Robot path dynamic coverage method, system and device and storage medium

Abstract

The embodiment of the invention provides a robot path dynamic coverage method, system and device and a storage medium, and relates to the technical field of battery manufacturing, and the method comprises the steps: dividing a passable region on a target map into a plurality of sub-regions, and determining the adjacent relation between the sub-regions; the robot is controlled to advance according to the adjacent relation, and accumulated advancing data in the current sub-area are obtained in real time; calculating an overlapping distance based on the accumulative travel data and a preset adaptive error model; and extending the boundary of the next sub-region outwards by an overlapping distance to obtain a target coverage range of the next sub-region. In this way, the cumulative travel data of the robot are obtained in real time, the overlapping distance between the sub-regions is dynamically adjusted to compensate the cumulative error of the speedometer in the task execution process of the robot, the coverage range of the next sub-region is increased, the situation that scanning leakage occurs near the boundary of the next sub-region, and a vacuum belt difficult to repair is formed is reduced, and the service life of the robot is prolonged. Therefore, the completeness and reliability of the coverage operation are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of battery manufacturing, in particular to a robot path dynamic coverage method, system, device and storage medium. BACKGROUND

[0002] Autonomous mobile robots need to complete the complete coverage task of the specified area in cleaning, exploration, agricultural work and other scenes, and one of the core technologies is coverage path planning. The existing coverage path planning technology usually follows the mode of decomposition first and coverage later.

[0003] Decomposition first and coverage later is to divide the area to be covered into several sub-areas, and plan an efficient scanning path in each sub-area, such as zigzag, so as to realize systematic traversal. The robot enters each sub-area in turn according to the preset order to perform the coverage task, and a fixed width overlap band is usually set between adjacent areas to compensate for the edge error.

[0004] However, when the robot runs for a long time, its pose will gradually deviate from the true value over time and motion, forming a cumulative error of the odometer and gradually increasing, which may cause missed scanning near the boundary of the sub-area, forming a difficult-to-repair vacuum band, especially in high-frequency turning or complex terrain areas, the error accumulation is more significant, and the fixed overlap strategy cannot adjust the redundancy as needed, which is easy to cause energy waste or insufficient coverage. SUMMARY

[0005] Therefore, the purpose of the present application is to provide a robot path dynamic coverage method, system, device and storage medium, by real-time acquisition of the cumulative travel data of the robot, dynamic adjustment of the overlap distance between sub-areas to compensate for the cumulative error of the odometer during the execution of the task, reduce the situation of missed scanning near the boundary of the sub-area, forming a difficult-to-repair vacuum band, to improve the integrity and reliability of the coverage operation.

[0006] In order to achieve the above purpose, in a first aspect, the embodiment of the present application provides a robot path dynamic coverage method, the method comprising: dividing the passable area on the target map into a plurality of sub-areas, and determining the adjacent relationship between each of the sub-areas; controlling the robot to travel according to the adjacent relationship between each of the sub-areas, and real-time acquiring the cumulative travel data of the robot in the current sub-area; based on the cumulative travel data of the robot in the current sub-area and a preset adaptive error model, calculating an overlap distance; extending the boundary of the next sub-area corresponding to the current sub-area outward by the overlap distance, to obtain the target coverage range of the next sub-area.

[0007] In the embodiment, the cumulative travel data of the robot in real time during travel of the robot in the plurality of sub-regions according to the adjacency relationship is obtained through the adjacency relationship between the plurality of sub-regions, so as to calculate the overlap distance by combining the cumulative travel data of the robot with the preset adaptive error model, and then add the overlap distance to the range of the next sub-region to obtain the target coverage range of the next sub-region. In this way, the cumulative travel data of the robot is obtained in real time, the overlap distance between the sub-regions is dynamically adjusted to compensate for the odometer cumulative error of the robot during task execution, the coverage range of the next sub-region is increased, the situation that the next sub-region boundary is not scanned and a difficult-to-repair vacuum belt is formed is reduced, and the integrity and reliability of the coverage operation are improved.

[0008] In some embodiments, the cumulative travel data includes a cumulative travel distance and a cumulative turning angle, the preset adaptive error model includes a first basic overlap value, a first distance error coefficient and a second turning error coefficient, and the calculation of the overlap distance based on the cumulative travel data of the robot in the current sub-region and the preset adaptive error model includes: calculating a first compensation overlap value according to the cumulative travel distance, the cumulative turning angle, the first distance error coefficient and the first turning error coefficient; and obtaining the overlap distance based on the first compensation overlap value and the first basic overlap value.

[0009] In this way, the first compensation overlap value and the first basic overlap value are added to generate a complete overlap distance, which is used to cope with sudden small disturbances; and the two error coefficients reflect the contribution weights of different types of motion behaviors to the cumulative error, realize the sub-modeling of the path deviation, consider the basic safety margin, introduce a dynamic adjustment component, and improve the adaptability of the coverage strategy.

[0010] In some embodiments, the cumulative travel data includes a cumulative travel distance and a cumulative turning angle, each of the sub-regions has a corresponding region type label, the preset adaptive error model includes a second basic overlap value, a second distance error coefficient and a second turning error coefficient, each of the region type labels corresponds to a set of distance error coefficients and turning error coefficients; and the calculation of the overlap distance based on the cumulative travel data of the robot in the current sub-region and the preset adaptive error model includes: determining the set of distance error coefficients and turning error coefficients corresponding to the region type label of the current sub-region in which the robot is located as the second distance error coefficient and the second turning error coefficient respectively according to the region type label of the current sub-region in which the robot is located; calculating a second compensation overlap value according to the cumulative travel distance, the cumulative turning angle, the second distance error coefficient and the second turning error coefficient; and obtaining the overlap distance based on the second compensation overlap value and the second basic overlap value.

[0011] In this way, the distance error coefficient and the turning error coefficient corresponding to the environment complexity of the sub-area are selected in the sub-area where the different area type label is located, the scene adaptive configuration of the error modeling parameters is realized, and the coverage redundancy can be adjusted differently according to the actual operation condition of different areas.

