CN114253270A - Method and system for automatically exploring and building picture by indoor robot - Google Patents

Abstract

The invention provides a method and a system for automatically exploring and building a diagram by an indoor robot, which relate to the technical field of robots, and the method comprises the following steps: step S1: selecting maps with different proportions according to scenes, setting angular points as reference points, traversing four angular points to determine map boundaries, and then exploring regions which are not explored inside; step S2: acquiring occupied grids in an unexplored area, converting the occupied grids into laser data, acquiring laser radar data, combining the two data to generate a histogram, and selecting an area where a robot can pass; step S3: obtaining candidate points by obtaining boundary line segments of idle grids and unknown grids, and grading and selecting an optimal target point according to rules; step S4: setting preference parameters, modifying and selecting a target point strategy, adapting to different conditions, checking missing and filling up, and confirming and storing a map. The invention can be beneficial to reducing the pressure of deployment personnel and improving the deployment efficiency.

Description

Method and system for automatically exploring and building picture by indoor robot

Technical Field

The invention relates to the technical field of robots, in particular to a method and a system for an indoor robot to automatically explore and build a map.

Background

The map used by a common indoor robot needs to be constructed and deployed in advance, usually a field deployment person pushes a machine or a remote control machine to model the environment to generate a grid map for navigation and positioning, and if a hotel with dozens of floors needs to be deployed, the field deployment person is hard, so that an autonomous exploration type map construction method is beneficial to reducing the pressure of the deployment person, and meanwhile, the deployment efficiency can be improved.

The invention patent with publication number CN111638526A discloses a method for a robot to automatically build a figure in an unfamiliar environment, which comprises the following steps: s1, generating leading edge points between the known area and the unknown area according to a through-wall random tree algorithm, and performing cluster optimization on the leading edge points to obtain preliminarily screened leading edge points; s2, evaluating the preliminarily screened leading edge point to obtain a leading edge target point; s3, the robot moves forwards along the target point; and S4, when the distance between the current position of the robot and the last position is larger than the preset movement radius or the travel distance is larger than the preset step length, judging whether an unknown area exists, if so, returning to the step S1, otherwise, stopping advancing and finishing map drawing. The invention has lower efficiency and consumes more computing resources under the condition of larger area.

The invention patent with publication number CN110806211A discloses a method, device and storage medium for robot autonomous exploration mapping, wherein the method comprises: acquiring environmental data acquired by a sensor, constructing an SLAM algorithm by utilizing synchronous positioning and a map, and identifying an area to be visited on a current estimated SLAM initial environmental map; planning paths based on an active exploration mode aiming at the identified access areas, selecting exploration paths from the planned paths according to a highest utility principle, and executing path exploration operation; and establishing an environment map corresponding to the autonomous exploration according to the path exploration result. The method only uses the clustering technology for the unknown areas and the idle areas, and is executed when the threshold value is reached, the scheme is simple, and the possible effect in the use of the actual scene is poor.

Disclosure of Invention

Aiming at the defects in the prior art, the invention provides a method and a system for an indoor robot to automatically explore and build a map.

According to the method and the system for automatically exploring and building the map by the indoor robot, the scheme is as follows:

in a first aspect, a method for an indoor robot to automatically explore mapping is provided, the method comprising:

step S1: selecting maps with different proportions according to scenes, setting angular points as reference points, traversing four angular points to determine map boundaries, and then exploring regions which are not explored inside;

step S2: acquiring occupied grids in an unexplored area, converting the occupied grids into laser data, acquiring laser radar data, combining the two data to generate a histogram, and selecting an area where a robot can pass;

step S3: obtaining candidate points by obtaining boundary line segments of idle grids and unknown grids, and grading and selecting an optimal target point according to rules;

step S4: setting preference parameters, modifying and selecting a target point strategy, adapting to different conditions, checking missing and filling up, and confirming and storing a map.

Preferably, the step S1 includes: the robot screen selects an autonomous exploration mode, small graphs, medium graphs and large graphs are respectively selected according to scenes, a built-in map after selection initializes a map with the same size, the resolution ratio is 0.05m/cell by default, and the position of the robot is the map center by default.

