inspection path, the intersection of the inspection paths is the node, the inspection robot starts from the starting point, inspects the entire path and returns to the starting point, completing an area inspection;

[0012] Use the A* algorithm to calculate the initial optimal path, consider the moving distance and energy consumption to calculate the actual cost g(n), and use the Manhattan distance to estimate the estimated cost h(n) from the current node to the end point. Add a penalty term p(n) for unvisited nodes and paths to ensure that all nodes and paths are visited, expressed as:

[0013] g(n)=g prev (n)+d(n prev ,n)+e(n prev ,n)

[0014] h(n) = |x goal - x n | + |y goal - y n

[0015] p(n) = λ·(U(n) + V(n))

[0016] The optimal path is the path with the minimum total cost f(n), which is expressed as:

[0017] f(n) = g(n) + h(n) + p(n)

[0018] Among them, represents the cumulative cost from the previous node to the current node; d(n prev , n) represents the actual distance from the previous node to the current node; e(n prev , n) represents the energy consumption from the previous node to the current node; (x goal , y goal ) represents the coordinates of the end point; (x n , y n ) represents the coordinates of the current node; λ represents the penalty coefficient; U(n) represents the number of unvisited nodes; V(n) represents the number of unvisited paths; Use sensors to detect the path condition in real time. When there are obstacles on the inspection path, automatic obstacle avoidance is performed. After the inspection is completed, return to the starting point and send a signal indicating that the inspection is completed to the central terminal.

[0019] As a preferred solution of the centralized control method for the fast - response inspection robot described in the present invention, among them: when there are obstacles on the inspection path, automatic obstacle avoidance includes using sensors to detect the obstacle position (x o , y o ) and width W o . If the remaining space on the path is not enough for the inspection robot to pass, the inspection robot returns to the previous node position, marks the obstacle position as a break point on the irregular grid edge in the inspection area, and re - plans the optimal path of the uninspected path through the A* algorithm; if the remaining space on the path is enough for the inspection robot to pass, then according to the obstacle position (x o , y o ) and the inspection robot position (x r , y r calculate the obstacle - avoidance direction, control the robot to turn and move to avoid the obstacle, which is expressed as:

[0020]

[0021] θ avoid = θ obs + α

[0022] Among them, θ obsIndicates the direction of the obstacle relative to the robot; α represents the obstacle avoidance angle.

[0023] As a preferred solution of the centralized control method for the fast-response inspection robot described in the present invention, wherein: the data model includes that when the risk data is uploaded by the robot edge end, for the environmental data, the risk data is classified and divided by a time window, and the trend of the risk data is further analyzed, and the change rate of the risk data is calculated to judge the slope of the trend:

[0024]

[0025] Among them, ΔX i (t) represents the change rate at time t; X i (t) represents the original data uploaded by the robot edge end; Δt represents the time interval; the change rate threshold δ of the environmental data is set according to historical data X , when ΔX i (t) ≥ δ X It means that the data corresponding to ΔX i (t) grows too fast, and the central end issues a warning.

[0026] As a preferred solution of the centralized control method for the fast-response inspection robot described in the present invention, wherein: the data model further includes that when there is robot data in the risk data uploaded by the robot edge end, the environmental data is retrieved according to the robot data for collaborative analysis;

[0027] When the temperature data of the robot is risk data, the environmental temperature data at the same time point is retrieved. If the environmental temperature data is normal data and the temperature of the inspection robot continues to rise over time, an alarm for abnormal temperature of the inspection robot is issued, the inspection task of the corresponding inspection robot is stopped, and the staff is dispatched for inspection and repair. If the environmental temperature T env is risk data, the temperature of the inspection robot is predicted through a linear model and compared with the real-time data T robot as follows:

[0028]

[0029]

[0030] Among them, α 1 represents the linear increase; β 1 represents the coefficient; ΔT represents the difference between the real-time temperature and the predicted value; when the voltage and current data of the robot are risk data, the environmental humidity data at the same time point is retrieved. If the environmental humidity data is normal data, an alarm for hardware failure of the inspection robot is issued, the inspection task of the corresponding inspection robot is stopped, and the staff is dispatched for inspection and repair. If the environmental humidity Henv For risk data, the Kalman filter is used to predict the impact of environmental humidity on the current and voltage of the robot, and early warnings are made through state estimation and measurement updates, expressed as:

[0031]

[0032] Wherein, represents the predicted current; represents the predicted voltage; A, B, C, and D represent the state matrices of the Kalman filter.