[0012] In some embodiments, the cumulative travel data includes a cumulative travel distance and a cumulative turning angle, the preset adaptive error model includes a third basic overlap value, a third distance error coefficient, a third turning error coefficient, and an abnormal response overlap value; the calculating the overlap distance based on the cumulative travel data of the robot in the current sub-area and the preset adaptive error model includes: obtaining an average moving speed and a cumulative absolute steering angle of the robot in the current sub-area; in a case where the average moving speed and the cumulative absolute steering angle meet a preset condition, determining a first preset overlap value as the abnormal response overlap value; in a case where the average moving speed and the cumulative absolute steering angle do not meet the preset condition, determining a second preset overlap value as the abnormal response overlap value; the preset condition is that the average moving speed is less than a first preset value and the cumulative absolute steering angle is greater than a second preset value; a third compensation overlap value is calculated according to the cumulative travel distance, the cumulative turning angle, the third distance error coefficient, and the third turning error coefficient; and the overlap distance is obtained based on the third compensation overlap value, the third basic overlap value, and the abnormal response overlap value.

[0013] In this way, according to the dual determination of the average moving speed and the cumulative absolute steering angle, the corresponding first preset overlap value or second preset overlap value is introduced as the current abnormal response overlap value, which introduces an identification mechanism for abnormal motion behavior, and by means of additional superposition of a punitive overlap amount, the coverage path of the next area is greatly expanded outward, which helps to guide the robot to escape from the invalid circulation state.

[0014] In some embodiments, the cumulative travel data includes a covariance matrix representing the positioning error degree of the robot, and the preset adaptive error model includes a fourth basic overlap value and an adjustment coefficient; the calculating the overlap distance based on the cumulative travel data of the robot in the current sub-area and the preset adaptive error model includes: calculating a fourth compensation overlap value according to the covariance matrix and the adjustment coefficient; and the overlap distance is obtained based on the fourth basic overlap value and the fourth compensation overlap value.

[0015] In this way, by directly and accurately obtaining the positioning data of the robot, that is, the covariance matrix, the uncertainty degree of the current positioning of the robot is known, so as to directly utilize the internal state information of the positioning, and the accurate response to the path uncertainty is realized.

[0016] In some embodiments, the accumulated travel distance is accumulated by travel distances of the robot in each of the sub-regions, and the accumulated turning angle is accumulated by turning angles of the robot in each of the sub-regions.

[0017] In this way, even if the robot pauses or detours obstacles halfway, the accumulated data can still accurately reflect the overall motion intensity, avoid misjudgment of the model due to temporary interruption, ensure the integrity and time sequence consistency of the data, and provide a reliable basis for subsequent adaptive adjustment based on the travel history.

[0018] In some embodiments, the dividing the passable region on the target map into a plurality of sub-regions comprises: performing binaryzation processing on the target map to obtain a corresponding binary grid map; and dividing the passable region in the binary grid map into a plurality of sub-regions.

[0019] In this way, the dependence on high-precision vector maps is reduced, and the versatility and deployment convenience of the entire path coverage system are enhanced.

[0020] In a second aspect, an embodiment of the present application provides a robot path dynamic coverage system, comprising: an acquisition module configured to divide a passable region on a target map into a plurality of sub-regions and determine adjacent relationships between the sub-regions; a robot configured to travel according to the adjacent relationships between the sub-regions and acquire accumulated travel data of the robot in a current sub-region in real time; and a processing module configured to calculate an overlap distance based on the accumulated travel data of the robot in the current sub-region and a preset adaptive error model, and extend a boundary of a next sub-region corresponding to the current sub-region outward by the overlap distance to obtain a target coverage range of the next sub-region.

[0021] In a third aspect, an embodiment of the present application provides an electronic device comprising a processor and a memory, wherein the memory stores a computer program capable of being executed by the processor, and the processor can execute the computer program to implement the robot path dynamic coverage method according to the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the robot path dynamic coverage method according to the first aspect.

[0023] In order to make the above objectives, characteristics and advantages of the present application more apparent and easy to understand, the following preferred embodiments are specifically described below with reference to the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some of the embodiments of the present application, and therefore should not be regarded as limiting the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0025] Figure 1 A flow chart of a robot path dynamic covering method provided by the embodiment of the present application; Figure 2 A flow chart of the sub-steps S301-S302 of the step S300 in the embodiment of the present application; Figure 1 A flow chart of the sub-steps S303-S305 of the step S300 in the embodiment of the present application; Figure 3 Figure 1 A flow chart of the sub-steps S306-S309 of the step S300 in the embodiment of the present application; Figure 4 A flow chart of the sub-steps S306-S309 of the step S300 in the embodiment of the present application; Figure 1 A flow chart of the sub-steps S306-S309 of the step S300 in the embodiment of the present application; Figure 5 Figure 1 A flow chart of the sub-steps S310-S311 of the step S300 in the embodiment of the present application; Figure 6 A functional module schematic diagram of a robot path dynamic covering system provided by the embodiment of the present application; Figure 7 A block schematic diagram of an electronic device 2000 provided by the embodiment of the present application.

[0026] Icon: 1000-robot path dynamic covering system; 1100-acquisition module; 1200-processing module; 2000-electronic device; 2100-processor; 2200-memory; 2300-bus; 2400-communication interface. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations.

[0028] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of the present application.

[0029] ​​It is to be understood that the terms "first" and "second" and similar such relational terms are used solely to distinguish one entity or action from another without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without more limitations, an element defined by an "comprising" statement serves as a means plus function alternative.

