Robot patrol inspection control method and device, equipment and storage medium

Abstract

The application provides a robot inspection control method and device, equipment and a storage medium, and belongs to the technical field of inspection control. The method comprises the following steps: obtaining a region to be inspected, and dividing the region to be inspected into a plurality of target inspection regions; for each target inspection region, performing a target inspection operation, wherein the target inspection operation comprises the following steps: acquiring environmental data of a target robot during an inspection process in the target inspection region; determining a potential value corresponding to each environmental data according to the environmental data, and determining a dynamic environmental potential matrix of the target inspection region according to the potential values; determining an inspection path in the target inspection region based on the dynamic environmental potential matrix and through a biological foraging algorithm, and then controlling the target robot to perform inspection according to the inspection path. The robot inspection control method and device, equipment and storage medium provided by the application can realize adaptive environmental inspection and improve the ability of the robot to cope with complex environments during the inspection process.

Description

Technical Field

[0001] This application belongs to the field of inspection control technology, and more specifically, relates to a robot inspection control method, device, equipment, and storage medium. Background Technology

[0002] Industrial inspection is a crucial link in ensuring production safety and normal equipment operation. It can promptly detect potential equipment malfunctions, prevent accidents, and reduce economic losses. With the continuous improvement of industrial intelligence, inspection work is also gradually developing from traditional manual inspection to automation and intelligence.

[0003] Currently, robotic inspection technology has been applied in various fields such as power and rail transportation. Existing robotic inspection systems mainly rely on pre-planned fixed paths to carry out inspection tasks, collecting environmental data and equipment status information by carrying various sensors.

[0004] However, existing robotic inspection systems have significant limitations. Most systems fix the inspection path before performing the inspection task. This approach cannot adapt to complex and ever-changing dynamic environments. When temporary obstacles appear or environmental conditions change, the robot often cannot respond effectively and can only stop operating and wait for human intervention. Summary of the Invention

[0005] The purpose of this application is to provide a robot inspection control method, device, equipment, and storage medium to achieve adaptive environmental inspection and improve the robot's ability to cope with complex environments during inspection.

[0006] A first aspect of this application provides a robot inspection control method, including:

[0007] Obtain the area to be inspected and divide it into multiple target inspection areas;

[0008] For each target inspection area, perform the target inspection operation;

[0009] The target inspection operation includes:

[0010] Acquire environmental data during the inspection process of the target robot in the target inspection area. The environmental data includes temperature data, obstacle height data, and signal strength data. The signal strength data is the signal strength between the target robot and the control device for information exchange.

[0011] Based on the environmental data, determine the potential energy value corresponding to each environmental data point, and then determine the dynamic environmental potential energy matrix of the target inspection area based on the potential energy value corresponding to each environmental data point.

[0012] Based on the dynamic environmental potential energy matrix and using a biological foraging algorithm, the inspection path within the target inspection area is determined.

[0013] The target robot is controlled to perform inspections within the target inspection area based on the inspection path.

[0014] A second aspect of this application provides a robot inspection control device, comprising:

[0015] The target inspection area determination module is used to obtain the area to be inspected and divide the area to be inspected into multiple target inspection areas.

[0016] The inspection control module is used to execute target inspection operations for each target inspection area;

[0017] Specifically, when executing target inspection operations, the inspection control module is used for:

[0018] Acquire environmental data during the inspection process of the target robot in the target inspection area. The environmental data includes temperature data, obstacle height data, and signal strength data. The signal strength data is the signal strength between the target robot and the control device for information exchange.

[0019] Based on the environmental data, determine the potential energy value corresponding to each environmental data point, and then determine the dynamic environmental potential energy matrix of the target inspection area based on the potential energy value corresponding to each environmental data point.

[0020] Based on the dynamic environmental potential energy matrix and using a biological foraging algorithm, the inspection path within the target inspection area is determined.

[0021] The target robot is controlled to perform inspections within the target inspection area based on the inspection path.

[0022] A third aspect of this application provides a control device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the robot inspection control method described above.

[0023] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the robot inspection control method described above.

[0024] The beneficial effects of the robot inspection control method, device, equipment, and storage medium provided in this application embodiment are as follows: This application embodiment acquires environmental data of the target robot before inspection and constructs a dynamic environmental potential energy matrix based on the acquired environmental data. Based on this dynamic environmental potential energy matrix, the inspection path of the target robot is planned through a biological foraging algorithm. The dynamic environmental potential energy matrix of this embodiment integrates the potential energy information corresponding to each environmental data (temperature data, obstacle height data, and signal strength data), quantifies the environmental characteristics of different locations in the target inspection area in matrix form, and the biological foraging algorithm can plan the optimal inspection path based on the dynamic environmental potential energy matrix information constructed from the environmental data. This avoids problems such as detours, repeated inspections, or omissions of key areas caused by data errors, reduces the risk of collisions during the inspection process, avoids interruption of the inspection task due to sudden environmental changes, improves the adaptability of the target robot to complex environments during inspection, and improves the efficiency of inspection work. Furthermore, temperature data can accurately reflect the situation around the equipment, which helps to detect abnormalities such as overheating of the equipment in a timely manner, and thus determine whether there are potential faults in the equipment. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 A flowchart illustrating a robot inspection control method provided in an embodiment of this application;

[0027] Figure 2 A flowchart illustrating a target inspection operation method provided in an embodiment of this application;

[0028] Figure 3 A schematic diagram illustrating the division of a target inspection area is provided in one embodiment of this application;

[0029] Figure 4 This is a schematic diagram illustrating another way of dividing a target inspection area according to an embodiment of this application;

[0030] Figure 5 This is a schematic diagram of the temperature anomaly potential energy value change curve provided in an embodiment of this application;

[0031] Figure 6 This is a schematic diagram illustrating the determination of candidate paths during robot inspection, provided in an embodiment of this application.

[0032] Figure 7This is a structural block diagram of a robot inspection control device provided in one embodiment of this application;

[0033] Figure 8 This is a schematic block diagram of a control device provided in an embodiment of this application. Detailed Implementation

[0034] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0035] It is understood that in the embodiments of this application, data such as user information are involved. When the embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with relevant laws, regulations and standards.

[0036] It should be noted that the terms "first," "second," etc., used in the specification, claims, and drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in sequences other than those illustrated or described herein.

[0037] To make the objectives, technical solutions, and advantages of this application clearer, the following description will be provided in conjunction with the accompanying drawings and specific embodiments.

[0038] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a robot inspection control method according to an embodiment of this application. The robot inspection control method provided in this embodiment is executed by a control device and can be applied to scenarios such as wind power plants and substations. The method may include:

[0039] S101: Obtain the area to be inspected and divide it into multiple target inspection areas.

[0040] In this embodiment, the location information of the area to be inspected, input by the user, is obtained, and the area to be inspected is determined based on the location information input by the user. Alternatively, the area drawn by the user on the map of the inspection area is used as the area to be inspected, or a pre-defined area is used as the area to be inspected. Furthermore, after obtaining the area to be inspected, it is divided into multiple target inspection areas.

[0041] The process of dividing the area to be inspected into multiple target inspection areas includes: acquiring equipment-related data for the area to be inspected and standardizing the equipment-related data; and determining the complexity index of the area to be inspected based on the standardized equipment-related data and the area complexity index calculation formula.

[0042] The formula for calculating the regional complexity index is as follows: ,in, For standardized equipment density data, n The standardized number of equipment types, f This is standardized historical anomaly frequency data. These are the dynamic weights corresponding to equipment density data, the number of equipment types, and historical anomaly frequency data, respectively.

[0043] Based on the above embodiments, the area to be inspected is divided into multiple target inspection areas. Specifically, this may include: if the complexity index exceeds a preset threshold, the area to be inspected is divided into multiple sub-areas, and each sub-area is used as a target inspection area. If the complexity index does not exceed the preset threshold, the area to be inspected is used as a target inspection area.