[0033] In a second aspect, the present invention also provides a centralized control system for a fast-response inspection robot, including a data acquisition module that collects environmental data and robot data through sensors, preprocesses the data at the robot edge, and uploads the preprocessed data to the central end; a walking drive module, where the central end calculates the initial optimal path through the A* algorithm and sends it to the inspection robot to perform the inspection task. The robot edge automatically avoids obstacles during the inspection. When the obstacle cannot be avoided, a signal is sent to the central end to recalculate the optimal path; an early warning module, where the central end analyzes the trend of environmental data based on the uploaded data and performs collaborative analysis and early warning on robot data and environmental data to achieve centralized control of the inspection robot.

[0034] In a third aspect, the present invention also provides a computing device, including: a memory and a processor;

[0035] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the centralized control method for the fast-response inspection robot are implemented.

[0036] In a fourth aspect, the present invention also provides a computer-readable storage medium that stores computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the centralized control method for the fast-response inspection robot are implemented.

[0037] Advantages of the present invention: In the method of the present invention, the robot preprocesses data at the edge end, effectively reducing the amount of data transmitted to the central system, optimizing the communication bandwidth and delay. The central system plans the inspection path based on the preprocessed data, maximizing the inspection efficiency and energy utilization rate, and ensuring comprehensive and efficient inspection coverage. At the same time, through the establishment of a data model for real-time analysis and early warning, abnormal situations can be discovered and responded to in a timely manner, improving the operation safety, reducing the downtime caused by sudden failures, and facilitating timely repair and maintenance, thereby ensuring the long-term stable operation of the equipment. Description of the Drawings

[0038] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0039] Figure 1 It is the overall flowchart of a centralized control method for a fast-response inspection robot provided by an embodiment of the present invention. Specific embodiments

[0040] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the specific embodiments of the present invention in detail with reference to the accompanying drawings of the specification. Obviously, the described embodiments are some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0041] Embodiment 1

[0042] Refer to Figure 1 , which is an embodiment of the present invention, providing a centralized control method for a fast-response inspection robot, including:

[0043] S1: The inspection robot collects target data, and the robot edge end preprocesses the collected data.

[0044] Furthermore, the target data is divided into environmental data and robot data. The environmental data includes temperature, humidity, wind speed, and road condition data, and the robot data includes the temperature, current, voltage, and running time data of the inspection robot itself.

[0045] The environmental temperature is collected by sensors, and the thermal imaging function can be added to identify high-temperature hotspots in the environment to prevent fires; the road condition data is real-time data collected by the inspection robot during the inspection process, and the path conditions and obstacle positions and distances are captured by the camera group.

[0046] Among the robot data, the collection points of the robot temperature are set near the key electronic components inside the robot, such as the main control chip, power motor, and battery pack positions. These components are easily affected by the environmental temperature, and their temperature changes have a direct impact on the performance of the robot.

[0047] The robot current is collected at the output line of the battery or the input end of the motor controller, which can monitor the current consumption of the entire system. The voltage is collected at the output end of the battery or the input end of the main control circuit board. The voltage changes at these positions can reflect the health status of the power supply system.

[0048] Furthermore, a low-pass filter is used to remove high-frequency noise, retain low-frequency signals, and smooth the collected target data, which is expressed as:

[0049]

[0050] where x(t) represents the filtered data; x(τ) represents the collected target data, and T represents the time window.

[0051] Calculate the Z-score for each data point, and determine and eliminate the data points that exceed the Z-score threshold. First, calculate the mean μ and standard deviation σ of each type of data:

[0052]

[0053] Then, calculate the Z-score based on the mean and standard deviation and detect outliers. When the calculated z(i) exceeds the set threshold, the corresponding data point is determined as an outlier and eliminated, which is expressed as:

[0054]

[0055] where N represents the total number of data; x(i) represents the i-th data.

[0056] For each type of collected data, set the risk threshold δ l and the warning threshold δ h , and the edge side of the robot compares the collected real-time data with the set thresholds. When x(t) < δ l , mark x(t) as normal data; when δ l ≤x(t) < δ h , mark x(t) as risk data and upload it to the central end for trend analysis; when x(t) ≥ δ h , it means that the data has exceeded the safe range, the inspection robot issues an alarm, and sends a warning signal to the central end for warning at the central end simultaneously.

[0057] S2: After the inspection robot reaches the predetermined starting point, the central end plans the inspection path according to the inspection range and sends it to the inspection robot.