[0030] As described in the background, a key task of autonomous mobile robots in cleaning, exploration, agricultural spraying, etc. is to achieve full coverage of the designated work area. Existing coverage path planning techniques usually follow the pattern of decomposition first and then coverage. Area decomposition is to decompose an irregular work area into a series of simple and regular sub-units. For example, trapezoidal decomposition decomposes the map into multiple trapezoids, which is computationally complex. For example, approximate unit decomposition decomposes the map into a series of simple units by scan line method, but the effect of processing concave and hole in the map is not good. For example, grid-based decomposition is directly performed on the grid map, but it is difficult to form large areas suitable for efficient coverage. For example, sub-area coverage, within each simple sub-unit, the robot performs coverage along a preset path (such as a "zigzag" shape).

[0031] The core defect of this technical route is the static boundary problem. The sub-area boundary generated by decomposition is fixed, while the actual motion trajectory of the robot is affected by various error sources. Systematic errors such as inconsistent wheel diameters of the robot, inaccurate wheelbase measurement, etc. Non-systematic errors such as ground skidding, bumping, etc. Therefore, after the robot runs for a long time, its pose will gradually deviate from the true value over time and motion, forming a cumulative error of the odometer and gradually increasing, which may cause missed scanning near the sub-area boundary, forming a difficult-to-repair vacuum belt, especially in high-frequency turning or complex terrain areas, error accumulation is more significant, and the fixed overlap strategy cannot adjust the redundancy as needed, which is easy to cause energy waste or insufficient coverage.

[0032] Therefore, the embodiment of the present application provides a robot path dynamic coverage method, which acquires the cumulative travel data of the robot in real time, dynamically adjusts the overlap distance between sub-areas to compensate for the cumulative error of the odometer during the execution of the task, increases the coverage range of the next sub-area, reduces the missed scanning near the boundary of the next sub-area, and forms a difficult-to-repair vacuum belt, so as to improve the integrity and reliability of the coverage operation. Referring to Figure 1 , Figure 1A robot path dynamic coverage method flow chart is provided for an embodiment of the present application. The robot path dynamic coverage method comprises steps S100-S400. S100, dividing the passable area on the target map into a plurality of sub-regions, and determining the adjacent relationship between each sub-region.

[0033] In this embodiment, before the robot performs the global coverage task, the working environment needs to be structured and modeled to reduce the complexity of path planning. Therefore, the original target map information is received. The map is usually expressed in a grid form based on sensor collected data, where each grid cell is marked as an obstacle or a passable state. By implementing a geometric segmentation algorithm on the passable area, it is divided into several rectangular geographical units, i.e. sub-regions, which together constitute the complete workspace division result. On this basis, it is analyzed whether there is a shared boundary segment between any two sub-regions, and if so, the adjacent relationship record therebetween is established. The extraction of this topological connection relationship makes the complex continuous space navigation problem into a traversal operation on the graph structure, providing logical support for subsequent orderly travel. The key of this embodiment is that through the spatial decoupling in the preprocessing stage, the irregular terrain is effectively organized, enabling the robot to complete the area coverage in a modular manner, greatly reducing the complexity of online calculation and ensuring the systematicness and completeness of the coverage. At the same time, the adjacent relationship directly relates the shared boundary information of two rectangles, providing a direct index for subsequent overlap calculation.

[0034] In some embodiments, the target map is binarized to obtain a corresponding binary grid map; the passable area in the binary grid map is divided into a plurality of sub-regions.

[0035] In this embodiment, before the robot performs the coverage task, the original environment information needs to be converted into a structured expression form suitable for computational processing. Specifically, the initial map data collected and constructed by the sensor is received, which usually contains multiple levels of gray scale or floating point values to reflect the passing confidence of different areas. To simplify subsequent analysis, a binarization operation is performed on the map, that is, each grid cell is mapped to one of the two discrete states: obstacle or passable area. This process is completed by setting a reasonable threshold, ensuring that real non-traversable objects are accurately marked, while retaining enough free space for the robot to move. The binary grid map generated thereby becomes the basis for subsequent spatial division. On this basis, the maximum rectangular coverage algorithm is used to geometrically decompose the passable area, representing the entire effective working surface with the least number of large-area rectangular blocks as much as possible. Each such rectangle constitutes a sub-area, with clear boundaries and internal connectivity, facilitating the planning of regular scanning paths. This division method not only improves the efficiency of area management, but also provides a clear topological basis for the extraction of adjacent relationships. The key of this embodiment lies in the two-stage processing flow of binarization and segmentation, which realizes the conversion from continuous perception data to discrete logical units, making it possible to organize and traverse complex environments.

[0036] S200, controlling the robot to travel according to the adjacent relationship between the sub-areas, and acquiring cumulative travel data of the robot in the current sub-area in real time.

[0037] In this embodiment, the robot generates an access sequence according to the constructed adjacency relationship set and uses a graph search strategy to visit the uncovered sub-areas one by one. In the process of entering a sub-area to perform a coverage path (such as a zigzag trajectory), key parameters reflecting its motion behavior are continuously collected and accumulated. These parameters include the total distance traveled by the robot since the start of the task and the total amount of direction change, collectively referred to as cumulative travel data. The above data is derived from the wheel encoder or other motion sensing devices, and is continuously updated and saved in the state variables of the robot over time. It is worth noting that this data is not limited to the local movement amount within the current sub-area, but is a global cumulative value throughout the entire task process, thus effectively reflecting the trend of positioning error over time and path growth. The core idea of this embodiment is to use dynamic information in actual operation to replace static assumptions, providing an objective basis for subsequent error compensation.

[0038] S300, calculating the overlap distance based on the cumulative travel data of the robot in the current sub-area and the preset adaptive error model.

[0039] In the embodiment, the accumulated travel data is input into a pre-configured mathematical model for operation processing. The model has a function expression describing the evolution law of error embedded therein, and can output a numerical value for boundary expansion, i.e., the overlap distance, according to the input motion parameters. The value of the overlap distance depends on the current motion history of the robot and the characteristics of the environment where the robot is located, and reflects the response capability to potential positioning deviation. Due to the differences in hardware configuration and working conditions of the robot in different application scenarios, the model has multiple implementation forms, and the applicable calculation logic can be selected according to the specific requirements. As can be seen, the transition from a fixed redundancy strategy to an on-demand adjustment mechanism is realized, and the flexibility and adaptability of the coverage strategy are improved.

