Fitting speed limiting method of cleaning robot

Abstract

The invention relates to a welt speed limiting method for a cleaning robot, which belongs to the technical field of cleaning robots and comprises the following operation steps of: 1, planning a path to obtain a local path; and 2, carrying out speed limit decision to obtain a path with speed limit. And 3, carrying out speed planning to obtain a path with speed and time, namely a finally planned track. A local path planning result is used as input, and after decision optimization of the scheme, a speed limiting result is input to a local speed planning module for calculation and output. In the speed limit decision, refined speed limit calculation including emergency stop decision is carried out on the speed according to environment and map information. More importantly, an independent speed limiting module is provided, and speed setting related to the cleaning task can be integrated into the speed limiting module.

Description

Technical Field

[0001] This invention relates to the field of cleaning robot technology, and more specifically to a method for limiting the speed of a cleaning robot at its edges. Background Art

[0002] Speed ​​planning for cleaning robots is a crucial part of their navigation and task execution. It requires considering various factors to ensure the robot maximizes cleaning efficiency while maintaining effectiveness. Cleaning service robots need to perceive the cleaning environment in real time and combine this information with map data to plan cleaning paths and corresponding speeds. For example, in congested areas, the robot may need to slow down to avoid collisions, while in more open areas it can accelerate.

[0003] The prior art, "A Method, Apparatus, Electronic Device, Chip and Medium for Determining an Obstacle Avoidance Path", application number CN202110895344.3, discloses a method for selecting a starting obstacle avoidance position and an ending obstacle avoidance position on an initial trajectory; determining an intermediate obstacle avoidance position based on the starting and ending obstacle avoidance positions; and determining an obstacle avoidance path based on the intermediate, starting, and ending obstacle avoidance positions, wherein the jerk corresponding to the obstacle avoidance path is continuous.

[0004] Cleaning robots need to consider several factors when making speed-limiting decisions:

[0005] (1) Room / shopping mall layout and obstacles: The cleaning robot needs to use sensors to perceive the layout of the room and any existing obstacles in order to determine a safe path and appropriate speed to avoid collisions and damage to furniture or other items.

[0006] (2) Floor type and condition: The cleaning robot may need to adjust the cleaning speed according to the type of floor (such as wood flooring, carpet, etc.) and condition (wet or dry) to ensure the floor cleaning effect.

[0007] (3) People and other moving objects: In home or shopping mall environments, cleaning robots also need to identify the location of people and other moving objects and slow down or pause accordingly to avoid accidents.

[0008] (4) Nature of cleaning task: Depending on the nature of the cleaning task, such as inspection or edge cleaning, the robot may need to adjust its speed and cleaning method to ensure the cleaning effect.

[0009] Currently, navigation planning technology for indoor cleaning robots can be broken down into global planning and local planning. Robots use local planning to dynamically adjust their paths to avoid dynamic obstacles, such as DWA and TEB. Most existing indoor robots couple path planning and velocity planning together during local planning, completing the process using the same algorithm. The drawbacks of existing technologies are:

[0010] (1) From the perspective of the algorithm as a whole, coupled local planning requires simultaneous optimization of path and speed, resulting in a large computational load and a limited local path length in the optimized output. Since it cannot make reasonable plans for trajectories over long distances and longer periods, coupled local planning is prone to getting trapped in local optima. In contrast, local planning methods that separate path and speed only optimize a single objective require less computation and can optimize a longer distance with the same computing power.

[0011] (2) Because speed planning is coupled with path planning, an algorithm must make a trade-off between the two for the sake of overall performance. The result is often that speed performance is sacrificed to ensure path stability. This will lead to the performance of the speed planning result not being guaranteed.

[0012] (3) Speed ​​planning and path planning are coupled. Since the path cannot be determined before speed planning, speed decision is difficult to calculate accurately. The final result is that speed planning is difficult to optimize precisely.