Preferably, the step S2 includes: acquiring grids occupied by unexplored areas in a map, regarding the grids as obstacles, and converting the grids into laser data scan _ obs; acquiring laser radar data scan _ original; combining the two data to form scan _ combination; generating a histogram from the scan _ combination data, wherein the closer the barrier distance is, the higher the barrier distance is, and similar mountains and valleys are generated;

and extracting a valley region, and filtering the valley region with the length smaller than the diameter of the robot.

Preferably, the step S3 includes: the method comprises the steps of obtaining an unknown area and an idle area in a map, extracting boundaries to obtain some boundary line segments, filtering line segments with the length smaller than a set threshold value, and taking the middle points of the line segments in the remaining line segments as candidate target points.

Preferably, the step S3 of scoring the target point rule includes:

1) if the target point does not return to-1 in the valley area, the candidate point is directly removed;

2) calculating the distance and the angle between the target point and the initial target point;

angle＝|θ1-θ2|

s1＝p1+p2

wherein, (x, y) is an initial target point coordinate, (x1, y1) is a selected coordinate, theta1 is an included angle between the initial target point and the robot, and theta2 is an included angle between the selected coordinate and the robot;

dist calculates the distance between the candidate target point and the initial target point coordinate;

the smaller the angle difference value is, the closer the candidate target point is to the position of the initial target point;

p1 is a value obtained by the obtained dist distance according to the formula (1), and when the distance between the selected candidate point and the starting target point is less than 1m, a constant value of 10 is directly assigned to p 1; when the calculated dist distance is between 1m and 100m, the interval obtained by p1 is between 10 m and 0.1 m, and is a decreasing function;

the p2 is a difference value obtained by two angles, the obtained angle difference value is within 1 degree according to the value obtained by the formula (2), the two points are on the same line, a constant value of 10 is directly assigned to p2, and when the difference value is between 1 and 180 degrees, the value of p2 is between 10 and 0.05, and is a decreasing function;

s1 is to add the two obtained scores of p1 and p2 to obtain a comprehensive score;

3) calculating the length of the line segment of the target point:

s2＝4*ln(x+1)，x∈(0，100)，s2∈(0，18.46)

x is the length of the obtained line segment, the second score s2 is calculated from above; the value of x is 0-100 m, s2 is 0-18.46 in the interval, and the longer the length of x, the larger s2 is.

Preferably, the step S4 includes: giving each score a weight, making preference settings,

score＝k1*s1+k2*s2

score is the sum of the first score and the second score, k1 and k2 as two fractional weight coefficients;

if the robot is allowed to walk away from the wide area, k2 is increased;

if the robot is led to the initial target point position, k1 is increased;

and selecting the target point with the highest score, reaching the target point after the target point is selected, and continuing to select the next target point.

Preferably, the method further comprises the following steps: when the distance between the position of the robot and the initial target point is less than 1.0m, the robot is considered to explore the edge of the preset map;

at this time, if the laser data is outside the map, the map needs to be expanded, the distance of the effective laser data is acquired, then the map is expanded in the direction to change the initial target point, the initial target point is set as the upper right corner point, then the target point acquisition process is continued, the robot is driven to advance to generate the map until the robot reaches the point, and the effective laser data is in the map, at this time, the direction search is considered to be finished.

Preferably, the method further comprises the following steps: and acquiring corner points of the map anticlockwise, completing exploration of the four corner points in sequence, returning to the upper right corner point, determining the boundary of the map, determining that an unknown area still exists in the map and is not explored, automatically adjusting a strategy according to the condition that the four corner points are explored, setting k1 to be 0, and completely selecting a target point according to the extracted boundary.

Preferably, the method further comprises the following steps: when the lengths of the obtained unknown boundary line segment and the obtained idle boundary line segment are smaller than a termination threshold value, ending the exploration;

and (5) the deployment personnel confirm the map, if corners are missed, the map is completed, and the map is stored.