[0040] S400, extending the boundary of the next sub-region corresponding to the current sub-region outward by the overlap distance to obtain a target coverage range of the next sub-region.

[0041] In the embodiment, when the robot is about to transfer from the current sub-region where the coverage has been completed to the adjacent next sub-region, the overlap distance calculated in the previous step is called to perform a spatial expansion operation on the original geographical boundary of the next sub-region. Specifically, the boundary is offset by a specified distance in the normal direction thereof to the inside of the original region, forming a new virtual working surface. Before formally entering the sub-region, the robot will first perform a coverage action on the expanded region, ensuring that even if there is a positional deviation, the scope of the cleaning tool or detection device can still cover the theoretical boundary area. By introducing this dynamic adjustment mechanism, the aim is to improve the coverage reliability at the junction of adjacent regions and avoid the phenomenon of missed scanning caused by odometry drift. It can be understood that this processing method actively increases the safety margin and enhances the overall working quality without relying on high-precision positioning.

[0042] In some embodiments, the accumulated travel data includes an accumulated travel distance and an accumulated turning angle, and the preset adaptive error model includes a first basic overlap value, a first distance error coefficient, and a second turning error coefficient. For step S300, a possible implementation of the embodiment of the present application is provided, which is described with reference to Figure 2 , Figure 2 For Figure 1 the flowchart of sub-steps S301-S302 of step S300, steps S301-S302 include: S301, calculating a first compensation overlap value according to the accumulated travel distance, the accumulated turning angle, the first distance error coefficient, and the first turning error coefficient.

[0043] In this embodiment, a linear weighted combination method, i.e., a linear weighted model, is used to quantify the influence of the two main error sources into a unified length unit output. The cumulative travel distance multiplied by the first distance error coefficient represents the cumulative deviation component caused by linear motion; the cumulative turning angle multiplied by the first turning error coefficient represents the uncertainty increment caused by direction adjustment. The sum of these two is the first compensation overlap value, which serves as a supplement to the basic overlap value in the final decision. This method is suitable for resource-constrained embedded platforms because of its low computational overhead and clear physical meaning, facilitating engineering deployment. This processing flow embodies a simplified modeling approach for robot motion characteristics, effectively reflecting the error growth trend while ensuring practicality.

[0044] For example, the formula for the first compensation overlap value is: L + |, among which (Distance Error Coefficient): A very small number representing approximately how many meters of positional deviation the robot will produce for every meter it moves. (Turning Error Coefficient): Another number representing the approximate positional deviation in meters per revolution (or arc) of the robot. (The error caused by turning is usually much greater than that caused by straight-line movement). The entire formula adds the "total error caused by distance" and the "total error caused by turning" to obtain the total first compensation overlap value, providing a simple and fast method to estimate the robot's cumulative positioning error, suitable for robots with limited computing resources.

[0045] S302. Obtain the overlap distance based on the first compensation overlap value and the first basic overlap value.

[0046] In this embodiment, the first compensated overlap value generated in the previous step is superimposed with a fixed reference value, which is the first basic overlap value. This reference value is used to ensure the most basic coverage redundancy and prevent the situation where there is no overlap due to model underestimation. The first basic overlap value is a fixed safety value (e.g., 10 cm), mainly used to compensate for the physical width of the robot itself, ensuring that the edges are always covered. The superimposed result serves as the final overlap distance output, used to guide the boundary expansion operation, taking into account multiple uncertainties in real-world applications. This includes not only time-varying performance deviations but also retains a minimum safety boundary. Therefore, this combination method balances efficiency and robustness, avoiding energy waste caused by large-scale overlap throughout the process while providing sufficient fault tolerance at critical nodes.

[0047] In some embodiments, the accumulated travel distance is accumulated by travel distance of the robot in each sub-region, and the accumulated turning angle is accumulated by turning angle of the robot in each sub-region.

[0048] In the embodiment, when the robot performs the global coverage task, the positioning system of the robot relies on kinematic calculation to estimate the position and pose of the robot, and the core input of the process is the displacement and direction change information continuously fed back by the motion sensor. For example, the distance increment of each movement and the angle change of each turning of the robot are collected in real time by a wheel encoder or an inertial measurement unit, and these local data are continuously accumulated along the time axis to form the accumulated travel distance and the accumulated turning angle. The accumulation operation is not limited to the local trajectory in the current sub-region, but is throughout the entire task period, from the initial start time to the current time. Therefore, the two parameters can effectively reflect the overall motion intensity and direction adjustment frequency experienced by the robot since the start of the task, and thus serve as important proxy indicators for measuring the accumulation degree of the odometer error. It is worth noting that since the error growth is usually positively correlated with the travel distance and the number of turns, especially in the presence of mechanical bias or ground disturbance, frequent turning is more likely to cause direction drift, so including these two parameters in the modeling process helps to more accurately predict the positioning uncertainty level in the current state. The basic idea of the embodiment is to use the directly obtainable motion sensing data to build a dynamically evolving error evaluation basis, thereby providing a reliable basis for subsequent adaptive overlap control.

[0049] In some embodiments, the accumulated travel data of the robot includes the accumulated travel distance and the accumulated turning angle, each sub-region has a corresponding region type label, the preset adaptive error model includes a second basic overlap value, a second distance error coefficient and a second turning error coefficient, each region type label corresponds to a set of distance error coefficients and turning error coefficients; for step S300, a possible implementation of the embodiment of the application is provided, which is described with reference to Figure 3 , Figure 3 For Figure 1 the flowchart of sub-steps S303-S305 of step S300 in the embodiment, steps S303-S305 include: S303, according to the region type label corresponding to the current sub-region where the robot is located, determining a set of distance error coefficients and turning error coefficients corresponding to the region type label as the second distance error coefficient and the second turning error coefficient, respectively.