[0044] In this embodiment, the equipment-related data for the area to be inspected includes historical fault frequency data, equipment density data, and the number of equipment types. The historical fault frequency data, equipment density data, and the number of equipment types for each device in the area to be inspected are obtained through the control equipment in that area.

[0045] Based on the number of times each piece of equipment in the area to be inspected failed within a historical time period and the corresponding equipment runtime, determine the historical failure frequency data for each piece of equipment within a unit runtime. Determine the equipment density data based on the number of devices per unit area within the area to be inspected. Determine the number of different types of equipment within the area to be inspected.

[0046] The historical fault frequency data, equipment density data, and equipment type quantity are standardized using a ratio method. For historical fault frequency data, taking last month's equipment fault frequency data as an example, the standardized historical fault frequency data is determined by the ratio of the historical fault frequency data to the preset fault frequency standard value, based on the preset fault frequency standard value set for each equipment. For example, if the preset fault frequency standard value is set to 5 times / month, and the acquired historical fault frequency data is 3 times / month, then the standardized historical fault frequency data is 0.6. For equipment density data, a preset density standard value is set, and the standardized equipment density data is determined by the ratio of the current equipment density data to the preset density standard value. For example, if the preset density standard value is 4 units / m², the standardized equipment density data is determined by the ratio of the current equipment density data to the preset density standard value. 3 The current equipment density data obtained is 0.4 units / m².3 If the number of equipment types is 5, and the number of equipment types in the current target inspection area is 6, then the standardized equipment density data is 0.1. For the number of equipment types, a preset standard value is set, and the standardized number of equipment types is determined based on the ratio of the number of equipment types in the current target inspection area to the preset standard value. For example, if the preset standard value is 5, and the number of equipment types in the current target inspection area is 6, then the standardized number of equipment types is 1.2.

[0047] The complexity index of the area to be inspected is determined based on the standardized equipment data, the corresponding dynamic weights, and the formula for calculating the area complexity index.

[0048] In this embodiment, each dynamic weight can be determined based on actual inspection needs, historical experience, or the analytic hierarchy process, and is used to characterize the importance of data related to different devices in measuring regional complexity.

[0049] The complexity index within the inspection area is compared with a preset threshold. If the complexity index exceeds the preset threshold, it indicates that the complexity of the area to be inspected is high. The area to be inspected is then divided into multiple sub-areas, with each sub-area serving as a target inspection area. In this embodiment, when dividing the area to be inspected, factors such as the geographical boundaries of the area and the distribution characteristics of equipment within the area can be considered to divide it into multiple sub-areas, with each sub-area serving as a target inspection area. This allows for more targeted inspection operations to be performed on each target inspection area subsequently.

[0050] If the complexity index does not exceed the preset threshold, it indicates that the complexity of the area to be inspected is low, and there is no need to divide it into sub-regions. The area to be inspected can be directly used as the target inspection area.

[0051] S102: Perform target inspection operations for each target inspection area.

[0052] In this embodiment, a target inspection operation is performed for each target inspection area to ultimately complete the inspection of the area to be inspected. In this embodiment, the target inspection operation can be performed on each target inspection area at different time periods, or the target robot can perform the target inspection operation on each target inspection area separately according to certain rules, or different target robots can perform the target inspection operation on different target inspection areas.

[0053] Specifically, such as Figure 2 As shown, for each target inspection area, the target inspection operation is performed in step S102, which may include steps Sa-Sd, where,

[0054] Sa: Acquire environmental data of the target robot during its inspection process within the target inspection area.

[0055] In this embodiment, the environmental data includes multiple sets of data collected from different locations. Each set of data includes: temperature data, obstacle height data, and signal strength data. The signal strength data represents the signal strength used for information exchange between the target robot and the control device. Temperature sensors, LiDAR, and cameras are installed in the target inspection area.

[0056] This embodiment acquires temperature data using a temperature sensor. A laser beam emitted by a lidar is used to determine the distance and height information between obstacles and the target robot based on the reflection time and angle of the laser beam. Depth information in the scene to be inspected is acquired by a camera, and a 3D point cloud map is determined based on this depth information. The height data of obstacles is then determined from this 3D point cloud map. Alternatively, the obstacle height data within the target inspection area can be based on the previous inspection measurements by the target robot, or the obstacle height data can be pre-set. A signal detection device is installed within the target inspection area to acquire signal strength data during information interaction between the target robot and the control device.

[0057] In this embodiment, after acquiring the environmental data of the target robot before it inspects the target inspection area, the timestamp corresponding to each data in each data set is determined. The timestamp of a data represents the time of data collection, and the data in the set is aligned in the time dimension based on the timestamp.

[0058] Specifically, the acquisition time of temperature data, obstacle height data, and signal strength data in each set of environmental data is recorded, generating corresponding timestamps. A suitable time point is selected from the timestamps of each data set as the reference time. In this embodiment, the earliest timestamp is selected, but other representative timestamps can also be selected according to specific needs. For each data timestamp, the difference between it and the reference time is calculated to obtain the time offset corresponding to each data. Based on this time offset, the corresponding data is adjusted in time to align the data set in the time dimension, obtaining the time-aligned data set. Based on the time-aligned data sets, the potential energy value groups corresponding to each data set are determined, constructing a dynamic environmental potential energy matrix for the target inspection area. Each potential energy value group includes the potential energy values ​​corresponding to the temperature data, obstacle height data, and signal strength data.

[0059] Sb: Determine the potential energy value corresponding to each environmental data point based on the environmental data, and determine the dynamic environmental potential energy matrix of the target inspection area based on the potential energy value corresponding to each environmental data point.

[0060] To enable the target robot to better adapt to the complexity of the target inspection area, this embodiment divides the target inspection area into multiple target grid cells before determining the potential energy values ​​corresponding to each environmental data point. Each target grid cell contains a data collection location for collecting environmental data. The division can be based on a grid map of the target inspection area or on the regional characteristics of the target inspection area to obtain multiple target grid cells.

[0061] This embodiment can acquire environmental information in the target inspection area based on LiDAR and visual sensors, and determine a grid map in the target inspection area based on this environmental information. This grid map divides the environment corresponding to the target inspection area into regular grids, specifically as follows: Figure 3 As shown.

[0062] In one possible implementation, each grid cell can represent a different state (that is not specified in the attribute tag). Figure 3 (This is reflected in the grid). For example, the state of a grid can be represented by a specific value; a value of 0 can represent an idle state, and a value of 1 can represent a state with an obstacle. Depending on the inspection needs, several adjacent grids can be merged into a single target grid unit. In this embodiment, the grid size is 1m × 1m, and four adjacent grids are merged into one target grid unit, meaning the target grid unit in this embodiment is 2m × 2m in size. In this embodiment, when merging multiple grids into a single target grid unit, the overall environmental characteristics of the target grid unit can be reflected by the state of each grid, improving the robot's adaptability to the complex environment of the target inspection area.

[0063] Within each target grid cell, a suitable point location is selected as the data acquisition location based on the environmental information in the raster map. For example, a grid cell with no obstacles and a relatively open field of view can be selected as the data acquisition location.

[0064] In this embodiment, the target inspection area can also be divided according to its regional characteristics. Specifically, this embodiment can divide the target inspection area based on terrain undulations and differences in obstacle distribution. Specifically, this embodiment can extract features of the target inspection area using image recognition technology. These features may include terrain undulations, building distribution features, and road orientation features. Based on the extracted features, the target inspection area is divided into different sub-regions. The resulting grid cells may be irregular, such as... Figure 4 As shown, Figure 4 The image only shows the target inspection area being divided into irregular sub-regions.

[0065] In one possible embodiment, larger grid cells can be used for areas with flat terrain and no obvious obstacles, while smaller grid cells can be used for areas with complex terrain or many obstacles.

[0066] For each target grid cell, the acquisition location is determined based on the corresponding regional characteristics. For example, in densely built-up areas, the acquisition location is set in a position where buildings can be observed well; in areas with roads, the acquisition location is set in the center of the road or near the roadside for easy detection of road surface conditions.