In a second aspect, a system for automatic indoor robot exploration mapping is provided, the system comprising:

module M1: selecting maps with different proportions according to scenes, setting angular points as reference points, traversing four angular points to determine map boundaries, and then exploring regions which are not explored inside;

module M2: acquiring occupied grids in an unexplored area, converting the occupied grids into laser data, acquiring laser radar data, combining the two data to generate a histogram, and selecting an area where a robot can pass;

module M3: obtaining candidate points by obtaining boundary line segments of idle grids and unknown grids, and grading and selecting an optimal target point according to rules;

module M4: setting preference parameters, modifying and selecting a target point strategy, adapting to different conditions, checking missing and filling up, and confirming and storing a map.

Compared with the prior art, the invention has the following beneficial effects:

1. the invention adapts to different scenes by selecting maps with different proportions;

2. the invention solves the problem of random walking without purpose of the robot (usually, a closed area needs to be manually defined on an interface display, then random search is carried out in the area, the target point frequently oscillates back and forth or cannot be completely searched back and forth) by setting a corner point as an initial target point G0, and the problem is solved by utilizing G0 for guidance;

3. according to the method, the map boundary is determined by traversing four corner points, and then the unexplored area in the map is explored, so that the exploration efficiency is improved;

4. according to the invention, the area where the robot can pass is selected by combining the map obstacle and the laser data to generate the histogram, so that the efficiency is improved, (the histogram is low in calculation complexity, reliable and high in efficiency, and the idea of artificial potential field is used for reference);

5. the invention obtains candidate points by obtaining boundary line segments of idle grids and unknown grids (relative to a conventional method), and simultaneously scores and selects an optimal target point according to rules, two strategies are used, each strategy obtains a score, and then comprehensive addition is carried out to improve robustness;

6. the invention modifies the strategy of selecting the target point by setting the preference parameters, thereby adapting to different conditions (convenient adjustment, namely adjusting parameters k1 and k2 according to specific conditions).

Drawings

Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:

FIG. 1 is a sensor _ combination histogram;

FIG. 2 is a schematic diagram of extracting boundary line segments;

FIG. 3 is a schematic view of angles between a candidate target point and an initial target point;

FIG. 4 is a schematic view of angles between a candidate target point and an initial target point;

FIG. 5 is a schematic diagram of the upper right corner of the initial target point G0;

FIG. 6 is a schematic diagram of selecting an optimal target towards the initial target point G0;

FIG. 7 is a schematic diagram of the initial target point G0 being changed to the upper left corner;

FIG. 8 is a schematic view of the lower left corner of the initial target point G0;

FIG. 9 is a schematic view of the lower right corner of the initial target point G0;

FIG. 10 is a schematic diagram of the initial target point G0 returning to the upper right corner;

fig. 11 is a schematic diagram of the completion of the boundary search of four corner points and the search of the internal region.

Detailed Description

The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the present invention.

The embodiment of the invention provides a method for automatically exploring and building an image for an indoor robot. Referring to fig. 1, the method includes:

step S1: selecting maps with different proportions according to scenes, setting angular points as reference points, traversing four angular points to determine map boundaries, and then exploring regions which are not explored inside;

step S2: acquiring occupied grids in an unexplored area, converting the occupied grids into laser data, acquiring laser radar data, combining the two data to generate a histogram, and selecting an area where a robot can pass;

step S3: obtaining candidate points by obtaining boundary line segments of idle grids and unknown grids, and grading and selecting an optimal target point according to rules;

step S4: setting preference parameters, modifying and selecting a target point strategy, adapting to different conditions, checking missing and filling up, and confirming and storing a map.

Specifically, step S1 includes: the robot screen selects an autonomous exploration mode, small graphs, medium graphs and large graphs are respectively selected according to scenes, a built-in map after selection initializes a map with the same size, the resolution ratio is 0.05m/cell by default, and the position of the robot is the map center by default.

In step S2: acquiring grids occupied by unexplored areas in a map, regarding the grids as obstacles, and converting the grids into laser data scan _ obs; acquiring laser radar data scan _ original; combining the two data to form scan _ combination; generating a histogram from the scan _ combination data, wherein the closer the barrier distance is, the higher the barrier distance is, and similar mountains and valleys are generated; and extracting a valley region, and filtering the valley region with the length smaller than the diameter of the robot.