[0058] Furthermore, the central end assigns an inspection area to the inspection robot, divides the inspection area into irregular grids, the edges of the grids represent the inspection paths, the positions where the inspection paths intersect are nodes, and the inspection robot starts from the starting point, inspects all the paths and returns to the starting point to complete an area inspection.

[0059] Calculate the initial optimal path using the A* algorithm, calculate the actual cost g(n) considering the moving distance and energy consumption, estimate the estimated cost h(n) from the current node to the end point through the Manhattan distance, and add a penalty term p(n) for unvisited nodes and paths to ensure that all nodes and paths are visited, expressed as:

[0060] g(n) = g prev (n) + d(n prev , n) + e(n prev , n)

[0061] h(n) = |x goal - x n | + |y goal - y n

[0062] p(n) = λ · (U(n) + V(n))

[0063] The optimal path is the path with the minimum total cost f(n), expressed as:

[0064] f(n) = g(n) + h(n) + p(n)

[0065] Among them, it represents the cumulative cost from the previous node to the current node; d(n prev , n) represents the actual distance from the previous node to the current node; e(n prev , n) represents the energy consumption from the previous node to the current node; (x goal , y goal ) represents the coordinates of the end point; (x n , y n ) represents the coordinates of the current node; λ represents the penalty coefficient; U(n) represents the number of unvisited nodes; V(n) represents the number of unvisited paths.

[0066] The actual distance d(n prev , n) is the Euclidean distance or Manhattan distance from the previous node to the current node. For the actual distance in the inspection area grid, it is calculated using the Manhattan distance through the grid coordinates:

[0067] d(n prev , n) = |x prev - x n | + |y prev - y n |

[0068] The energy consumption e(n prev, n) It is estimated based on the empirical data during the actual operation of the robot or a pre-determined energy consumption model. The factors considered include the moving distance, terrain conditions, and the load carried by the robot. When finding the optimal path, overly precise calculation of energy consumption is not required. Calculation through a simple linear model ensures the path-finding efficiency, and only the relationship between the moving distance and energy consumption needs to be considered, which is expressed as:

[0069] e(n prev , n) = k · d(n prev , n)

[0070] Among them, k represents the energy consumption per unit distance.

[0071] It should be noted that since the inspection robot needs to pass through all paths for each inspection, generally speaking, the terrain conditions have little impact on energy consumption. Under normal circumstances, the inspection robot does not carry too much load. Therefore, these two points, namely the terrain conditions and the load carried by the robot, can be ignored. The path conditions are detected in real time using sensors. When there are obstacles on the inspection path, automatic obstacle avoidance is performed. After the inspection is completed, the robot returns to the starting point and sends a signal indicating the completion of the inspection to the central terminal.

[0072] It should be noted that in the inspection area, it is inevitable that there may be some inspection areas that can only be reached by one path, and this path has only one intersection point with other paths. The inspection robot must enter this path for inspection and then return the same way. Therefore, a penalty term p(n) is added to ensure that the inspection robot passes through all nodes and paths.

[0073] The actual cost g(n) reflects the true cost from the current node to the end point, including the moving distance and energy consumption. The estimated cost h(n) provides an estimated cost from the current node to the end point to ensure the global optimality of the path. The penalty term p(n) is used to ensure that all nodes and paths are visited, so as to find the optimal path with the least consumption passing through all inspection areas and paths.

[0074] Furthermore, sensors are used to detect the position of the obstacle (x o , y o ) and the width W o . If the remaining space on the path is not sufficient for the inspection robot to pass through, the inspection robot returns to the previous node position, marks the position of the obstacle on the irregular grid edge of the inspection area as a breakpoint, and re-plans the optimal path for the un-inspected path through the A* algorithm.

[0075] If the remaining space on the path is sufficient for the inspection robot to pass through, then according to the position of the obstacle (x o , y o ) and the position of the inspection robot (x r , y r)Calculate the obstacle avoidance direction, control the robot to turn and move, and avoid obstacles, expressed as:

[0076]

[0077] θ avoid = θ obs + α

[0078] Among them, θ obs represents the direction of the obstacle relative to the robot; α represents the obstacle avoidance angle. This obstacle avoidance method can be directly implemented on the edge chip of the inspection robot, without the need to upload it to the central end for analysis. While achieving fast obstacle avoidance, the saved part of the traffic is used to transmit other collected data, ensuring the real-time nature of early warning.