[0067] This embodiment determines the potential energy value corresponding to each environmental data point based on the environmental data, and determines the dynamic environmental potential energy matrix of the target inspection area based on the potential energy value corresponding to each environmental data point. Specifically, it includes:

[0068] Based on the time-aligned data sets, determine the potential energy value groups corresponding to each data set. Each potential energy value group includes the potential energy values ​​corresponding to temperature data, obstacle height data, and signal strength data. Construct the dynamic environmental potential energy matrix of the target inspection area based on the potential energy value groups corresponding to each data set.

[0069] In this embodiment, for each group of data after time alignment, features are extracted from each data in the group to obtain the potential energy value corresponding to each data, and each potential energy value is normalized accordingly.

[0070] Determine the weights corresponding to each type of environmental data;

[0071] For each target grid cell, the potential energy value of the target grid cell is determined based on the normalized potential energy value corresponding to the target grid cell and the weights corresponding to each environmental data.

[0072] The dynamic environmental potential energy matrix of the target inspection area is constructed based on the potential energy value corresponding to each target grid cell.

[0073] This embodiment establishes a mapping relationship between temperature and potential energy: the greater the temperature data deviates from the standard temperature range, the higher the corresponding potential energy value; it also establishes a mapping relationship between obstacle height and potential energy: the higher the obstacle, the greater the obstruction to the target robot's inspection, and thus the lower the corresponding potential energy value; finally, it establishes a mapping relationship between signal strength data and potential energy: signal strength data is related to communication quality; the stronger the signal, the better the communication quality, and the higher the corresponding potential energy value; the weaker the signal, the worse the communication quality, and the lower the corresponding potential energy value.

[0074] This embodiment extracts features from each data point in the dataset to obtain the potential energy value corresponding to each data point. Specifically, the standard temperature range is set to 25℃-30℃ based on historical environmental data. If the temperature data is higher or lower than the standard temperature range in the historical environmental data, the temperature data is converted into a temperature anomaly potential energy value. The spatial obstruction coefficient of the obstacle is determined based on the ratio of the obstacle height data to a preset height threshold. The obstacle potential energy value is then determined based on the spatial obstruction coefficient and the obstacle dynamic risk coefficient.

[0075] For example, in this embodiment, the method for determining the abnormal temperature potential energy value is as follows: the temperature data in the data set is compared with the standard temperature range. If the temperature data is higher than 30℃ or lower than 25℃, the temperature data is marked as abnormal temperature data. The abnormal temperature data is mapped to an initial abnormal temperature potential energy value according to a preset temperature range. For example, if the temperature range is 30℃ < T ≤ 35℃, the corresponding initial abnormal temperature potential energy value can be set to 4; if the temperature range is 35℃ < T ≤ 40℃, the corresponding initial abnormal temperature potential energy value can be set to 7; if the temperature range is T > 40℃, the corresponding initial abnormal temperature potential energy value can be set to 10. For cases where the temperature data is lower than 25℃, if the temperature range is 20℃ < T ≤ 25℃, the corresponding initial abnormal temperature potential energy value can be set to 3; if the temperature range is 15℃ < T ≤ 20℃, the corresponding initial abnormal temperature potential energy value can be set to 6; and if the temperature range is T ≤ 15℃, the corresponding initial abnormal temperature potential energy value can be set to 9. The initial temperature anomaly potential energy value is linearly normalized to determine the temperature anomaly potential energy value. The relationship between the temperature range and the initial temperature anomaly potential energy value can be preset.

[0076] In this embodiment, the spatial obstruction coefficient of an obstacle is determined by setting a preset height threshold. H t This preset height threshold can be determined based on factors such as the target robot's size and mobility. It is based on the acquired obstacle height data. H With preset height threshold H t Determine the spatial obstruction coefficient. The spatial obstruction coefficient of obstacles can be obtained using a linear mapping method, based on the acquired obstacle height data. H With preset height threshold H t Determine the height ratio r , The spatial obstruction coefficient of the obstacle is determined based on this height ratio. At this point, the spatial obstruction coefficient... M =height ratio r Alternatively, a piecewise function can be used to determine the spatial obstruction coefficient of the obstacle. H ≤ Ht If the height of the obstacle within the target inspection area does not obstruct the robot's inspection, then the spatial obstruction coefficient is determined to be zero. M =0; if H > H t If the height of an obstacle within the target inspection area obstructs the robot's inspection, then the spatial obstruction coefficient is determined. ,in, H max The maximum obstacle height within the target inspection area. The formula for the obstacle potential energy value in this embodiment is: ,in, E O This represents the potential energy of the obstacle. D The obstacle dynamic risk coefficient is used to characterize the degree of risk posed by obstacles to the target robot's inspection. This coefficient can be determined based on historical inspection data. The obstacle potential energy value is then normalized.

[0077] This embodiment sets an ideal value for signal strength, which can be determined based on the actual area conditions or historical experience. The signal deviation is determined by comparing the actual signal strength with the ideal value. This deviation is then converted into an information potential energy value, which is normalized. A larger signal deviation may result in a higher potential energy value.

[0078] In this embodiment, the signal deviation is determined using a deviation degree formula and based on the actual signal strength value and the ideal signal strength value. The deviation degree formula in this embodiment is as follows: ,in, This represents the deviation between the actual and ideal values ​​of signal strength, also known as signal deviation as described above. | represents the absolute value. This represents the actual value of the signal strength. The ideal value of the signal strength is denoted as , where . ≠0.

[0079] Historical environmental data of the target inspection area is acquired. Based on this data, the impact of each environmental factor on equipment failures within the target inspection area is determined, and the weights of each environmental factor are adjusted. For example, for temperature data, other environmental data are kept constant, and the failure rates corresponding to temperatures above 40℃ and below 40℃ are determined. The correlation between this environmental data and failures is determined using the Pearson correlation coefficient, and the weight of the temperature data is adjusted accordingly. This adjustment can be achieved by: if the correlation is higher than a preset correlation threshold, it indicates a higher impact of temperature data on equipment failures, and the weight of the temperature data is increased.

[0080] In this embodiment, the target inspection area is divided into multiple target grid cells, each corresponding to a data acquisition location. The target grid cells can be set according to the size and inspection accuracy of the target inspection area. For each target grid cell, the potential energy value corresponding to each environmental data is acquired at the acquisition location of that target grid cell. The potential energy values ​​are then normalized to obtain the normalized potential energy value corresponding to the location of that target grid cell. Based on the normalized potential energy value corresponding to the location of that target grid cell and the weights corresponding to each environmental data, the potential energy value of that target grid cell is determined.

[0081] This embodiment establishes the mapping relationships between temperature, obstacle height, and signal strength and potential energy values, respectively. This allows for the representation of the impact of environmental data on equipment malfunctions from different dimensions. For example, when the temperature data exceeds the standard temperature range, it indicates that the corresponding equipment may be faulty. In this case, path planning can be prioritized for target grid cells with higher temperatures, enabling earlier detection of potential equipment malfunctions, timely elimination of equipment failure risks, and prevention of further escalation of the fault. By determining the degree of influence of each environmental data on equipment malfunctions through historical environmental data and adjusting the corresponding weights, the weights of each environmental data are determined based on the magnitude of their impact on inspection (equipment malfunction related) when calculating the potential energy of the target grid cell. This ensures that the target robot can better adapt to complex environments and reduce losses caused by equipment malfunctions.

[0082] Based on the arrangement order of the target grid cells within the target inspection area, the potential energy value of each target grid cell is sequentially filled into the corresponding position of the dynamic environment potential energy matrix, thereby determining the dynamic environment potential energy matrix of the target inspection area. The rows of this dynamic environment potential energy matrix correspond one-to-one with the row numbers in the target grid cells, and the columns of this dynamic environment potential energy matrix correspond one-to-one with the column numbers in the target grid cells. For example, for the column numbered (… i , j After obtaining the potential energy value of the target mesh cell, this potential energy value is used as the first element in the dynamic environment potential energy matrix. i Line number j Column position.