In step S3: the method comprises the steps of obtaining an unknown area and an idle area in a map, extracting boundaries to obtain some boundary line segments, filtering line segments with the length smaller than a set threshold value, and taking the middle points of the line segments in the remaining line segments as candidate target points.

The target point rule scoring in step S3 includes:

1) if the target point does not return to-1 in the valley area, the candidate point is directly removed;

2) calculating the distance and the angle between the target point and the initial target point;

angle＝|θ1-θ2|

s1＝p1+p2

wherein, (x, y) is an initial target point coordinate, (x1, y1) is a selected coordinate, theta1 is an included angle between the initial target point and the robot, and theta2 is an included angle between the selected coordinate and the robot;

referring to fig. 3 and 4, dist calculates the distance between the candidate target point and the initial target point coordinate, and among a group of candidate target points, the closer the candidate target point is to the initial target point, the closer the candidate coordinate is;

angle represents the difference between theta1 and theta2, with smaller differences resulting in candidate target points closer to the location of initial target point G0.

p1 is obtained from the obtained dist distance according to the formula (1), and when the distance between the selected candidate point and the starting target point G0 is less than 1m, the p1 is directly assigned with a constant value of 10; when the computed dist distance is between 1m and 100m, the interval obtained by p1 is between 10 m and 0.1 m, and is a decreasing function.

The p2 is obtained by a difference value obtained by two angles according to the formula (2), the obtained angle difference value is within 1 degree, the two points are on the same line, a constant value of 10 is directly assigned to the p2, and when the difference value is between 1 and 180 degrees, the value of p2 is between 10 and 0.05, and is a decreasing function.

s1 is to add the obtained p1 and p2 to obtain a composite score.

3) Calculating the length of the line segment of the target point:

s2＝4*ln(x+1)，x∈(0，100)，s2∈(0，18.46)

x is the length of the obtained line segment, the second score s2 is calculated from above; the value of x is 0-100 m, s2 is 0-18.46 in the interval, and the longer the length of x, the larger s2 is.

In step S4, a weight is given to each score, preference setting is made,

score＝k1*s1+k2*s2

wherein score is the sum of the first score and the second score, and k1 and k2 are two fractional weight coefficients;

k1 and k2 represent the preference of picking targets, k1 s1 represents the sub-fraction of candidate target points towards the initial target point G0, k2 s2 represents the sub-fraction of the length of the free and free grid segments, raising the total score by raising k1 s1 indicates that targets are more likely to pick towards the initial target point G0; increasing the total score by increasing k2 × s2 indicates that the target points on the longer segments of the free and free grid segments are more likely to be selected, which represents a more open area. Default k1 equals 1 and k2 equals 1.

If the robot is allowed to walk away from the wide area, k2 is increased;

if the robot is enabled to face the initial target point position as far as possible, k1 is increased;

and selecting the target point with the highest score, reaching the target point after the target point is selected, and continuing to select the next target point.

When the distance between the position of the robot and the initial target point is less than 1.0m, the robot is considered to explore the edge of the preset map; at this time, if the laser data is outside the map, the map needs to be expanded, the distance of the effective laser data is acquired, then the map is expanded in the direction to change the initial target point, the initial target point is set as the upper right corner point, then the target point acquisition process is continued, the robot is driven to advance to generate the map until the robot reaches the point, and the effective laser data is in the map, at this time, the direction search is considered to be finished.

And acquiring corner points of the map anticlockwise, completing exploration of the four corner points in sequence, returning to the upper right corner point, determining the boundary of the map, determining that an unknown area still exists in the map and is not explored, automatically adjusting a strategy according to the condition that the four corner points are explored, setting k1 to be 0, and completely selecting a target point according to the extracted boundary.

When the lengths of the acquired unknown and free boundary line segments are both smaller than the termination threshold (1.4m is the general corridor width), the exploration is ended. And (5) the deployment personnel confirm the map, if corners are missed, the map is completed, and the map is stored.

Next, the present invention will be described in more detail.

(1) The robot screen selects an autonomous exploration mode, small maps (20 m), medium maps (50 m), large maps (100 m), built-in maps after selection are initialized to be the same size, the resolution is 0.05m/cell by default, the position of the robot is the map center by default, the global planning allows unknown area planning, the local planning tracks the global path, and the robot screen has dynamic obstacle avoidance capability.