[0050] In this embodiment, the current sub-region is read to obtain the region type label assigned in the offline modeling stage, such as "open area", "narrow area" or "complex area", which reflects the geometric characteristics and the difficulty of passing through the region. According to the label, a set of special error coefficients matching the label are obtained by indexing the pre-stored parameter mapping table. These coefficients are calibrated according to the measured data in different environments, and higher weight values are set in high-interference areas to reflect the more serious risk of positioning degradation. By dynamically switching the model parameters, the compensation strength can be adjusted in advance, which is especially suitable for task scenarios where challenging terrain is known. The basic idea of this embodiment is to use a context-aware strategy to improve the adaptability of the coverage strategy in diverse environments, which can be used as an environment-aware adaptive model.

[0051] For example, the robot selects the corresponding distance error coefficient and turn error coefficient according to the region type label and turn error coefficient , , generally, > > For example, in a complex area The turn error coefficient is set to be high because the turn is prone to error in a complex area, so as to adapt to the upcoming challenges in advance. For example, before entering a known complex area, the robot is ready to use a more conservative (i.e., larger) overlap strategy.

[0052] S304, calculating a second compensation overlap value according to the accumulated travel distance, the accumulated turn angle, the second distance error coefficient and the second turn error coefficient.

[0053] In this embodiment, the second distance error coefficient and the second turn error coefficient selected in the previous step are used to recalculate the additional deviation caused by motion, and the specific calculation formula is the same as that of the first compensation overlap. Compared with the unified parameter model, this method can more accurately reflect the error propagation characteristics in a specific region. For example, in a "complex area" with dense furniture, the turn is frequent and prone to be affected by shielding, so a higher turn error coefficient is allocated to generate a larger compensation value. This operation process still follows the linear superposition principle, but because the input parameters are more targeted, the output results are more in line with actual needs. As can be seen, by introducing environmental context information, the mechanism improves the fine degree of error prediction, which helps to realize differentiated coverage control.

[0054] S305, obtaining an overlap distance based on the second compensation overlap value and the second basic overlap value.

[0055] In the embodiment, the second compensation overlap value is summed with another preset reference value, i.e., a second basic overlap value, which has a similar function as the aforementioned basic overlap value and is used to maintain the minimum coverage margin. The final overlap distance is thus determined and transmitted to the boundary expansion module. Such a structure allows the basic value and the compensation value to be configured independently, facilitating parameter optimization for different models or task requirements. It can be understood that, on the basis of maintaining a universal architecture, the processing manner supports personalized coverage strategy adjustment for various sub-regions, thereby enhancing the configurability and engineering applicability.

[0056] In some embodiments, the cumulative travel data of the robot includes a cumulative travel distance and a cumulative turning angle, and the preset adaptive error model includes a third basic overlap value, a third distance error coefficient, a third turning error coefficient, and an abnormal response overlap value. For step S300, a possible implementation manner of the embodiment of the application is provided with reference to Figure 4 , Figure 4 For Figure 1 the flowchart of steps S306-S309 of step S300, steps S306-S309 include: S306, obtaining an average moving speed and a cumulative absolute turning angle of the robot in the current sub-region.

[0057] In the embodiment, the speed sampling sequence and the angular velocity integral value of the robot during the running in the current sub-region are extracted from the motion log, and after statistical processing, two auxiliary judgment indexes are obtained, which can be used as a hybrid penalty model. The average moving speed reflects the overall travel rhythm, and the cumulative absolute turning angle reflects the degree of direction adjustment. These data do not directly participate in the conventional error modeling, but are used to identify abnormal behavior patterns, especially those operation situations that may cause positioning failure in a short time. By introducing an additional behavior observation dimension, the response capability to non-steady state working conditions is expanded beyond the simple motion accumulation model. Such a multi-dimensional monitoring method provides a data basis for subsequent intelligent discrimination.

[0058] S307, in a case where the average moving speed and the cumulative absolute turning angle meet a preset condition, determining the first preset overlap value as the abnormal response overlap value; in a case where the average moving speed and the cumulative absolute turning angle do not meet the preset condition, determining the second preset overlap value as the abnormal response overlap value; and the preset condition is that the average moving speed is less than a first preset value and the cumulative absolute turning angle is greater than a second preset value.

[0059] In this embodiment, the robot is evaluated whether it is in a state of suspected stuck or path planning failure. When detecting a very low moving speed but frequent turning activity, it is inferred that it may be in a situation of local oscillation or repeated attempts to escape, at which time the high-priority compensation mechanism is activated, and a larger first preset overlap value is assigned as the abnormal response overlap value. Conversely, if the composite condition is not met, a smaller second preset overlap value is used, indicating the default response level in the normal operating state; for example wherein, may be the first preset value and the second preset value, is a larger constant, that is, the first preset overlap value, and 0 is the second preset overlap value. This model can respond to non-ideal behavior of the robot (such as repeated attempts caused by local path planning failure) in real time, and a large punitive overlap is used to ensure that the boundary of the chaotic area is completely covered after escaping. This threshold determination mechanism can autonomously identify potential failure scenarios without relying on external intervention, embodying the fault-tolerant design idea for edge cases. As can be seen, this branch logic enhances the robust performance under non-ideal behavior, preventing permanent coverage loss caused by temporary loss of control.

[0060] S308, calculating a third compensation overlap value according to the accumulated travel distance, the accumulated turning angle, a third distance error coefficient and a third turning error coefficient.

[0061] In this embodiment, the linear modeling paradigm is continued, and a basic compensation amount is generated using globally accumulated motion parameters combined with a set of special coefficients. Although the calculation form is similar to other compensation values, the third distance error coefficient and the third turning error coefficient used are specially set for the hybrid penalty model, which may be specially calibrated to cooperate with the overall effect of the abnormal response mechanism, such as the third compensation overlap value being equal to This value exists independently and is used later to participate in the synthesis of the final overlap distance together with the abnormal response term. This modular design facilitates functional expansion, allowing the basic error estimation and sudden disturbance response to work together in the same framework, improving the integration and maintainability.