[0083] In one possible embodiment, the target inspection area is divided into several sub-regions based on its regional characteristics. As can be seen from the above, these sub-regions might be as follows: Figure 4 For the irregular region shown, for each sub-region, the potential energy value corresponding to each environmental data point is collected at the collection location within that sub-region. Each potential energy value is normalized to obtain the normalized potential energy value corresponding to the target grid cell location. Based on the normalized potential energy value corresponding to the target grid cell location and the weights corresponding to each environmental data point, the potential energy value of the target grid cell is determined.

[0084] For example Figure 4 Each sub-region shown is indexed to determine the mapping relationship between the sub-region and the dynamic potential energy matrix. For example, the sub-regions include the first to the seventh sub-regions, corresponding to... Figure 4 As shown in the diagram, z1-z7 represent the following sub-regions: z1 corresponds to position (1,1) of the dynamic environment potential energy matrix; z2 corresponds to position (1,2); z3 corresponds to position (2,1); z4 corresponds to position (2,2); z5 corresponds to position (3,1); z6 corresponds to position (3,2); and z7 corresponds to position (4,1). Based on this mapping, the potential energy values ​​of each sub-region are sequentially filled into the corresponding positions in the dynamic environment potential energy matrix. For position (4,2), since there is no corresponding sub-region in this embodiment, the potential energy value at that position is padded with zeros, thus determining the four-row, two-column dynamic environment potential energy matrix.

[0085] Sc: Based on the dynamic environmental potential energy matrix and using a biological foraging algorithm, the inspection path within the target inspection area is determined.

[0086] In this embodiment, the inspection path within the target inspection area is determined based on the dynamic environmental potential energy matrix and a biological foraging algorithm, specifically including:

[0087] A candidate path set is determined based on a biological foraging algorithm. The candidate path set contains multiple candidate paths, and each candidate path is composed of multiple target grid cells.

[0088] Obtain the fitness value corresponding to each candidate path;

[0089] By copying candidate paths with fitness values ​​higher than the first preset fitness value and eliminating candidate paths with fitness values ​​lower than the second preset fitness value, the adjusted candidate path set is determined.

[0090] Based on the adjusted candidate path set, the optimal candidate path is determined through a path intersection mechanism. This optimal candidate path is the candidate path with the highest fitness value.

[0091] The optimal candidate path is used as the inspection path within the target inspection area;

[0092] The methods for obtaining the fitness value of each candidate path include:

[0093] Determine the potential energy value corresponding to the candidate path in the dynamic environment potential energy matrix;

[0094] The cost of the candidate path is determined based on the potential energy value;

[0095] The fitness value of a candidate path is determined based on its cost.

[0096] In this embodiment, the basic parameters of the biological foraging algorithm are first determined, such as the number of candidate paths (population size) and the grid range (search space) of the target inspection area covered by the dynamic environmental potential energy matrix. Within the search space corresponding to the dynamic environmental potential energy matrix, multiple candidate paths are randomly generated, forming a candidate path set. Each candidate path in the set is composed of multiple target grid cells connected in a certain order, which can simulate the possible paths the target robot might take during inspection in the target inspection area, ensuring that the target grid cells are within the target inspection area and that the construction of the candidate paths meets the actual inspection requirements.

[0097] For each candidate path, the potential energy value corresponding to each target grid cell is obtained from the dynamic environment potential energy matrix, and these potential energy values ​​are summed to obtain the potential energy value corresponding to the candidate path.

[0098] The relationship between cost and the potential energy value of a candidate path can be established by setting the cost equal to the potential energy value, or by determining the cost of a candidate path using a cost calculation formula. In this embodiment, the cost of the candidate path is determined using its potential energy value and a cost calculation formula, which is as follows: ,in, k 1 represents a preset proportional coefficient, set based on empirical values. Different candidate paths correspond to the same coefficient. k 1, E This represents the potential energy value of the candidate path. Cost The cost of the candidate path represents the direct correlation between the cost and the potential energy value of the candidate path. In this embodiment, the higher the potential energy value of the candidate path, the higher the corresponding cost, indicating that the candidate path is less conducive to the target robot's inspection.

[0099] In this embodiment, the fitness value of a candidate path is set to be inversely proportional to its cost. The fitness value of a candidate path is determined according to the formula for calculating the fitness value of a candidate path, which is as follows: ,in, F The fitness value of the candidate path. This is the value after standardizing the cost. The value is a preset, extremely small positive number, which can be 0.001, to avoid the denominator of the fitness value calculation formula being zero. In this embodiment, the lower the cost of the candidate path, the higher the corresponding fitness value, indicating that the candidate path is better.

[0100] A first preset fitness value and a second preset fitness value are set, wherein the first preset fitness value is greater than the second preset fitness value. Candidate paths with fitness values ​​higher than the first preset fitness value are copied from the candidate path set to increase the number of high-quality candidate paths in subsequent iterations; candidate paths with fitness values ​​lower than the second preset fitness value are eliminated to reduce the interference of low-quality candidate paths on the search process. The candidate path set is adjusted based on the copying and elimination operations.

[0101] This embodiment performs a path crossing operation on the candidate paths in the adjusted candidate path set. For example, two or more candidate paths can be selected, and their path segments (sequences of grid cells) can be exchanged at certain intersection points to generate new candidate paths. The fitness values ​​of the new candidate paths obtained after the crossing and the fitness values ​​of the candidate paths that did not participate in the crossing operation are obtained, and the candidate path with the highest fitness value is selected as the current optimal candidate path.

[0102] The preset termination condition in this embodiment can be that the number of iterations reaches a set maximum value, or that the fitness value of the optimal candidate path no longer improves significantly after multiple consecutive iterations (the improvement range of the fitness value of the optimal candidate path is within a preset range).

[0103] If the preset termination condition is met, the iteration stops, and the candidate path with the highest fitness value in the adjusted candidate path set is taken as the inspection path in the target inspection area; if the preset termination condition is not met, the fitness value calculation and adjustment of the current candidate path set continues, and the next iteration begins.

[0104] Sd: Controls the target robot to perform inspections within the target inspection area based on the inspection path.

[0105] In this embodiment, the optimal inspection path (represented by a sequence of target grid cells) obtained by the biological foraging algorithm is converted into motion commands for the target robot, such as the direction and distance of each path segment, so as to control the target robot to perform inspections within the target inspection area.

[0106] In one possible embodiment, this embodiment employs a bacterial foraging optimization algorithm, which is specifically implemented as follows:

[0107] Candidate paths are mapped to bacterial individuals in the algorithm, with each bacterial individual's position corresponding to a candidate path composed of a sequence of target grid cells. In this embodiment, the population size can be set to 30.

[0108] The fitness value of a candidate path is determined based on its cost.

[0109] For each bacterial individual, the starting point of the current candidate path is used as the starting point of the bacterial individual. An adjacent non-high-risk target grid cell (the potential energy value of the target grid cell is less than the preset risk threshold) is randomly selected as the direction of movement. During the movement of the bacterial individual, the starting point and ending point of the candidate path remain unchanged. The step size of each movement of the bacterial individual can be set to 1 target grid cell. A new candidate path is determined, and the fitness value of the new candidate path is calculated. If the fitness value of the new candidate path is greater than the fitness value of the original candidate path, the new candidate path is retained. If the fitness value of the new candidate path is not greater than the fitness value of the original candidate path, a new path is randomly generated in the neighborhood with the current position of the bacterial individual as the starting point. The random movement continues until the fitness value of the determined new candidate path is greater than the fitness value of the original candidate path, or the preset maximum number of random movements is reached. The preset maximum number of random movements can be set to 5 to avoid getting trapped in a local loop. In this embodiment, if after reaching the preset maximum number of random moves, the fitness value of the new candidate path is still not greater than the fitness value of the original path, then the candidate path with the largest fitness value among the five new candidate paths can be selected as the new candidate path. In this embodiment, if all adjacent grids at the current position are high-risk (the potential energy value of the target grid cell is ≥ a preset risk threshold), then the grid with the lowest high-risk value among the five new candidate paths can be temporarily selected to prevent deadlock where movement is impossible.