[0013] (4) When speed planning and path planning are coupled, when facing dynamic obstacles, it is easy to cause left and right oscillations in order to bypass them. In this case, it is more appropriate to use reasonable speed planning to handle the parking situation.

[0014] On the other hand, while intelligent driving technology boasts relatively good planning capabilities, it relies on the entire autonomous driving technology stack to function. Due to limitations in sensors, maps, computing platforms, and other aspects, it's difficult to directly apply intelligent driving technology to indoor service robots. In summary, indoor robot navigation technology has limitations, while intelligent driving, despite its powerful functions, struggles to penetrate the market, creating a significant gap between the two. Summary of the Invention

[0015] This invention primarily addresses the shortcomings of existing technologies by providing a method for edge-limiting speed control for cleaning robots. It takes the results of local path planning as input, optimizes the process, and then inputs the speed limit result back to the local speed planning module for calculation and output. The speed limit decision involves refined speed calculation based on environmental and map information, including emergency braking decisions. More importantly, with a dedicated speed limit module, speed settings related to cleaning tasks can be integrated into this module.

[0016] The above-mentioned technical problems of the present invention are mainly solved by the following technical solutions:

[0017] A method for limiting the speed of a cleaning robot at edges includes the following steps:

[0018] Step 1: Perform path planning to obtain local paths.

[0019] Step 2: Make a speed limit decision to obtain a route with a speed limit.

[0020] The third step is to perform speed planning to obtain a path with speed and time, which is the final planned trajectory.

[0021] As a preferred option, the "path" in the speed limit decision module is a prerequisite. For indoor cleaning service robots, the path is obtained by the existing coupled local planner or by an independent local path planner.

[0022] As a preferred approach, in the "speed limit decision" module, the obstacle information output by perception, the obstacle information on the map, and the deceleration area are all represented in a polygonal geometric manner, and the speed limit decision is implemented using a unified "path-polygon" model.

[0023] As a preferred option, let A be the path planned by the robot. Assuming the robot's width is D, we can approximate the robot's shape with a circle to obtain two outer envelope lines B and C along A. The obstacles perceived or on the map can be represented as polygons G. Projecting each vertex of G onto A yields the longest path. Calculate the distance from each vertex on G to A, obtain the minimum distance L, and convert L to the distance L′ on the side of the robot, L′=LD / 2.

[0024] As a preferred option, in edge-fitting scenarios, design function g EdgeCleaning (L′)=a(bc×e -d×L′ ), where a, b, c, and d are the parameters of the function; in edge-fitting scenarios, design the threshold distance at which edge-fitting speed limiting takes effect. EdgeCleaning When L′ is satisfied <threshold EdgeCleaning G will trigger a speed limit, with the speed limit value being V. G =g EdgeCleaning (L′), speed limit zone For the deceleration zone Q on the map, there is a corresponding finite speed value V. Q If Q intersects A, the intersecting path is obtained. Then Q will trigger a speed limit, with a speed limit value of V. Q Speed ​​limit area

[0025] As a preferred approach, after traversing all obstacles and deceleration zones near A, a series of speed limit decisions (V) are obtained. limit ) i ∈[Ui′ W i ], [U i′ W i ] indicates the speed limit zone on A. A is essentially a series of points, A = A i , i∈[0,n], according to (V limit ) i ∈[U i’ W i The set yields A i Speed ​​limit set at points (V) limit ) j ∈A i Calculate A i Minimum speed limit at the point After the speed limit decision, the speed limit for path A is:

[0026] As a preferred approach, the speed-limiting decision model is applied to the edge-to-edge cleaning scenario, and a specific decision function is designed accordingly. Based on an edge distance of 10 cm and an edge-to-edge cleaning speed of 0.4 m / s, the speed-limiting decision function is as follows: g EdgeCleaning (L′)=0.405(1-e -45L′ When the robot is less than or equal to 0.1 meters away from an obstacle, the obstacle will make a speed limit decision.