[0062] S309, obtaining an overlap distance based on the third compensation overlap value, a third basic overlap value and an abnormal response overlap value.

[0063] In this embodiment, three inputs are integrated at this stage: the third compensation overlap value generated by regular motion, the third basic overlap value that guarantees basic redundancy, and the abnormal response overlap value triggered by behavior discrimination. The three are added together to generate the final overlap distance output. In particular, the abnormal response overlap value can significantly increase the total value in extreme cases, achieving an "abnormal response overlap" that ensures that the robot can thoroughly sweep the boundary area that may be missed after escaping from the chaotic state. This superimposed structure allows different types of influencing factors to act in parallel, reflecting both long-term cumulative effects and responding to instantaneous abnormal events. It ensures that even if the robot experiences local path planning failure or abnormal behavior, a very large overlap area can still be used to compensate for the large coverage gap that may be generated. In most cases, there will be no abnormal response overlap, and nothing will be done. Once an abnormality is detected, a large penalty amount, i.e., the abnormal response overlap value, will be immediately imposed on the final overlap value to ensure that the coverage is foolproof.

[0064] In some embodiments, the cumulative travel data includes a covariance matrix representing the degree of positioning error of the robot, and the preset adaptive error model includes a fourth basic overlap value and an adjustment coefficient; for step S300, one possible implementation of the embodiment of the present application is described with reference to Figure 5 , Figure 5 For Figure 1 the flowchart of sub-steps S310-S311 of step S300, steps S310-S311 include: S310, calculating a fourth compensation overlap value according to the covariance matrix and the adjustment coefficient.

[0065] In this embodiment, in a high-level robot equipped with a simultaneous localization and mapping (SLAM) system, the positioning module can directly output an uncertainty measure of the pose estimate, i.e., a covariance matrix, which can be used as a high-precision covariance model for more advanced robots with a SLAM (simultaneous localization and mapping) system. The elements of this matrix reflect the width of the confidence interval of the robot in each degree of freedom, with particular attention to the dispersion of the position component, for example, Σp = , Σp is the covariance matrix, representing the "uncertainty" of the robot in the X and Y directions. The larger these two values, the greater the error of the robot. An appropriate norm or trace value is selected as the error scale index, and multiplied by the adjustment coefficient, thereby converting the abstract probability distribution into a specific physical length dimension. This processing process avoids indirect speculation of the error source and instead relies on real-time perception data for direct quantification, improving the accuracy of compensation decisions. As can be seen, this method makes full use of the output capabilities of modern perception, achieving higher-level state feedback control.

[0066] S311. Obtain the overlap distance based on the fourth basic overlap value and the fourth compensated overlap value.

[0067] In this embodiment, the fourth compensated overlap value derived from the covariance is added to a preset fourth basic overlap value to form the final overlap distance suitable for a high-precision platform. For example, the overlap distance can be proportional to a measure of positional uncertainty. ,in (Basic Overlap): This is a fixed safety value (e.g., 10 cm), mainly used to compensate for the robot's physical width, ensuring that the edges are always covered. γ is an adjustment coefficient used to convert uncertainty into overlap distance. The former, the fourth compensation overlap value derived from the covariance matrix, reflects the current positioning reliability, while the latter, the fourth basic overlap value, maintains the minimum coverage requirement. The combination of the two constitutes a complete compensation strategy, suitable for application scenarios with extremely high reliability requirements, and can automatically enhance coverage redundancy when SLAM degrades. Based on the above description, the entire methodology exhibits good scalability, compatible with different technical routes from low-end odometry to high-end fusion positioning, providing a unified coverage assurance framework for diverse robot products.

[0068] In some embodiments, taking a cleaning robot as an example, the overlap distance is described in detail, and an area consisting of an "open office (R1)" and a "dense cubicle area (R2)" is cleaned. For example, an environmental perception adaptive model is used. First, R1 is marked as an "open area" and R2 is marked as a "complex area" through offline analysis. Then, the parameters are set as follows: D_base (base overlap value) = 0.1m. Open area coefficient (distance error coefficient k_{L,o} = 0.01, turning error coefficient k_{θ,o} = 0.05); complex area coefficient (distance error coefficient k_{L,c} = 0.015, turning error coefficient k_{θ,c} = 0.15).

[0069] During the robot's coverage process, the cumulative travel distance L_traveled = 0, and the cumulative turning angle θ_accumulated = 0. Starting from R1, the robot covers R1 (open area): an open area coefficient of (0.01, 0.05) is applied, the robot performs coverage, traveling ΔL_1 = 100 meters and turning Δθ_1 = 20π. After completing R1, the state is updated: L_traveled = 100, θ_accumulated = 20π. Next, the robot makes a decision and enters R2 (complex area): the robot queries neighbor relationships and finds that R1's neighbor is R2. Furthermore, it detects that R2 is labeled "complex area" and immediately switches the error coefficient to (0.015, 0.15).

[0070] After that, the overlap distance is calculated: D_overlap_1 = D_base + k_{L,c} * L_traveled + k_{θ,c} *θ_accumulated = 0.1 + 0.015 * 100 + 0.15 * 20 * pi = 0.1 + 1.5 + 9.42 =11.02 meters (Note: Again, actual parameters will be smaller to get reasonable values, such as 0.5 meters. The calculation here is intended to illustrate that more conservative parameters are used for the calculation because of the complexity of R2). So the robot enters R2 with an overlap of 0.5 meters. Finally cover R2 (complex area): the robot continues to accumulate errors using the complex area coefficient. Due to the frequent turns,θ_accumulated grows rapidly, resulting in a larger overlap distance D_overlap_2 if there is R3 later.