[0110] All bacterial individuals are sorted in descending order of fitness value, and the top 50% of high-fitness bacterial individuals are retained. In this embodiment, the top 15 bacterial individuals in descending order are retained. The retained high-fitness bacterial individuals are replicated, with each high-fitness bacterial individual splitting and replicating once. The bottom 50% of low-fitness individuals are eliminated to ensure that the population size remains unchanged.

[0111] Select the bacterial individual with the highest fitness value from the remaining bacterial individuals, and use the candidate path corresponding to this bacterial individual as the optimal candidate path.

[0112] The algorithm terminates and outputs the optimal inspection path when any of the following preset conditions are met:

[0113] The algorithm terminates when the number of iterations reaches a preset value (e.g., 20 times).

[0114] If the fluctuation value of the optimal fitness value calculated in each of the last 5 iterations is less than 0.005, then the algorithm is considered to have converged to the global optimum and terminated. The fluctuation value can be determined based on the optimal fitness value corresponding to the current iteration number and the optimal fitness value 5 iterations ago. The formula for calculating the fluctuation value is as follows: b is the fluctuation value. For the first kThe optimal fitness value in the next iteration. For the first k - The optimal fitness value after 4 iterations.

[0115] As can be seen from the above, when the target robot in this embodiment inspects the target inspection area, it acquires environmental data and determines the potential energy value corresponding to each environmental data point based on the acquired environmental data, thereby constructing a dynamic environmental potential energy matrix. This dynamic environmental potential energy matrix can dynamically update each potential energy value in the matrix according to environmental changes, and the optimal inspection path of the target robot is determined based on the dynamic environmental potential energy value. This embodiment can also update the dynamic environmental potential energy matrix according to environmental changes. When environmental changes occur in the target inspection area, such as the appearance of temporary obstacles, the target robot can determine a new dynamic environmental potential energy matrix based on the new environmental data, and then re-plan the optimal path based on the new dynamic environmental potential energy matrix and the biological foraging algorithm, thereby improving its adaptability to complex environments. Since the target robot can adapt to changing environmental data, no human intervention is required, thus ensuring the continuity of inspection work and improving inspection efficiency.

[0116] In one embodiment of this application, after determining the potential energy value corresponding to the candidate path in the dynamic environment potential energy matrix, the method further includes:

[0117] Based on the potential energy values ​​corresponding to each target grid cell traversed by the candidate path, a multidimensional change curve corresponding to the candidate path is determined. The multidimensional change curve includes: temperature anomaly potential energy value change curve, obstacle potential energy value change curve, and information potential energy value change curve. Each dimension change curve is determined based on the potential energy value corresponding to each target grid cell in that dimension and is used to characterize the change in potential energy value corresponding to each target grid cell traversed by the candidate path in that dimension.

[0118] Target grid cells that meet at least one of the preset conditions are identified as high-risk grid cells;

[0119] If there is a preset number of high-risk grid cells in the candidate path, then a high-risk path segment is determined. The preset number is greater than a preset number threshold. The high-risk path segment consists of a preset number of high-risk grid cells.

[0120] The alternative path is determined based on the local avoidance strategy, and a new candidate path is formed based on the alternative path;

[0121] The cost of determining the candidate path based on the potential energy value includes:

[0122] The cost of the new candidate path is determined based on the potential energy value;

[0123] The preset conditions include:

[0124] The potential energy value of temperature anomaly is greater than or equal to the temperature anomaly risk threshold.

[0125] The obstacle's potential energy value is greater than or equal to the obstacle risk threshold;

[0126] The information potential value is less than or equal to the signal risk threshold;

[0127] The rate of change of the change curve in any dimension between two adjacent target grid cells is greater than or equal to a preset rate of change threshold.

[0128] In this embodiment, for each target grid cell traversed by the candidate path, the temperature anomaly potential energy value, obstacle potential energy value, and information potential energy value corresponding to each target grid cell are extracted from the dynamic environment potential energy matrix. Based on the temperature anomaly potential energy value, obstacle potential energy value, and information potential energy value corresponding to each target grid cell, a multi-dimensional change curve corresponding to the candidate path is determined. Each dimension change curve is determined based on the potential energy value corresponding to that dimension for each target grid cell, and is used to characterize the change in potential energy value corresponding to each target grid cell traversed by the candidate path in that dimension. The multi-dimensional change curve includes the temperature anomaly potential energy value change curve, the obstacle potential energy value change curve, and the information potential energy value change curve.

[0129] Specifically, the multi-dimensional change curves are determined as follows: The temperature anomaly potential energy change curve is plotted with the order of the target grid cells in the candidate path on the horizontal axis and the temperature anomaly potential energy value on the vertical axis, reflecting the potential energy changes in the temperature dimension along the path. The obstacle potential energy change curve is plotted with the order of the grid cells on the horizontal axis and the comprehensive obstacle potential energy value on the vertical axis. The information potential energy change curve is plotted with the order of the grid cells on the horizontal axis and the information potential energy value on the vertical axis.

[0130] For each target grid cell, whether it is a risk grid cell is determined based on at least one preset condition. In this embodiment, the preset conditions are: the temperature anomaly potential energy value is greater than or equal to the temperature anomaly risk threshold; the obstacle potential energy value is greater than or equal to the obstacle risk threshold; the information potential energy value is less than or equal to the signal risk threshold; and the rate of change of the change curve of any dimension (temperature, obstacle height, information intensity) between two adjacent target grid cells is greater than or equal to a preset rate of change threshold. This rate of change can be determined by dividing the potential energy difference between adjacent target grid cells by the distance between the adjacent target grid cells. The temperature anomaly risk threshold, obstacle risk threshold, signal risk threshold, and preset rate of change threshold are all pre-set.

[0131] If the target grid cell meets at least one of the above preset conditions, the target grid cell will be identified as a high-risk grid cell.

[0132] Based on the above embodiments, it is checked whether there is a continuous preset number of high-risk grid cells in the candidate path. If there is a continuous preset number of high-risk grid cells and the preset number is greater than the preset number threshold, the path segment composed of these continuous high-risk grid cells is determined as a high-risk path segment.

[0133] For candidate paths with high-risk segments, a local avoidance strategy is adopted to plan alternative paths to avoid these high-risk segments. The alternative paths replace the high-risk segments in the original candidate paths, forming new candidate paths. The cost of the new candidate paths is determined based on their potential energy values.

[0134] This embodiment uses multidimensional change curves and preset conditions for judging high-risk grid cells to accurately identify high-risk areas (high-risk grid cells and high-risk path segments) on the inspection path, and promptly detects risk points such as abnormal temperature, obstacle threats, and poor signal in the inspection environment. This embodiment adopts a local avoidance strategy to avoid high-risk path segments, prevent the target robot from inspecting in high-risk areas, ensure the operational safety of the target robot, and improve the inspection quality.