[0027] As a preferred option, when the robot is 0.1 meters away from the obstacle, the speed limit is 0.4005 m / s; when the robot is close to the obstacle at around 0.1 meters, the speed limit is around 0.4 m / s, with little speed variation; and as the distance between the robot and the obstacle decreases, the speed limit decreases exponentially.

[0028] Global planning: When a robot performs navigation and path planning, it considers the overall environment and task requirements and makes reasonable decisions to complete the navigation task efficiently and safely.

[0029] Local planning: When a robot already has a global path plan, it needs to plan and adjust local paths during movement, such as when encountering dynamic obstacles, when the plan is inaccurate, or when it needs to avoid changes in the environment.

[0030] DWA (Dynamic Window Approach) is a robot path planning algorithm used for local path planning and obstacle avoidance in dynamic environments for mobile robots. The algorithm mainly includes several steps: motion space search, evaluation and selection, and motion execution.

[0031] TEB (Timed-Elastic-Band) is a path planning algorithm primarily used for local path planning and trajectory optimization of mobile robots in dynamic environments. The optimization algorithm aims to minimize trajectory execution time (time-optimal objective), maintain distance from obstacles, and adhere to kinematic constraints such as maximum velocity and acceleration. This algorithm helps generate trajectories that conform to the robot's dynamic characteristics and environmental changes, while ensuring path smoothness.

[0032] Frenet: The Frenet Coordinate System is a coordinate system used to describe position on a curve. This system is based on the normal and tangent directions. In the Frenet coordinate system, the normal direction (N) is perpendicular to the tangent direction (T) of the curve. Additionally, there is a direction (s) that moves forward along the curve, usually expressed as the integral of the tangent direction. These three directions (s, N, T) constitute the Frenet coordinate system.

[0033] The present invention can achieve the following effects:

[0034] This invention provides a method for edge-limiting speed control of cleaning robots. Compared with existing technologies, it uses the results of local path planning as input, optimizes the decision process, and then inputs the speed limit result into the local speed planning module for calculation and output. The speed limit decision involves refined speed calculation based on environmental and map information, including emergency stopping decisions. More importantly, with an independent speed limit module, speed settings related to cleaning tasks can be integrated into this module.

[0035] (1) The speed decision adopts an independent modular design, making it easier for the upper-level business logic to develop and maintain speed-related configurations.

[0036] (2) An independent speed decision module can calculate the speed limit more accurately. Only with accurate speed quantification can cleaning efficiency be further improved.

[0037] (3) By adopting the method of separating path and speed, only static obstacles can be considered in path planning, and dynamic obstacles can be included in the speed decision to calculate the speed limit. This can largely solve the path jitter problem caused by coupled local planners that do not distinguish between dynamic and static obstacles (oscillation occurs when turning left / right around dynamic obstacles).

[0038] (4) By adopting the decision-making framework and edge-limiting method of this scheme, we can ensure safe stopping in case of emergencies while maintaining smooth operation at the edge. Attached Figure Description

[0039] Figure 1 This is a flowchart of the overall technical solution of the present invention.

[0040] Figure 2 This is a flowchart of the speed limit decision method of the present invention.

[0041] Figure 3 This is a structural schematic diagram of the implementation process of the present invention.

[0042] Figure 4 This is the g of the present invention EdgeCleaning (L′)=0.405(1-e -45L′ (Function curve graph) Detailed Implementation

[0043] The technical solution of the invention will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0044] Example: Figure 1-4 As shown, a method for limiting the speed of a cleaning robot at its edges includes the following steps:

[0045] Step 1: Perform path planning to obtain local paths.

[0046] Step 2: Make a speed limit decision to obtain a route with a speed limit.

[0047] In the speed limit decision module, the "path" is a prerequisite. For indoor cleaning service robots, the path is obtained by the existing coupled local planner or by an independent local path planner. In the "speed limit decision" module, obstacle information output by perception, obstacle information on the map, and deceleration areas are all represented in a polygonal geometric way. Speed ​​limit decision is implemented using a unified "path-polygon" model.

