Obstacle avoidance method, system and storage medium for hotel intelligent robot

Abstract

The invention discloses an obstacle avoidance method, system and storage medium for a hotel intelligent robot, relates to the technical field of robot obstacle avoidance control, and solves the problem that the prior art builds a route prediction model, and the processes of model construction, learning and prediction increase the computing power demand of the robot, thereby causing computing delays and affecting the real-time and accuracy of the robot's obstacle avoidance; the invention also partitions the operation area according to obstacle information and frequency-time thresholds; obtains a first path of the robot according to the partition result; judges whether the obstacles appearing on the first path are dynamic obstacles; if so, calculates the active avoidance rate of the corresponding dynamic obstacles based on historical avoidance data, and obtains a third path according to the active avoidance rate; if not, judges whether the current obstacle is a surmountable obstacle; if so, issues a continue driving instruction; if not, replans and obtains a second path; the invention simplifies the avoidance algorithm, reduces the computing power of the robot during operation, and saves time cost.

Description

Technical Field

[0001] The invention belongs to the field of robot control and relates to a robot obstacle avoidance technology, in particular to an obstacle avoidance method, system and storage medium for a hotel intelligent robot. Background Art

[0002] Building a hotel intelligent robot obstacle avoidance system is beneficial. Through the precise obstacle avoidance system, the robot can sense and avoid obstacles in real time, thereby significantly reducing failures and downtime caused by collisions; the obstacle avoidance system can dynamically adjust the robot's driving path according to real-time obstacle information to ensure that the robot can complete the task in the shortest and safest path; the obstacle avoidance system reduces the time wasted by the robot due to obstacle avoidance, enabling it to complete tasks faster, such as delivering meals and items; the obstacle avoidance system can ensure that the robot will not collide with hotel employees or guests during operation, thereby ensuring personnel safety.

[0003] The prior art (CN115328130A) discloses a robot obstacle avoidance system and an obstacle avoidance method thereof, including a terminal processor, an input end of the terminal processor being bidirectionally electrically connected to an environment acquisition module, and an output end of the terminal processor being electrically connected to an environment model building module; indoor environment images such as shopping malls are collected by the environment image acquisition module and injected into the environment model building module, the environment model building module imports the data and generates it into a route prediction model, the neural network prediction model performs self-learning and judgment according to the robot's travel route, selects a path with fewer obstacles through the route prediction model, and controls the servo motor to start and move through a drive circuit; the prior art constructs a route prediction model through the environment model building module, and performs self-learning and judgment based on the neural network prediction model, thereby selecting a path with fewer obstacles through the route prediction model, and the process of model construction, learning and prediction increases the computing power demand of the robot, thereby causing calculation delays, affecting the real-time and accuracy of the robot's obstacle avoidance.

[0004] The present invention provides an obstacle avoidance method, system and storage medium for a hotel intelligent robot to solve the above technical problems. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an obstacle avoidance method, system and storage medium for a hotel intelligent robot, which are used to solve the technical problem that the prior art builds a route prediction model through an environmental model establishment module, and the process of model construction, learning and prediction increases the computing power demand of the robot, thereby causing calculation delays and affecting the real-time and accuracy of the robot's obstacle avoidance.

[0006] To achieve the above object, a first aspect of the present invention provides an obstacle avoidance method for a hotel intelligent robot, comprising:

[0007] Obtain the operating area of ​​the robot and obstacle information of obstacles in the operating area, and partition the operating area according to the obstacle information and the frequency-time threshold to obtain the partition result; perform path planning according to the partition result to obtain the first path of the robot;

[0008] Determine whether the obstacle currently appearing on the first path is a dynamic obstacle;

[0009] If yes, then the type of the currently appearing dynamic obstacle is obtained, the active avoidance rate of the corresponding dynamic obstacle is obtained based on the type of the dynamic obstacle, a third path is obtained according to the active avoidance rate, and the robot is controlled to travel along the third path;

[0010] If no, determine whether the current fixed obstacle is surmountable; if yes, issue a continue driving instruction; if no, replan to obtain a second path, and drive along the second path.

[0011] Preferably, partitioning the operation area according to the obstacle information and the frequency-time threshold includes:

[0012] Divide the operating area into several sub-blocks based on the type of obstacles and number them;

[0013] Determine whether there is an obstacle in the sub-block; if yes, mark the corresponding sub-block as an obstacle area; if no, mark the corresponding sub-block as an obstacle-free area; and count the activity time of obstacles in the obstacle area through the database;

[0014] Determine whether the activity time of the obstacle in the obstacle area is 0;

[0015] If yes, the corresponding sub-block is marked as a fixed obstacle area;

[0016] If not, determine whether the activity time is lower than the frequency-time threshold a1; if yes, mark the corresponding block as a low-frequency obstacle area; if not, mark the corresponding block as a high-frequency obstacle area; wherein, 0<a1<24.

[0017] The present invention realizes refined management of the operating area by dividing the operating area into several blocks and making a detailed judgment on the obstacle situation of each block; at the same time, it also helps the robot or related system to avoid obstacles more accurately when planning paths or performing tasks, thereby improving safety and efficiency; the blocks are classified into obstacle areas and obstacle-free areas, so that the path planning algorithm can select the optimal path according to the characteristics of each block, thereby improving the flexibility of path planning.