[0135] For example, a power substation is used as the target inspection area, with a preset quantity threshold of 2. Based on the candidate path of the target robot, multiple target grid cells traversed by the candidate path are determined. The temperature anomaly potential energy value, obstacle potential energy value, and information potential energy value of each target grid cell are obtained, and corresponding temperature anomaly potential energy value change curves, obstacle potential energy value change curves, and information potential energy value change curves are plotted respectively. Based on preset conditions, target grid cells are selected where the temperature anomaly potential energy value is greater than or equal to the temperature anomaly risk threshold, the obstacle potential energy value is greater than or equal to the obstacle risk threshold, the information potential energy value is less than or equal to the signal risk threshold, and the rate of change of any dimension's change curve between two adjacent target grid cells exceeds a preset rate of change threshold. (Reference) Figure 5 A schematic diagram of the temperature anomaly potential energy value change curve, showing the rate of change between the seventh and eighth target grid cells in the candidate path. In this embodiment, the rate of change between other adjacent target grid cells besides the seventh and eighth target grid cells is not shown in the diagram. Figure 5The diagram shows the sequence of target grid cells in the candidate path, where the horizontal axis x represents the order of the target grid cells, and the vertical axis y represents the rate of change of the temperature anomaly potential energy value change curve between two adjacent target grid cells. In this embodiment, the temperature anomaly potential energy value, obstacle potential energy value, and information potential energy value of each target grid cell are all within their corresponding preset potential energy values. However, in the path segment corresponding to target grid cell A to target grid cell B, the rate of change of the temperature anomaly potential energy change curve between two adjacent target grid cells is greater than a preset rate of change threshold. The corresponding target grid cells are then filtered out and marked as high-risk grid cells. If four consecutive high-risk grid cells are found in the candidate path, exceeding a preset number threshold (the preset number threshold is 2), the path segment corresponding to these high-risk grid cells is confirmed as a high-risk path segment. (Reference) Figure 6 This is a schematic diagram illustrating the determination of candidate paths during the inspection process of the target robot. Figure 6 The path segment corresponding to the four consecutive high-risk grid cells in the shaded area (from target grid cell A to target grid cell B) is designated as a high-risk path segment.

[0136] For high-risk path segments, an alternative path is planned to bypass the high-risk path segment based on the local avoidance strategy. The high-risk path segment is then replaced by the alternative path to form a new candidate path.

[0137] refer to Figure 6 As an example, in this embodiment, the original candidate path is O→C→A→B→D→M. Since the path from target grid cell A to target grid cell B is a high-risk path segment, it needs to be avoided. Furthermore, the risk attributes of a high-risk path segment are not isolated; surrounding target grid cells may have risk associations, such as abnormal temperature diffusion or electromagnetic interference radiation. To further mitigate hidden risks, following the principle of local avoidance and the collaborative logic of the biological foraging algorithm, when planning alternative paths, the biological foraging algorithm first filters out a set of low-potential-energy candidate grid cells based on the dynamic environmental potential energy matrix (for example, the low-potential-energy candidate grid cell set includes target grid cell G, the right side of G, and the right side of D). Based on each candidate grid cell in this set, the corresponding path cost and the risk corresponding to that path cost are calculated. Finally, the target grid cell with the lowest cost and lowest risk is selected as the relay unit. In this embodiment, grid cell G satisfies the conditions of having the lowest cost, shortest path, and lowest risk among the candidate grid cells. That is, in this embodiment, grid cell G is selected as the relay cell. In other words, the path segment corresponding to the target grid cell C to the target grid cell G and then to the target grid cell D can be selected as the alternative path to obtain the new candidate path O→C→G→D→M. This new candidate path O→C→G→D→M is only a possible alternative path and is not intended to be a limitation.

[0138] Based on all target grid cells traversed by the new candidate path, the potential energy value of the new candidate path is determined, and the cost of the new candidate path is determined based on the potential energy value of the new candidate path.

[0139] In one embodiment of this application, after forming a new candidate path based on the alternative path, the method further includes:

[0140] Obtain the multi-dimensional change curves corresponding to the new candidate paths;

[0141] For each dimension of the change curve, based on the curve difference method, the rate of change of the change curve of that dimension between two adjacent target grid cells is determined according to the potential energy value of the change curve of that dimension in the adjacent target grid cells.

[0142] The path segments corresponding to target grid cells with a rate of change greater than or equal to a preset rate of change threshold are designated as out-of-standard road segments.

[0143] Based on the road sections exceeding the standards, high-risk route sections, and the preset expansion range, determine the expanded high-risk areas;

[0144] Based on expanding the high-risk area, new candidate paths are adjusted to obtain the adjusted candidate paths;

[0145] Based on the potential energy value, determine the cost of the adjusted candidate path.

[0146] In this embodiment, all target grid cells traversed by the new candidate path are determined. For each target grid cell, the potential energy value corresponding to each dimension (temperature data, obstacle height data, and signal strength data) of the target grid cell is obtained. The change curves of each dimension are plotted with the order of the target grid cells in the new candidate path as the horizontal axis and the potential energy of each dimension as the vertical axis.

[0147] For each dimension's variation curve, the target grid cell corresponding to the substitution path and the target grid cells adjacent to the target grid cell corresponding to the substitution path are determined. Based on the determined target grid cells, two adjacent target grid cells are selected, and the rate of change of the variation curve for that dimension is determined using the curve difference calculation formula. The rate of change between each pair of adjacent target grid cells corresponding to the substitution path is obtained, thus obtaining the set of rate of change of all adjacent target grid cells in that dimension.

[0148] For example, for temperature data, the formula for calculating the curve difference corresponding to the temperature anomaly potential energy value change curve is: ,in, The rate of change of the curves of two adjacent target grid cells in this dimension. For the first i +1 target mesh cell's temperature anomaly potential energy value, For the first i Temperature anomaly potential energy value of each target grid cell d This represents the distance between adjacent target grid cells.

[0149] Based on the preset change rate thresholds for each dimension, if there is a path segment corresponding to an adjacent target grid cell with a change rate greater than or equal to the corresponding preset change rate threshold, then that path segment is identified as an out-of-standard road segment.

[0150] For each road segment exceeding the standard, an extended high-risk area is determined within the target inspection area by combining the preset expansion range and the identified high-risk path segments. In this embodiment, the preset expansion range is based on the road segment exceeding the standard and the high-risk path segments, extending one target grid cell in each of the front, back, left, and right directions of the path to form the extended high-risk area within the target inspection area. Alternatively, the preset expansion range can be determined according to the actual situation.

[0151] The new candidate path is adjusted based on the expanded high-risk area to obtain the adjusted candidate path. The cost of the adjusted candidate path is determined based on the potential energy values ​​of all target grid cells corresponding to the adjusted candidate path.

[0152] After obtaining an alternative path, based on the preset rate of change thresholds for each dimension, if there is no path segment corresponding to an adjacent target grid cell with a rate of change greater than or equal to the corresponding preset rate of change threshold, then the alternative path is taken as a new candidate path. The cost of this new candidate path is determined based on the potential energy values ​​of all target grid cells corresponding to it.

[0153] Based on the same inventive concept, this application also provides a robot inspection control device for implementing the robot inspection control method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more robot inspection control device embodiments provided below can be found in the limitations of the robot inspection control method described above, and will not be repeated here.

[0154] This application provides a robot inspection control device, such as... Figure 7 As shown, the robot inspection control device 70 includes a target inspection area determination module 71 and an inspection control module 72.

[0155] In one embodiment of this application, the target inspection area determination module 71 is used to obtain the area to be inspected and divide the area to be inspected into multiple target inspection areas.

[0156] The inspection control module 72 is used to perform target inspection operations for each target inspection area;

[0157] Specifically, when performing target inspection operations, the inspection control module 72 is used for:

[0158] Acquire environmental data during the inspection process of the target robot in the target inspection area. The environmental data includes temperature data, obstacle height data, and signal strength data. The signal strength data is the signal strength between the target robot and the control device for information exchange.

[0159] Based on the environmental data, determine the potential energy value corresponding to each environmental data point, and then determine the dynamic environmental potential energy matrix of the target inspection area based on the potential energy value corresponding to each environmental data point.

[0160] Based on the dynamic environmental potential energy matrix and using a biological foraging algorithm, the inspection path within the target inspection area is determined.

[0161] The target robot is controlled to perform inspections within the target inspection area based on the inspection path.

[0162] In one embodiment of this application, the environmental data includes multiple sets of data, which are data collected at different acquisition locations. Each set of data includes: temperature data, obstacle height data, and signal strength data. After acquiring the environmental data of the target robot's inspection process within the target inspection area, the inspection control module 72 is specifically used for:

[0163] For each set of data, determine the timestamp corresponding to each data point in the set. The timestamp of a data point represents the time when the data was collected. Based on the timestamps, align the set of data in the time dimension.

[0164] Based on the time-aligned data, determine the potential energy value group corresponding to each data group. A potential energy value group includes the potential energy values ​​corresponding to temperature data, obstacle height data, and signal strength data, respectively.