(2) Taking the large map as an example, the initial target G0 is selected as the upper right corner (50.0 ), and the engineering selection generally subtracts one map resolution, so as to facilitate the normal operation of the path planning (49.95 ). After the paths of the robot and the target point are obtained, the robot starts to move towards the target, meanwhile, a built-in slam program generates occupied and idle maps, and the target point selection strategy is started to be executed after the robot runs for a distance (0.5m) or an angle (10 degrees).

(3) Acquiring the occupied grids in the map and converting the occupied grids into laser data scan _ obs by referring to the data 1; acquiring laser radar data scan _ original; combining the two data to form scan _ combination; the scan _ combine data is generated into a histogram, and the closer the obstacle distance is, the higher the height is, the like high mountains and low valleys are generated. And extracting a valley region, and filtering the valley region with the length smaller than the diameter of the robot.

(4) Referring to fig. 2, an unknown area and an idle area in a map are obtained, boundaries are extracted to obtain some boundary line segments, line segments with lengths smaller than a threshold value are filtered, and the middle points of the line segments in the remaining line segments are used as candidate target points.

(5) And scoring the target point:

1) if the target point does not return to-1 in the valley area, the candidate point is directly removed;

2) the distance and angle between the target point and the initial target point G0 are calculated, and the smaller the distance and angle errors are, the robot does not change the direction greatly, and the robot fits the guiding direction of the G0 point more.

angle＝|θ1-θ2|

s1＝p1+p2

Wherein, (x, y) is the G0 coordinate of the initial target point, (x1, y1) is the selected coordinate, theta1 is the included angle between the initial target point and the robot, and theta2 is the included angle between the selected coordinate and the robot;

referring to fig. 3 and 4, dist calculates the distance between the candidate target point and the initial target point coordinate, and among a group of candidate target points, the closer the candidate target point is to the initial target point, the closer the candidate coordinate is;

angle represents the difference between theta1 and theta2, with smaller differences resulting in candidate target points closer to the location of initial target point G0.

p1 is obtained according to formula (1) from the obtained dist distance, and when the distance between the selected candidate point and the starting target point G0 is less than 1m, the direct p1 is assigned with a constant value of 10; when the computed dist distance is between 1m and 100m, the interval obtained by p1 is between 10 m and 0.1 m, and is a decreasing function.

The p2 is obtained by a difference value obtained by two angles according to the formula (2), the obtained angle difference value is within 1 degree, the two points are on the same line, a constant value of 10 is directly assigned to the p2, and when the difference value is between 1 and 180 degrees, the value of p2 is between 10 and 0.05, and is a decreasing function.

s1 is to add the obtained p1 and p2 to obtain a composite score.

3) Calculating the length of the line segment of the target point:

s2＝4*ln(x+1)，x∈(0，100)，s2∈(0，18.46)

x is the length of the obtained line segment, the second score s2 is calculated from above; the value of x is 0-100 m, s2 is 0-18.46 in the interval, and the longer the length of x, the larger s2 is.

In step S4, a weight is given to each score, preference setting is made,

score＝k1*s1+k2*s2

wherein score is the sum of the first score and the second score, and k1 and k2 are two fractional weight coefficients;

k1 and k2 represent the preference of picking targets, k1 s1 represents the sub-fraction of candidate target points towards the initial target point G0, k2 s2 represents the sub-fraction of the length of the free and free grid segments, raising the total score by raising k1 s1 indicates that targets are more likely to pick towards the initial target point G0; increasing the total score by increasing k2 × s2 indicates that the target points on the longer segments of the free and free grid segments are more likely to be selected, which represents a more open area. Default k1 equals 1 and k2 equals 1.

If one wants to get the robot as far away as possible, k2 is raised;

if one wants to orient the robot as much as possible towards the initial target point G0, k1 is increased;

and selecting the target point with the highest score, and continuing to select the target point after the target point is selected and reaching the target point to prevent the target from oscillating due to switching back and forth.