[0079] S3: The edge end of the robot uploads the preprocessed data to the central end, establishes a data model for data analysis and real-time early warning, and realizes the centralized control of the inspection robot.

[0080] Furthermore, the inspection robot uploads the preprocessed data to the central end server through a wireless network. When the edge end of the robot uploads risk data, for environmental data, the risk data is classified and divided through a time window, and the trend of the risk data is further analyzed, and the change rate of the risk data is calculated to judge the slope of the trend:

[0081]

[0082] Among them, ΔX i (t) represents the change rate at time t; X i (t) represents the original data uploaded by the edge end of the robot; Δt represents the time interval; the change rate threshold δ of the environmental data is set according to historical data X , when ΔX i (t) ≥ δ X , it means that the data corresponding to ΔX i (t) grows too fast, and the central end issues an early warning.

[0083] It should be noted that the change rate threshold δ X can be set according to the inspection environment. For example, when there are high-temperature devices in the inspection environment, the value of δ X can be increased to prevent false alarms.

[0084] Furthermore, when there is robot data in the risk data uploaded by the edge end of the robot, the environmental data is retrieved according to the robot data for collaborative analysis.

[0085] When the temperature data of the robot is risk data, retrieve the environmental temperature data at the same time point. If the environmental temperature data is normal data and the temperature of the patrol robot continues to rise over time, an alarm for abnormal temperature of the patrol robot is issued, the patrol task of the corresponding patrol robot is stopped, and staff is dispatched for inspection and repair. If the environmental temperature T env is risk data, then predict the temperature of the patrol robot through a linear model and compare it with the real-time data T robot as follows:

[0086]

[0087]

[0088] where α 1 represents the linear increase; β 1 represents the coefficient; ΔT represents the difference between the real-time temperature and the predicted value.

[0089] When the voltage and current data of the robot are risk data, retrieve the environmental humidity data at the same time point. If the environmental humidity data is normal data, an alarm for hardware failure of the patrol robot is issued, the patrol task of the corresponding patrol robot is stopped, and staff is dispatched for inspection and repair. If the environmental humidity H env is risk data, then use the Kalman filter to predict the influence of environmental humidity on the current and voltage of the robot, and issue a warning through state estimation and measurement update, which is expressed as:

[0090]

[0091] where represents the predicted current; represents the predicted voltage; A, B, C, D represent the state matrices of the Kalman filter.

[0092] If the calculated predicted values of voltage and current are the same as the collected ones, it means that too high humidity or water vapor in the environment affects the circuit of the patrol robot, and an alarm for high environmental humidity is issued, and staff is dispatched for handling.

[0093] It should be noted that the environmental temperature has a direct impact on the temperature of the patrol robot. By monitoring whether the temperature of the robot rises due to the increase in environmental temperature, corresponding cooling measures are taken.

[0094] The threshold value of the temperature difference should be a range rather than a fixed value. Within this range ΔT th it is judged that the change in the robot's temperature is caused by the change in the environmental temperature. When ΔT exceeds the threshold range of the temperature difference ΔT thWhen it indicates that the temperature of the inspection robot's own chip or circuit is too high, the central terminal sends a signal to the robot to pause the inspection task and continuously monitors the robot's temperature. If the temperature cannot drop to the normal data range without task load, it is determined that there is a hardware failure in the inspection robot, an alarm is issued, and staff are dispatched for inspection and repair.

[0095] This embodiment also provides a centralized control system for a fast-response inspection robot, including a data acquisition module that collects environmental data and robot data through sensors, and the robot edge preprocesses the data and uploads the preprocessed data to the central terminal; a walking drive module, where the central terminal calculates the initial optimal path through the A* algorithm and sends it to the inspection robot to execute the inspection task. The robot edge automatically avoids obstacles during the inspection. When an obstacle cannot be avoided, a signal is sent to the central terminal to recalculate the optimal path; an early warning module, where the central terminal analyzes the trend of environmental data based on the uploaded data and collaboratively analyzes and warns the robot data and environmental data to achieve centralized control of the inspection robot.

[0096] This embodiment also provides a computing device, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for centralized control of a fast-response inspection robot as proposed in the above embodiment.

[0097] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the method for centralized control of a fast-response inspection robot as proposed in the above embodiment.