[0165] The dynamic environmental potential energy matrix of the target inspection area is constructed based on the potential energy values ​​corresponding to each group of data.

[0166] In one embodiment of this application, the target inspection area is divided into multiple target grid cells, each containing a sampling location. When determining the potential energy value corresponding to each environmental data based on the environmental data, and determining the dynamic environmental potential energy matrix of the target inspection area based on the potential energy value corresponding to each environmental data, the inspection control module 72 is specifically used for:

[0167] For each set of data after time alignment, feature extraction is performed on each data point in the set to obtain the potential energy value corresponding to each data point, and the potential energy value is normalized accordingly.

[0168] Determine the weights corresponding to each type of environmental data;

[0169] For each target grid cell, the potential energy value of the target grid cell is determined based on the normalized potential energy value corresponding to the target grid cell and the weights corresponding to each environmental data.

[0170] The dynamic environmental potential energy matrix of the target inspection area is constructed based on the potential energy value corresponding to each target grid cell.

[0171] In one embodiment of this application, when determining the inspection path within the target inspection area based on the dynamic environmental potential energy matrix and a biological foraging algorithm, the inspection control module 72 is specifically used for:

[0172] A candidate path set is determined based on a biological foraging algorithm. The candidate path set contains multiple candidate paths, and each candidate path is composed of multiple target grid cells.

[0173] Obtain the fitness value corresponding to each candidate path;

[0174] By copying candidate paths with fitness values ​​higher than the first preset fitness value and eliminating candidate paths with fitness values ​​lower than the second preset fitness value, an adjusted set of candidate paths is determined.

[0175] Based on the adjusted candidate path set, the optimal candidate path is determined through a path intersection mechanism. The optimal candidate path is the candidate path with the highest fitness value.

[0176] The optimal candidate path is used as the inspection path within the target inspection area;

[0177] The methods for obtaining the fitness value of each candidate path include:

[0178] Determine the potential energy value corresponding to the candidate path in the dynamic environment potential energy matrix;

[0179] The cost of the candidate path is determined based on the potential energy value;

[0180] The fitness value of a candidate path is determined based on its cost.

[0181] In one embodiment of this application, after determining the potential energy value corresponding to the candidate path in the dynamic environment potential energy matrix, the inspection control module 72 is specifically used for:

[0182] Based on the potential energy values ​​corresponding to each target grid cell traversed by the candidate path, a multidimensional change curve corresponding to the candidate path is determined. The multidimensional change curve includes: temperature anomaly potential energy value change curve, obstacle potential energy value change curve, and information potential energy value change curve. Each dimension change curve is determined based on the potential energy value corresponding to each target grid cell in that dimension and is used to characterize the change in potential energy value corresponding to each target grid cell traversed by the candidate path in that dimension.

[0183] Target grid cells that meet at least one of the preset conditions are identified as high-risk grid cells;

[0184] If there is a preset number of high-risk grid cells in the candidate path, then a high-risk path segment is determined. The preset number is greater than a preset number threshold. The high-risk path segment consists of a preset number of high-risk grid cells.

[0185] The alternative path is determined based on the local avoidance strategy, and a new candidate path is formed based on the alternative path;

[0186] The cost of determining the candidate path based on the potential energy value includes:

[0187] The cost of the new candidate path is determined based on the potential energy value;

[0188] The preset conditions include:

[0189] The potential energy value of temperature anomaly is greater than or equal to the temperature anomaly risk threshold.

[0190] The obstacle's potential energy value is greater than or equal to the obstacle risk threshold;

[0191] The information potential value is less than or equal to the signal risk threshold;

[0192] The rate of change of the change curve in any dimension between two adjacent target grid cells is greater than or equal to a preset rate of change threshold.

[0193] In one embodiment of this application, after forming a new candidate path based on the alternative path, the target inspection area determination module 71 is specifically used for:

[0194] Obtain the multidimensional change curves corresponding to the new candidate paths;

[0195] For each dimension of the change curve, based on the curve difference method, the rate of change of the change curve in each of the two adjacent target grid cells is determined according to the potential energy value of the change curve in the adjacent target grid cells.

[0196] If there are target grid cells with a rate of change greater than or equal to a preset rate of change threshold, then the path segment corresponding to the target grid cell with a rate of change greater than or equal to the preset rate of change threshold is regarded as an out-of-standard road segment.

[0197] Based on the road sections exceeding the standards, high-risk route sections, and the preset expansion range, determine the expanded high-risk areas;

[0198] Based on expanding the high-risk area, new candidate paths are adjusted to obtain the adjusted candidate paths;

[0199] The cost of determining the new candidate path based on the potential energy value includes:

[0200] Based on the potential energy value, determine the cost of the adjusted candidate path.

[0201] In one embodiment of this application, before dividing the area to be inspected into multiple target inspection areas, the target inspection area determination module 71 is specifically used for:

[0202] Acquire equipment-related data for the area to be inspected, and standardize the equipment-related data, which includes historical failure frequency data, equipment density data, and the number of different types of equipment.

[0203] Based on the standardized equipment-related data and the regional complexity index calculation formula, the complexity index of the area to be inspected is determined.

[0204] If the complexity index exceeds the preset threshold, the area to be inspected will be divided into multiple sub-areas, and each sub-area will be used as a target inspection area.

[0205] The formula for calculating the regional complexity index is as follows: ,in, For standardized equipment density data, n The standardized number of equipment types, f This is standardized historical anomaly frequency data. These are the dynamic weights corresponding to equipment density data, the number of equipment types, and historical anomaly frequency data, respectively.

[0206] See Figure 8 , Figure 8 This is a schematic block diagram of a control device provided in one embodiment of this application. Figure 8The control device 800 shown in this embodiment may include one or more processors 801, one or more input devices 802, one or more output devices 803, and one or more memories 804. The processors 801, input devices 802, output devices 803, and memories 804 communicate with each other via a communication bus 805. The memory 804 stores computer programs, including program instructions. The processor 801 executes the program instructions stored in the memory 804. The processor 801 is configured to invoke the program instructions to perform the functions of each module / unit in the above-described device embodiments, for example... Figure 7 The functions of the target inspection area determination module 71 and the inspection control module 72 are shown.

[0207] It should be understood that, in the embodiments of this application, the processor 801 may be a central processing unit (CPU), but it may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0208] Input device 802 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 803 may include a display (LCD, etc.), a speaker, etc.

[0209] The memory 804 may include read-only memory and random access memory, and provides instructions and data to the processor 801. A portion of the memory 804 may also include non-volatile random access memory. For example, the memory 804 may also store environmental data of the target inspection area, the potential energy values ​​corresponding to each environmental data, the dynamic environmental potential energy matrix, and information such as the target inspection path.

[0210] In specific implementations, the processor 801, input device 802, and output device 803 described in the embodiments of this application can execute the implementation method described in the robot inspection control method provided in the embodiments of this application, or they can execute the implementation method of the control device described in the embodiments of this application, which will not be repeated here.

[0211] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the methods described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0212] The computer-readable storage medium can be an internal storage unit of the control device in any of the foregoing embodiments, such as a hard disk or memory of the control device. The computer-readable storage medium can also be an external storage device of the control device, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc., provided on the control device. Furthermore, the computer-readable storage medium can include both internal storage units and external storage devices of the control device. The computer-readable storage medium is used to store computer programs and other programs and data required by the control device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0213] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0214] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the control device and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0215] In the several embodiments provided in this application, it should be understood that the disclosed control devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of modules / units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules, units, or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces or modules / units, or it may be an electrical, mechanical, or other form of connection.

[0216] The modules / units described as separate components may or may not be physically separate. Similarly, the components shown as modules / units may or may not be physical modules / units; they may be located in one place or distributed across multiple network modules / units. Some or all of the modules / units can be selected to achieve the purpose of the embodiments of this application, depending on actual needs.

[0217] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The integrated modules / units described above can be implemented in hardware or in the form of software functional modules / units.