[0071] Based on the above method, the embodiment of the application also provides a system corresponding to the above method, as shown in Figure 6 , the embodiment of the application provides a functional module schematic diagram of the robot path dynamic coverage system 1000. It should be noted that the robot path dynamic coverage system 1000 provided in the embodiment has the same basic principle and technical effects as the above method embodiment, and for the sake of brief description, the part not mentioned in the embodiment can be referred to the corresponding content in the method embodiment. Figure 6 , the embodiment of the application provides a functional module schematic diagram of the robot path dynamic coverage system 1000. It should be noted that the robot path dynamic coverage system 1000 provided in the embodiment has the same basic principle and technical effects as the above method embodiment, and for the sake of brief description, the part not mentioned in the embodiment can be referred to the corresponding content in the method embodiment.

[0072] In the embodiment, the robot path dynamic coverage system 1000 includes an acquisition module 1100 and a processing module 1200. The acquisition module 1100 is configured to divide the passable area on the target map into a plurality of sub-areas, determine the adjacent relationship between the sub-areas, control the robot to travel according to the adjacent relationship between the sub-areas, and acquire the cumulative travel data of the robot in the current sub-area in real time. It can be understood that the acquisition module 1100 is configured to execute steps S100-S200.

[0073] The processing module 1200 is configured to calculate an overlap distance based on the cumulative travel data of the robot in the current sub-area and a preset adaptive error model, and extend the boundary of the next sub-area corresponding to the current sub-area outward by the overlap distance to obtain a target coverage range of the next sub-area. It can be understood that the processing module 1200 is configured to execute steps S300-S400.

[0074] In some embodiments, the accumulated travel data includes an accumulated travel distance and an accumulated turning angle, the preset adaptive error model includes a first basic overlap value, a first distance error coefficient and a second turning error coefficient, and the processing module 1200 is configured to calculate a first compensation overlap value according to the accumulated travel distance, the accumulated turning angle, the first distance error coefficient and the first turning error coefficient; and obtain the overlap distance based on the first compensation overlap value and the first basic overlap value. It can be understood that the processing module 1200 is configured to perform steps S301-S302 described above.

[0075] In some embodiments, the accumulated travel data of the robot includes an accumulated travel distance and an accumulated turning angle, each sub-region has a corresponding region type label, the preset adaptive error model includes a second basic overlap value, a second distance error coefficient and a second turning error coefficient, each region type label corresponds to a set of distance error coefficients and turning error coefficients, and the processing module 1200 is configured to determine a set of distance error coefficients and turning error coefficients corresponding to the region type label as the second distance error coefficient and the second turning error coefficient according to the region type label corresponding to the current sub-region where the robot is located; calculate a second compensation overlap value according to the accumulated travel distance, the accumulated turning angle, the second distance error coefficient and the second turning error coefficient; and obtain the overlap distance based on the second compensation overlap value and the second basic overlap value. It can be understood that the processing module 1200 is further configured to perform steps S303-S305 described above.

[0076] In some embodiments, the accumulated travel data of the robot includes an accumulated travel distance and an accumulated turning angle, the preset adaptive error model includes a third basic overlap value, a third distance error coefficient, a third turning error coefficient and an abnormal response overlap value. The processing module 1200 is configured to obtain an average moving speed of the robot in the current sub-region and an accumulated absolute turning angle; determine the first preset overlap value as the abnormal response overlap value in a case where the average moving speed and the accumulated absolute turning angle meet a preset condition; determine the second preset overlap value as the abnormal response overlap value in a case where the average moving speed and the accumulated absolute turning angle do not meet the preset condition; the preset condition is that the average moving speed is less than a first preset value and the accumulated absolute turning angle is greater than a second preset value; calculate a third compensation overlap value according to the accumulated travel distance, the accumulated turning angle, the third distance error coefficient and the third turning error coefficient; and obtain the overlap distance based on the third compensation overlap value, the third basic overlap value and the abnormal response overlap value. It can be understood that the processing module 1200 is further configured to perform steps S306-S309 described above.

[0077] In some embodiments, the accumulated travel data comprises a covariance matrix representing a degree of positioning error of the robot, and the preset adaptive error model comprises a fourth basic overlap value and an adjustment coefficient; the processing module 1200 is configured to calculate a fourth compensation overlap value according to the covariance matrix and the adjustment coefficient; and the overlap distance is obtained based on the fourth basic overlap value and the fourth compensation overlap value. It can be understood that the processing module 1200 is configured to perform the above steps S310-S311.

[0078] In some embodiments, the accumulated travel distance is obtained by accumulating the travel distance of the robot during the robot travels in each sub-region, and the accumulated turning angle is obtained by accumulating the turning angle of the robot during the robot travels in each sub-region.

[0079] In some embodiments, the obtaining module 1100 is configured to perform binaryzation processing on the target map to obtain a corresponding binary grid map; and divide the passable region in the binary grid map into a plurality of sub-regions.

[0080] Based on the same inventive concept disclosed above, the embodiments of the present application also provide a block schematic diagram of an electronic device 2000 for performing the above method. Please refer to Figure 7 , Figure 7 The block schematic diagram of the electronic device 2000 provided by the embodiments of the present application comprises a processor 2100, a memory 2200, a bus 2300 and a communication interface 2400. The processor 2100 and the memory 2200 are connected through the bus 2300, and the processor 2100 communicates with external devices through the communication interface 2400.

[0081] The processor 2100 can be an integrated circuit chip with signal processing capability. In the implementation process, each step of the above method can be completed by the integrated logic circuit of hardware or the instruction of software form in the processor 2100. The processor 2100 mentioned above can be a general processor 2100, including a central processing unit 2100 (CPU), a network processor 2100 (NP) and the like; it can also be a digital signal processor 2100 (DSP), an application specific integrated circuit (ASIC), a ready programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0082] The memory 2200 is configured to store a computer program, for example, the robot path dynamic covering system 1000 in the embodiment of the present application, including at least one software function module stored in the memory 2200 in the form of software or firmware, and the processor 2100 executes the program to realize the robot path dynamic covering method in the embodiment of the present application after receiving an execution instruction.