[0048] Let A be the path planned by the robot. Assuming the robot's width is D, we can approximate the robot's shape using a circle to obtain two outer envelope lines B and C along A. The obstacles perceived or on the map can be represented as polygons G. Projecting each vertex of G onto A yields the longest path. Calculate the distance from each vertex on G to A, obtain the minimum distance L, and convert L to the distance L′ on the side of the robot, L′=LD / 2.

[0049] In edge-fitting scenarios, design function g. EdgeCleaning (L′)=a(bc×e -d×L′ ), where a, b, c, and d are the parameters of the function; in edge-fitting scenarios, design the threshold distance at which edge-fitting speed limiting takes effect. EdgeCleaning When L′ is satisfied <threshold EdgeCleaning G will trigger a speed limit, with the speed limit value being V. G =g EdgeCleaninlng (L′), speed limit zone For the deceleration zone Q on the map, there is a corresponding finite speed value V. Q If Q intersects A, the intersecting path is obtained. Then Q will trigger a speed limit, with a speed limit value of V. Q Speed ​​limit area

[0050] After traversing the obstacles and deceleration zones near A, a series of speed limit decisions (V) are obtained. limit ) i ∈[U i’ W i ], [U i’ Wi] represents the speed limit zone on A. A is essentially a series of points, A = A i , i∈[0,n], according to (V limit ) i ∈[U i’ W i The set yields A i Speed ​​limit set at points (V) limit ) j ∈A i Calculate A i Minimum speed limit at the point After the speed limit decision, the speed limit for path A is:

[0051] In the model above, the deceleration zone Q can be a high-traffic area, such as a passageway in a shopping mall, or an area with blind spots, such as a T-junction or intersection with walls on both sides; it can also be information related to the cleaning task, such as special carpet areas, floor areas, etc.

[0052] The third step is to perform speed planning to obtain a path with speed and time, which is the final planned trajectory.

[0053] The speed-limiting decision model is applied to the edge-to-edge cleaning scenario, and a specific decision function is designed accordingly. Assuming an edge distance of 10 cm and an edge-to-edge cleaning speed of 0.4 m / s, the speed-limiting decision function is as follows: g EdgeCleaning (L′)=0.405(1-e -45L′ When the robot is less than or equal to 0.1 meters from the obstacle, the obstacle will make a speed limit decision. When the robot is 0.1 meters from the obstacle, the speed limit is 0.4005 m / s; when the robot is close to the obstacle at around 0.1 meters, the speed limit is around 0.4 m / s, with little speed change; as the distance between the robot and the obstacle continues to decrease, the speed limit decreases exponentially.

[0054] The more common term "projection" is used to represent the geometric relationships of the graphic. More professionally, the Frenet coordinate system can be used. Whether it's "projection" or "Frenet coordinates," the essence is the same: describing the geometric relationships along a path. Edge-hugging speed limiting can be represented in the general form g. EdgeCleaning (L′)=a(bc×e -d×L′ ), where a, b, c, and d are the parameters of the function.

[0055] In summary, the edge-limiting speed method for this cleaning robot is implemented between path planning and speed planning. It takes the results of local path planning as input, optimizes them through this solution, and then inputs the speed-limiting results into the local speed planning module for calculation and output. The speed-limiting decision involves refined speed calculations based on environmental and map information, including emergency stopping decisions. More importantly, with an independent speed-limiting module, speed settings related to cleaning tasks can be integrated into this module.

[0056] Compared to the computational methods used in autonomous driving, which require the use of vehicle outlines (e.g., rectangles) when calculating spatial relative positions due to high safety requirements, this method, applied to indoor low-to-medium speed robots, simplifies lateral relative distance calculations using a circular decision module. The advantage of this method is that while sacrificing some accuracy within permissible safety limits, it achieves a reduction in computational load. This is the key reason why this patent can be applied to indoor low-to-medium speed robots, and it is also a point of protection that distinguishes it from autonomous driving technologies.