[0098] The storage medium proposed in this embodiment and the method for centralized control of a fast-response inspection robot proposed in the above embodiment belong to the same inventive concept. Technical details not described in detail in this embodiment can be referred to in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0099] From the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software and necessary general hardware. Of course, it can also be implemented by hardware, but in many cases, the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk, or optical disc of a computer, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0100] Example 2

[0101] The following is an embodiment of the present invention, which provides a centralized control method for a fast-response inspection robot. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0102] A simulated inspection area for the simulation experiment is established. The simulation experiment of the path-finding effect is carried out by the inspection robot. The path-finding algorithms used are: the traditional A algorithm, the Dijkstra algorithm, and the improved A algorithm of the present invention. In the same inspection area, multiple nodes and obstacles are set to form a complex inspection path. The total moving distance, energy consumption, and time of each inspection are recorded, and multiple experiments are carried out and the average data is taken as shown in the following table.

[0103] Table 1 Data table of path-finding experiment

[0104]

[0105] It can be seen from the data table that the improved A* algorithm of the present invention is higher than the traditional algorithm in terms of distance, energy consumption, and time. However, according to the unvisited nodes and paths, the path-finding algorithm of the present invention makes the inspection robot preferentially select the optimal route that can access all paths and nodes through the penalty term, for the purpose of applying to the inspection area of the present invention rather than finding the path with the least consumption.

[0106] In the simulated inspection area, by artificially setting heat sources and humidifiers, the warning capabilities of the inspection robot in case of high temperature, high humidity, etc. are respectively simulated. The inspection robot performs inspection tasks in different areas, collects data such as temperature, humidity, and path conditions, and records the warning status of the robot.

[0107] Table 2 Data table of warning experiment

[0108]

[0109] It can be seen from the data that at 10:05, the inspection robot passed through a high-temperature area. The ambient temperature itself was within the acceptable range, but due to the rapid increase in temperature caused by artificial heating, the inspection robot issued a warning, verifying the analysis and warning of the temperature trend by the central end of the present invention.

[0110] After canceling the warning information, the inspection robot continued to work. At 10:10, it passed through a high-humidity area. The edge end of the robot judged that the humidity value exceeded the set humidity threshold, and the inspection robot directly reported an alarm. At the same time, the excessive humidity affected the current and voltage of the inspection robot. Through collaborative analysis, it was judged that the environmental humidity was too high, which was consistent with the judgment of the edge end of the robot, and the central end issued a warning synchronously.

[0111] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A fast-response inspection robot centralized control method, characterized in that: include: The inspection robot collects target data, and the robot edge end pre-processes the collected data; After the inspection robot arrives at the predetermined starting point, the central end plans the inspection path according to the inspection range and sends it to the inspection robot; The robot edge uploads the pre-processed data to the central end, builds a data model for data analysis and real-time warning, and realizes centralized control of the inspection robot.

2. The fast-response inspection robot centralized control method according to claim 1, characterized in that: The target data is divided into environmental data and robot data. The environmental data includes temperature, humidity, wind speed and road condition data, and the robot data includes the temperature, current, voltage and running time data of the inspection robot itself.

3. The fast-response inspection robot centralized control method according to claim 2, characterized in that: The preprocessing includes using a low-pass filter to remove high-frequency noise, retain low-frequency signals, and smooth the collected target data, which is expressed as: Among them, x ( t ) Represents the filtered data; x ( τ ) represents the target data collected, and T represents the time window; Calculate the Z-score of each data point, and remove data that exceeds the Z-score threshold as outliers; For each type of data collected, set the risk threshold δ based on historical normal data l and the warning threshold δ h , the robot edge compares the collected real-time data with the set threshold. ( t ) <δ l When x ( t ) Mark as normal data; When δ l ≤x ( t ) <δ h When x ( t ) Mark as risk data and upload to the central end for trend analysis; When x ( t ) ≥δ h When the data exceeds the safety range, the inspection robot will sound an alarm and send a warning signal to the central end, which will also issue a warning at the same time.