[0083] The memory 2200 can include a high-speed random access memory 2200 (RAM) and can also include a non-volatile memory 2200. Optionally, the memory 2200 can be a storage device built in the processor 2100 or a storage device independent of the processor 2100.

[0084] The bus 2300 can be an ISA bus 2300, a PCI bus 2300, or an EISA bus 2300, etc. Figure 7 Only one bidirectional arrow is used to represent, but it does not mean that there is only one bus 2300 or only one type of bus 2300.

[0085] The electronic device 2000 can be a mobile phone, a tablet computer, a notebook computer, a desktop computer, etc.

[0086] Based on the same inventive concept, the embodiment of the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by the processor 2100 to realize the robot path dynamic covering method as described above. The computer readable storage medium can include a U disk, a mobile hard disk, a read-only memory 2200 (Read-Only Memory, ROM), a random access memory 2200 (Random Access Memory, RAM), a magnetic disk or an optical disk, etc.

[0087] The above only describes the preferred embodiments of the present application and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for dynamic coverage of robot paths, characterized in that, The method includes: Divide the passable area on the target map into multiple sub-regions and determine the adjacency relationship between each sub-region; The robot is controlled to move according to the adjacency relationship between each of the sub-regions, and the cumulative movement data of the robot in the current sub-region is acquired in real time; The overlap distance is calculated based on the robot's cumulative travel data in the current sub-region and a preset adaptive error model; Extend the boundary of the next sub-region corresponding to the current sub-region outward by the overlapping distance to obtain the target coverage area of ​​the next sub-region.

2. The method according to claim 1, characterized in that, The cumulative travel data includes cumulative travel distance and cumulative turning angle. The preset adaptive error model includes a first basic overlap value, a first distance error coefficient, and a second turning error coefficient. The calculation of the overlap distance based on the robot's cumulative travel data in the current sub-region and the preset adaptive error model includes: The first compensation overlap value is calculated based on the cumulative travel distance, the cumulative turning angle, the first distance error coefficient, and the first turning error coefficient. The overlap distance is obtained based on the first compensated overlap value and the first basic overlap value.

3. The method according to claim 1, characterized in that, The cumulative travel data includes cumulative travel distance and cumulative turning angle. Each sub-region has a corresponding region type label. The preset adaptive error model includes a second basic overlap value, a second distance error coefficient, and a second turning error coefficient. Each region type label corresponds to a set of distance error coefficients and turning error coefficients. The calculation of the overlap distance based on the robot's cumulative travel data in the current sub-region and a preset adaptive error model includes: Based on the region type label corresponding to the current sub-region where the robot is located, a set of distance error coefficients and turning error coefficients corresponding to the region type label are respectively determined as the second distance error coefficient and the second turning error coefficient. The second compensation overlap value is calculated based on the cumulative travel distance, the cumulative turning angle, the second distance error coefficient, and the second turning error coefficient. The overlap distance is obtained based on the second compensation overlap value and the second basic overlap value.

4. The method according to claim 1, characterized in that, The cumulative travel data includes cumulative travel distance and cumulative turning angle, and the preset adaptive error model includes a third basic overlap value, a third distance error coefficient, a third turning error coefficient, and an abnormal response overlap value. The calculation of the overlap distance based on the robot's cumulative travel data in the current sub-region and a preset adaptive error model includes: Obtain the robot's average moving speed and cumulative absolute turning angle in the current sub-region; If the average moving speed and the cumulative absolute turning angle meet preset conditions, a first preset overlap value is determined as an abnormal response overlap value; if the average moving speed and the cumulative absolute turning angle do not meet the preset conditions, a second preset overlap value is determined as an abnormal response overlap value; the preset conditions are that the average moving speed is less than the first preset value and the cumulative absolute turning angle is greater than the second preset value. The third compensation overlap value is calculated based on the cumulative travel distance, the cumulative turning angle, the third distance error coefficient, and the third turning error coefficient. The overlap distance is obtained based on the third compensation overlap value, the third basic overlap value, and the abnormal response overlap value.

5. The method according to claim 1, characterized in that, The cumulative travel data includes a covariance matrix, which characterizes the degree of positioning error of the robot. The preset adaptive error model includes a fourth basic overlap value and an adjustment coefficient. The calculation of the overlap distance based on the robot's cumulative travel data in the current sub-region and a preset adaptive error model includes: The fourth compensation overlap value is calculated based on the covariance matrix and the adjustment coefficient. The overlap distance is obtained based on the fourth basic overlap value and the fourth compensated overlap value.

6. The method according to any one of claims 2-5, characterized in that, The cumulative travel distance is obtained by accumulating the travel distance of the robot during its travel in each of the sub-regions; the cumulative turning angle is obtained by accumulating the turning angle of the robot during its travel in each of the sub-regions.

7. The method according to claim 1, characterized in that, The process of dividing the passable area on the target map into multiple sub-regions includes: The target map is binarized to obtain the corresponding binary raster map; The passable area in the binary raster map is divided into multiple sub-regions.

8. A dynamic coverage system for robot paths, characterized in that, The system includes: The acquisition module is used to divide the passable area on the target map into multiple sub-regions and determine the adjacency relationship between each sub-region; control the robot to move according to the adjacency relationship between each sub-region, and acquire the cumulative movement data of the robot in the current sub-region in real time; The processing module is used to calculate the overlap distance based on the robot's cumulative travel data in the current sub-region and a preset adaptive error model; and to extend the boundary of the next sub-region corresponding to the current sub-region outward by the overlap distance to obtain the target coverage area of ​​the next sub-region.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing a computer program that can be executed by the processor to implement the robot path dynamic coverage method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the robot path dynamic coverage method as described in any one of claims 1-7.

Images