(7) And when the distance between the position of the robot and the initial target point G0 is less than 1.0m, the robot is considered to have explored the edge of a preset map, at this time, if the laser data is outside the map, the map needs to be expanded, the distance of the effective laser data is acquired, then the map is expanded in the direction to change the initial target point G0, the initial target point G0 point is set as the upper right corner point, then the target point acquiring process is continued, the robot is driven to advance to generate the map until the robot reaches the point, and the effective laser data are all in the map. At this time, the direction search is considered to be completed.

(8) Referring to fig. 5, 6, 7, 8, 9, 10 and 11, map corners are acquired counterclockwise, four corners are searched in sequence, and the boundaries of the map are determined when the map returns to the upper-right corner point, at this time, an unknown area possibly still exists in the map and is not searched, the strategy is automatically adjusted according to the judgment condition that the four corners are searched, k1 is set to 0, and a target point is selected completely according to the extracted boundaries.

(9) And when the acquired unknown and free boundary line lengths are both smaller than the set threshold (1.4m is the general corridor width), ending the search.

(10) And the deployment personnel confirm the map, complete the map if corners are missed, and store the map.

The embodiment of the invention provides a method and a system for automatically exploring and building an image for an indoor robot, which utilize an initial target point as orientation guidance, use the principle of a potential field method to generate a histogram model for an obstacle, and select a feasible area; simultaneously extracting segments of an unknown region and an idle region, and selecting candidate target points; finally obtaining an optimal target point by comparing the distance and the direction of the guide point and judging whether the feasible region can pass the robot or not and grading the dimensions such as the length of the line segment where the target point is located; and searching for the target in the map after the boundary area exploration is finished, finally meeting a termination threshold value, finishing the autonomous exploration, confirming the map by field personnel, and checking for missing and filling up the gap.

Those skilled in the art will appreciate that, in addition to implementing the system and its various devices, modules, units provided by the present invention as pure computer readable program code, the system and its various devices, modules, units provided by the present invention can be fully implemented by logically programming method steps in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Therefore, the system and various devices, modules and units thereof provided by the invention can be regarded as a hardware component, and the devices, modules and units included in the system for realizing various functions can also be regarded as structures in the hardware component; means, modules, units for performing the various functions may also be regarded as structures within both software modules and hardware components for performing the method.

The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.

Claims (10)

1. A method for automatically exploring and building a map by an indoor robot is characterized by comprising the following steps:

step S1: selecting maps with different proportions according to scenes, setting angular points as reference points, traversing four angular points to determine map boundaries, and then exploring regions which are not explored inside;

step S2: acquiring occupied grids in an unexplored area, converting the occupied grids into laser data, acquiring laser radar data, combining the two data to generate a histogram, and selecting an area where a robot can pass;

step S3: obtaining candidate points by obtaining boundary line segments of idle grids and unknown grids, and grading and selecting an optimal target point according to rules;

step S4: setting preference parameters, modifying and selecting a target point strategy, adapting to different conditions, checking missing and filling up, and confirming and storing a map.

2. The method for indoor robot to automatically search for mapping as claimed in claim 1, wherein the step S1 comprises: the robot screen selects an autonomous exploration mode, small graphs, medium graphs and large graphs are respectively selected according to scenes, a built-in map after selection initializes a map with the same size, the resolution ratio is 0.05m/cell by default, and the position of the robot is the map center by default.

3. The method for indoor robot to automatically search for mapping as claimed in claim 1, wherein the step S2 comprises: acquiring grids occupied by unexplored areas in a map, regarding the grids as obstacles, and converting the grids into laser data scan _ obs; acquiring laser radar data scan _ original; combining the two data to form scan _ combination; generating a histogram from the scan _ combination data, wherein the closer the barrier distance is, the higher the barrier distance is, and similar mountains and valleys are generated;

and extracting a valley region, and filtering the valley region with the length smaller than the diameter of the robot.

4. The method for indoor robot to automatically search for mapping as claimed in claim 1, wherein the step S3 comprises: the method comprises the steps of obtaining an unknown area and an idle area in a map, extracting boundaries to obtain some boundary line segments, filtering line segments with the length smaller than a set threshold value, and taking the middle points of the line segments in the remaining line segments as candidate target points.