4. The fast-response inspection robot centralized control method according to claim 3, characterized in that: The central end plans the inspection path according to the inspection range, including: the central end allocates an inspection area to the inspection robot, divides the inspection area into irregular grids, the edges of the grids represent the inspection paths, the intersections of the inspection paths are nodes, the inspection robot starts from the starting point, inspects the entire path and returns to the starting point, completing an area inspection; Use the A* algorithm to calculate the initial optimal path, consider the moving distance and energy consumption to calculate the actual cost g(n), and use the Manhattan distance to estimate the estimated cost h(n) from the current node to the end point. Add a penalty term p(n) for unvisited nodes and paths to ensure that all nodes and paths are visited, expressed as: g(n)=g prev (n)+d(n prev ,n)+e(n prev ,n) h(n)=|x goal -x n |+|y goal -y n | p(n)=λ·(U(n)+V(n)) The optimal path is the path with the smallest total cost f(n), expressed as: f(n)=g(n)+h(n)+p(n) Where, represents the cumulative cost from the previous node to the current node; d(n prev ,n) represents the actual distance from the previous node to the current node; e(n prev ,n) represents the energy consumption from the previous node to the current node; (x goal ,y goal ) represents the coordinates of the end point; (x n ,y n ) represents the coordinates of the current node; λ represents the penalty coefficient; U(n) represents the number of unvisited nodes; V(n) represents the number of unvisited paths; Use sensors to detect path conditions in real time, automatically avoid obstacles when there are any on the inspection path, return to the starting point after the inspection is completed, and send an inspection completion signal to the central end.

5. The fast-response inspection robot centralized control method according to claim 4, characterized in that: When there is an obstacle on the inspection path, the automatic obstacle avoidance includes using a sensor to detect the obstacle position in real time (x o ,y o ) and width W o ,If the remaining space in the path is insufficient for the inspection robot to pass, the inspection robot returns to the previous node position, marks the obstacle position on the edge of the irregular grid in the inspection area as a breakpoint, and replans the optimal path of the uninspected path through the A* algorithm; If the remaining space on the path is sufficient for the inspection robot to pass through, then according to the obstacle position (x o ,y o ) and the inspection robot position (x r ,y r ) calculates the obstacle avoidance direction, controls the robot to turn and move, and avoids obstacles, which can be expressed as: i avoid =θ obs +a Among them, θ obs represents the direction of the obstacle relative to the robot; α represents the obstacle avoidance angle.

6. The fast-response inspection robot centralized control method according to claim 5, characterized in that: The data model includes, when the robot edge uploads risk data, for environmental data, classifying the risk data and dividing it by time window, further analyzing the trend of the risk data, and calculating the rate of change of the risk data to determine the slope of the trend: Where ΔX i( t ) represents the rate of change at time t; X i( t ) represents the raw data uploaded by the robot edge; Δt represents the time interval; the change rate threshold δ of the environmental data is set according to the historical data X , when ΔX i( t ) ≥δ X When ΔX i( t ) The corresponding data grew too fast, and the central end issued an early warning.

7. The fast-response inspection robot centralized control method according to claim 6, characterized in that: The data model also includes, when there is robot data in the risk data uploaded by the robot edge, retrieving environmental data for collaborative analysis based on the robot data; When the robot's temperature data is risk data, the ambient temperature data at the same time point is retrieved. If the ambient temperature data is normal data and the temperature of the inspection robot continues to rise over time, an abnormal temperature alarm for the inspection robot is issued, the inspection task of the corresponding inspection robot is stopped, and staff are dispatched for inspection and maintenance. If the ambient temperature T env If the data is risk data, the temperature of the inspection robot is predicted by the linear model And with real-time data T robot For comparison, it is expressed as: Among them, α1 represents the linear increase; β1 represents the coefficient; ΔT represents the difference between the real-time temperature and the predicted value; When the voltage and current data of the robot are risk data, the ambient humidity data at the same time point is retrieved. If the ambient humidity data is normal data, an inspection robot hardware failure alarm is issued, the inspection task of the corresponding inspection robot is stopped, and staff are dispatched for inspection and maintenance. If the ambient humidity is H env If it is risk data, the Kalman filter is used to predict the impact of environmental humidity on the robot's current and voltage, and early warning is given through state estimation and measurement update, which is expressed as: in, represents the predicted current; Represents the predicted voltage; A, B, C, D represent the state matrix of Kalman filter.

8. A fast-response inspection robot centralized control system using the method as claimed in any one of claims 1 to 7, characterized in that: include, The data acquisition module collects environmental data and robot data through sensors, pre-processes the data at the edge of the robot, and uploads the pre-processed data to the central end; Walking drive module: The central end calculates the initial optimal path through the A* algorithm and sends it to the inspection robot to perform the inspection task. The edge end of the robot automatically avoids obstacles during the inspection. When the obstacle cannot be avoided, it sends a signal to the central end to recalculate the optimal path. In the early warning module, the central end analyzes the trend of environmental data based on the uploaded data, and conducts collaborative analysis and early warning of robot data and environmental data to achieve centralized control of inspection robots.

9. A computing device comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Images