[0057] The above description is only a specific embodiment of the present invention, but the structural features of the present invention are not limited thereto. Any changes or modifications made by those skilled in the art within the scope of the present invention are covered by the patent scope of the present invention.

Claims

1. A method for limiting the speed of a cleaning robot at its edge, characterized in that... The following steps are included: Step 1: Perform path planning to obtain local paths; Step 2: Make a speed limit decision to obtain a route with a speed limit; The third step is to perform speed planning to obtain a path with speed and time, which is the final planned trajectory.

2. The edge-limiting speed method for a cleaning robot according to claim 1, characterized in that: In the speed limit decision module, "path" is a prerequisite. For indoor cleaning service robots, the path is obtained by the existing coupled local planner or by an independent local path planner.

3. The edge-limiting speed method for a cleaning robot according to claim 2, characterized in that: In the "Speed ​​Limit Decision" module, obstacle information output by perception, obstacle information on the map, and deceleration areas are all represented in a polygonal geometric way. Speed ​​limit decisions are implemented using a unified "path-polygon" model.

4. The edge-limiting speed method for a cleaning robot according to claim 3, characterized in that: Let A be the path planned by the robot. Assuming the robot's width is D, we can approximate the robot's shape using a circle to obtain two outer envelope lines B and C along A. The obstacles perceived or on the map can be represented as polygons G. Projecting each vertex of G onto A yields the longest path. Calculate the distance from each vertex on G to A, obtain the minimum distance L, and convert L to the distance L′ on the side of the robot, L′=LD / 2.

5. The edge-limiting speed method for a cleaning robot according to claim 4, characterized in that: In edge-fitting scenarios, design function g. EdgeCleaning (L′)=a(bc×e -d×L′ ), where a, b, c, and d are the parameters of the function; in edge-fitting scenarios, design the threshold distance at which edge-fitting speed limiting takes effect. EdgeCleaning When L′ is satisfied <threshold EdgeCleaning G will trigger a speed limit, with the speed limit value being V. G =g EdgeCleaning (L′), speed limit zone For the deceleration zone Q on the map, there is a corresponding speed limit. If Q intersects A, the intersecting path is obtained. Then Q will trigger a speed limit, with a speed limit value of V. Q Speed ​​limit area 6. The edge-limiting speed method for a cleaning robot according to claim 4, characterized in that: After traversing the obstacles and deceleration zones near A, a series of speed limit decisions (V) are obtained. limit ) i ∈[U i′ W i ], [U i W i ] indicates the speed limit zone on A. A is essentially a series of points, A = A i , i∈[0,n], according to (V limit ) i ∈[U i′ W i The set yields A i Speed ​​limit set at points (V) limit ) j ∈A i Calculate A i Minimum speed limit at the point After the speed limit decision, the speed limit for path A is:

7. The edge-limiting speed method for a cleaning robot according to claim 4, characterized in that: The speed-limiting decision model is applied to the edge-to-edge cleaning scenario, and a specific decision function is designed accordingly. Assuming an edge distance of 10 cm and an edge-to-edge cleaning speed of 0.4 m / s, the speed-limiting decision function is as follows: g EdgeCleaning (L′)=0.405(1-e -45L′ When the robot is less than or equal to 0.1 meters away from an obstacle, the obstacle will make a speed limit decision.

8. The edge-limiting speed method for a cleaning robot according to claim 7, characterized in that: When the robot is 0.1 meters from the edge, the speed limit is 0.4005 m / s; when the robot is close to the edge at around 0.1 meters, the speed limit is around 0.4 m / s, with little change in speed; as the distance between the robot and the obstacle continues to decrease, the speed limit decreases exponentially.

Images