5. The method for indoor robot automatic exploration mapping according to claim 1, wherein said step S3 of scoring target point rule comprises:

1) if the target point does not return to-1 in the valley area, the candidate point is directly removed;

2) calculating the distance and the angle between the target point and the initial target point;

angle＝|θ1-θ2|

s1＝p1+p2

wherein, (x, y) is an initial target point coordinate, (x1, y1) is a selected coordinate, theta1 is an included angle between the initial target point and the robot, and theta2 is an included angle between the selected coordinate and the robot;

dist calculates the distance between the candidate target point and the initial target point coordinate;

the smaller the angle difference value is, the closer the candidate target point is to the position of the initial target point;

p1 is a value obtained by the obtained dist distance according to the formula (1), and when the distance between the selected candidate point and the starting target point is less than 1m, a constant value of 10 is directly assigned to p 1; when the calculated dist distance is between 1m and 100m, the interval obtained by p1 is between 10 m and 0.1 m, and is a decreasing function;

the p2 is a difference value obtained by two angles, the obtained angle difference value is within 1 degree according to the value obtained by the formula (2), the two points are on the same line, a constant value of 10 is directly assigned to p2, and when the difference value is between 1 and 180 degrees, the value of p2 is between 10 and 0.05, and is a decreasing function;

s1 is to add the two obtained scores of p1 and p2 to obtain a comprehensive score;

3) calculating the length of the line segment of the target point:

s2＝4*ln(x+1)，x∈(0,100),s2∈(0,18.46)

x is the length of the obtained line segment, the second score s2 is calculated from above; the value of x is 0-100 m, s2 is 0-18.46 in the interval, and the longer the length of x, the larger s2 is.

6. The method for indoor robot to automatically explore mapping according to claim 5, wherein the step S4 comprises: giving each score a weight, making preference settings,

score＝k1*s1+k2*s2

score is the sum of the first score and the second score, k1 and k2 as two fractional weight coefficients;

if the robot is allowed to walk away from the wide area, k2 is increased;

if the robot is led to the initial target point position, k1 is increased;

and selecting the target point with the highest score, reaching the target point after the target point is selected, and continuing to select the next target point.

7. The method for indoor robot to automatically explore mapping according to claim 6, further comprising: when the distance between the position of the robot and the initial target point is less than 1.0m, the robot is considered to explore the edge of the preset map;

at this time, if the laser data is outside the map, the map needs to be expanded, the distance of the effective laser data is acquired, then the map is expanded in the direction to change the initial target point, the initial target point is set as the upper right corner point, then the target point acquisition process is continued, the robot is driven to advance to generate the map until the robot reaches the point, and the effective laser data is in the map, at this time, the direction search is considered to be finished.

8. The method for indoor robot to automatically explore mapping according to claim 7, further comprising: and acquiring corner points of the map anticlockwise, completing exploration of the four corner points in sequence, returning to the upper right corner point, determining the boundary of the map, determining that an unknown area still exists in the map and is not explored, automatically adjusting a strategy according to the condition that the four corner points are explored, setting k1 to be 0, and completely selecting a target point according to the extracted boundary.

9. The method for indoor robot auto-discovery mapping according to claim 8, further comprising: when the lengths of the obtained unknown boundary line segment and the obtained idle boundary line segment are smaller than a termination threshold value, ending the exploration;

and (5) the deployment personnel confirm the map, if corners are missed, the map is completed, and the map is stored.

10. A system for automatically exploring and building a graph by an indoor robot is characterized by comprising:

module M1: selecting maps with different proportions according to scenes, setting angular points as reference points, traversing four angular points to determine map boundaries, and then exploring regions which are not explored inside;

module M2: acquiring occupied grids in an unexplored area, converting the occupied grids into laser data, acquiring laser radar data, combining the two data to generate a histogram, and selecting an area where a robot can pass;

module M3: obtaining candidate points by obtaining boundary line segments of idle grids and unknown grids, and grading and selecting an optimal target point according to rules;

module M4: setting preference parameters, modifying and selecting a target point strategy, adapting to different conditions, checking missing and filling up, and confirming and storing a map.

Images