[0218] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A robot inspection control method, characterized in that, Performed by a control device, the method includes: Obtain the area to be inspected and divide it into multiple target inspection areas; For each target inspection area, perform the target inspection operation; The target inspection operation includes: The environmental data of the target robot during its inspection within the target inspection area is acquired. The environmental data includes multiple sets of data, which are data collected at different acquisition locations. Each set of data includes: temperature data, obstacle height data, and signal strength data. The signal strength data is the signal strength between the target robot and the control device for information exchange. For each set of data, determine the timestamp corresponding to each data point in the set. The timestamp of a data point represents the collection time of that data. Based on the timestamps, align the set of data in the time dimension. Based on the time-aligned data, determine the potential energy value group corresponding to each data group. A potential energy value group includes the potential energy values ​​corresponding to temperature data, obstacle height data, and signal strength data, respectively. Construct a dynamic environmental potential energy matrix for the target inspection area based on the potential energy value groups corresponding to each set of data. A candidate path set is determined based on a biological foraging algorithm. The candidate path set contains multiple candidate paths, and each candidate path is composed of multiple target grid cells. Obtain the fitness value corresponding to each candidate path; By copying candidate paths with fitness values ​​higher than a first preset fitness value and eliminating candidate paths with fitness values ​​lower than a second preset fitness value, an adjusted set of candidate paths is determined. Based on the adjusted candidate path set, the optimal candidate path of the adjusted candidate path set is determined through a path intersection mechanism. The optimal candidate path is the candidate path with the highest fitness value. The optimal candidate path is used as the inspection path within the target inspection area; The methods for obtaining the fitness value of each candidate path include: Determine the potential energy value corresponding to the candidate path in the dynamic environment potential energy matrix; The cost of the candidate path is determined based on the potential energy value; The fitness value of the candidate path is determined based on its cost. The target robot is controlled to perform inspections within the target inspection area according to the inspection path.

2. The robot inspection control method as described in claim 1, characterized in that, The target inspection area is divided into multiple target grid cells, each containing a data acquisition location. The process of determining the potential energy value group corresponding to each set of data based on the time-aligned data includes: For each set of data after time alignment, feature extraction is performed on each data point in the set to obtain the potential energy value corresponding to each data point, and the potential energy value is normalized accordingly. The step of constructing the dynamic environmental potential energy matrix of the target inspection area based on the potential energy value groups corresponding to each set of data includes: Determine the weights corresponding to each type of environmental data; For each target grid cell, the potential energy value of the target grid cell is determined based on the normalized potential energy value corresponding to the target grid cell and the weights corresponding to each environmental data. The dynamic environmental potential energy matrix of the target inspection area is constructed based on the potential energy value corresponding to each target grid cell.

3. The robot inspection control method as described in claim 1, characterized in that, After determining the potential energy value corresponding to the candidate path in the dynamic environment potential energy matrix, the method further includes: Based on the potential energy values ​​corresponding to each target grid cell traversed by the candidate path, a multidimensional change curve corresponding to the candidate path is determined. The multidimensional change curve includes: temperature anomaly potential energy value change curve, obstacle potential energy value change curve, and information potential energy value change curve. Each dimension change curve is determined based on the potential energy value corresponding to each target grid cell in that dimension and is used to characterize the change in potential energy value corresponding to each target grid cell traversed by the candidate path in that dimension. Target grid cells that meet at least one of the preset conditions are identified as high-risk grid cells; If there is a consecutive preset number of high-risk grid cells in the candidate path, then a high-risk path segment is determined. The preset number is greater than a preset number threshold, and the high-risk path segment is composed of the consecutive preset number of high-risk grid cells. An alternative path is determined based on a local avoidance strategy, and a new candidate path is formed based on the alternative path; The step of determining the cost of the candidate path based on the potential energy value includes: The cost of the new candidate path is determined based on the potential energy value; The preset conditions include: The potential energy value of temperature anomaly is greater than or equal to the temperature anomaly risk threshold. The obstacle's potential energy value is greater than or equal to the obstacle risk threshold; The information potential value is less than or equal to the signal risk threshold; The rate of change of the change curve in any dimension between two adjacent target grid cells is greater than or equal to a preset rate of change threshold.

4. The robot inspection control method as described in claim 3, characterized in that, After forming new candidate paths based on the alternative paths, the method further includes: Obtain the multidimensional change curve corresponding to the new candidate path; For each dimension of the change curve, based on the curve difference method, the rate of change of the change curve in each of the two adjacent target grid cells is determined according to the potential energy value of the change curve in the adjacent target grid cells. If there is a target grid cell whose rate of change is greater than or equal to a preset rate of change threshold, then the path segment corresponding to the target grid cell whose rate of change is greater than or equal to the preset rate of change threshold is regarded as an out-of-standard road segment. Based on the road sections exceeding the standards, the high-risk road sections, and the preset expansion range, the expanded high-risk area is determined; Based on the expanded high-risk area, new candidate paths are adjusted to obtain the adjusted candidate paths. The step of determining the cost of the new candidate path based on the potential energy value includes: Based on the potential energy value, the cost of the adjusted candidate path is determined.

5. The robot inspection control method as described in claim 1, characterized in that, Before dividing the area to be inspected into multiple target inspection areas, the following steps are also included: Acquire equipment-related data for the area to be inspected, and standardize the equipment-related data, which includes historical fault frequency data, equipment density data, and the number of equipment types. Based on the standardized equipment-related data and the regional complexity index calculation formula, the complexity index of the area to be inspected is determined. The division of the area to be inspected into multiple target inspection areas includes: If the complexity index exceeds a preset threshold, the area to be inspected is divided into multiple sub-areas, and each sub-area is used as a target inspection area. The formula for calculating the regional complexity index is as follows: ,in, For standardized equipment density data, n The standardized number of equipment types, f This is standardized historical anomaly frequency data. These are the dynamic weights corresponding to equipment density data, the number of equipment types, and historical anomaly frequency data, respectively.

6. A robot inspection control device, characterized in that, The device includes: The target inspection area determination module is used to obtain the area to be inspected and divide the area to be inspected into multiple target inspection areas. The inspection control module is used to execute target inspection operations for each target inspection area; Specifically, when performing target inspection operations, the inspection control module is used for: The environmental data of the target robot during its inspection within the target inspection area is acquired. The environmental data includes multiple sets of data, which are data collected at different acquisition locations. Each set of data includes: temperature data, obstacle height data, and signal strength data. The signal strength data is the signal strength between the target robot and the control device for information exchange. For each set of data, determine the timestamp corresponding to each data point in the set. The timestamp of a data point represents the collection time of that data. Based on the timestamps, align the set of data in the time dimension. Based on the time-aligned data, determine the potential energy value group corresponding to each data group. A potential energy value group includes the potential energy values ​​corresponding to temperature data, obstacle height data, and signal strength data, respectively. Construct a dynamic environmental potential energy matrix for the target inspection area based on the potential energy value groups corresponding to each set of data. A candidate path set is determined based on a biological foraging algorithm. The candidate path set contains multiple candidate paths, and each candidate path is composed of multiple target grid cells. Obtain the fitness value corresponding to each candidate path; By copying candidate paths with fitness values ​​higher than a first preset fitness value and eliminating candidate paths with fitness values ​​lower than a second preset fitness value, an adjusted set of candidate paths is determined. Based on the adjusted candidate path set, the optimal candidate path of the adjusted candidate path set is determined through a path intersection mechanism. The optimal candidate path is the candidate path with the highest fitness value. The optimal candidate path is used as the inspection path within the target inspection area; Specifically, for each candidate path, the inspection control module, when obtaining the fitness value of that candidate path, is used to: Determine the potential energy value corresponding to the candidate path in the dynamic environment potential energy matrix; The cost of the candidate path is determined based on the potential energy value; The fitness value of the candidate path is determined based on its cost. The target robot is controlled to perform inspections within the target inspection area according to the inspection path.

7. A control device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 5.

Images