[0018] Preferably, the method for obtaining the frequency-time threshold a1 includes:

[0019] Obtain several groups of historical obstacle area existence times through the database and mark them as Tr, where r is the number of days, r=1, 2, 3, ..., n; n is a positive integer;

[0020] The total duration of obstacles in the first special node is marked as Tr1, and the total duration of obstacles in the second special node is marked as Tr2; the total duration of obstacles except Tr1 and Tr2 is marked as Tr3; wherein, the first special node is the time node when the operation area is in the activity period; the second special node is the node when the operation area is in the meal time of the day except the first special node; the third special node is the time node when obstacles exist in the operation area except Tr1 and Tr2;

[0021] Construct a frequency-time threshold weight table according to the first special node, the second special node, and the third special node; obtain the frequency-time threshold weight from the frequency-time threshold weight table and mark it as Qm, where m is the number of frequency-time threshold weights, m=1, 2, 3;

[0022] Determine whether the operation area of ​​the robot is in the first special node;

[0023] Yes, then by the formula The frequency-time threshold a1 is calculated;

[0024] If not, then determine whether the robot's operating area is in the second special node; if yes, then use the formula Calculate the frequency-time threshold a1; if not, use the formula The frequency-time threshold a1 is calculated; wherein n1, n2, n3 represent the total statistical days under each special node; r1, r2, r3 represent the statistical days under each special node; n1, n2, n3, r1, r2, r3 are all positive integers greater than 1.

[0025] The present invention records and analyzes in detail the time when obstacles existed in historical obstacle areas, accurately understands the distribution of obstacles in each time period and special nodes, and helps the robot avoid the period when high obstacles appear and choose a safer route when planning the path; by constructing a frequency-time threshold weight table, it is helpful to evaluate the impact of obstacles on the operation of the robot in each time period and special nodes, and also helps the robot to use resources more efficiently.

[0026] Preferably, performing path planning according to the partitioning conditions to obtain the first path of the robot includes:

[0027] S100: retrieve obstacle-free areas, fixed obstacle areas, low-frequency obstacle areas, and high-frequency obstacle areas; mark the obstacle-free areas and fixed obstacle areas as A1x, the low-frequency obstacle areas as A2d, and the high-frequency obstacle areas as A3g; where x, d, and g represent the number of corresponding areas, respectively; x≥0, d≥0, g≥0, and x+d+g≠0;

[0028] S120: Obtain the starting point and destination of the robot, obtain the number of all traversable paths through the path planning algorithm, and record it as u; count the type of area and the corresponding number of areas passed by each traversable path;

[0029] S130: Assign scores to the regions corresponding to A1x, A2d, and A3g as qf1, qf2, and qf3, respectively; wherein qf1 is the score of the first region, qf2 is the score of the second region, and qf3 is the score of the third region; and 0≤qf1<qf2<qf3≤100;

[0030] The total score of each area that the robot has gone through under each path is calculated by the calculation formula ZQFu=x×qf1+d×qf2+g×qf3;

[0031] S140: Obtain the time required for the robot to pass through the corresponding area and record it as ST1, ST2, and ST3 respectively;

[0032] The total time the robot takes on each path is calculated using the formula ZSCu=x×ST1+d×ST2+g×ST3;

[0033] S150: assign the total duration of the interval [(ZSCmax+ZSCmin) / 2, ZSCmax] to sf1;

[0034] Assign the total duration in the interval [(ZSCmax+ZSCmin) / 3, (ZSCmax+ZSCmin) / 2) to sf2;

[0035] Assign the total duration in the interval [0, (ZSCmax+ZSCmin) / 3) to sf3; where ZSCmax represents the total duration of the robot corresponding to the longest path under all paths; ZSCmin represents the total duration of the robot corresponding to the shortest path under all paths; where sf1 is the first duration value, sf2 is the second duration value, and sf3 is the third duration value; and 0≤sf1<sf2<sf3≤100;

[0036] S160: Mark the sum of the total scores of each area traversed by the robot under each path and the total time the robot has traversed under each path as the first total score, and mark the path corresponding to the lowest first total score as the first path of the robot; sort the paths except the first path from low to high according to the total scores; wherein p represents the total number of areas traversed by the robot under each path, and p=x+d+g.

[0037] The present invention takes various types of areas into consideration and assigns each type of area a score. The process can more comprehensively evaluate the advantages and disadvantages of each path, helping the robot to select a route with fewer obstacles and greater efficiency. By calculating the total score and total time the robot has traveled through each area under each path, and sorting them based on these values, the process can automatically select the optimal path, reducing the need for human intervention and improving the efficiency and accuracy of path planning.

[0038] Preferably, the active avoidance rate of the corresponding dynamic obstacle is calculated based on the historical avoidance data, including:

[0039] S210: Retrieving historical avoidance data from a database;

[0040] S220: Count the number of times the dynamic obstacle appears ahead and the number of times the obstacle is actively avoided, and mark the ratio of the number of times the dynamic obstacle appears to the number of times the obstacle is actively avoided as the active avoidance rate.

[0041] It should be noted that the avoidance threshold is set by personnel in this field according to actual working conditions.

[0042] When the active avoidance rate exceeds the set avoidance threshold, the robot chooses to wait on the spot instead of taking risks to avoid. This strategy helps to avoid collisions in highly dynamic or complex environments, thereby protecting the safety of the robot and the surrounding environment.

[0043] Preferably, the step of acquiring the third path according to the active avoidance rate includes:

[0044] S310: extract the active avoidance rate, and compare the active avoidance rate with the avoidance threshold; when the active avoidance rate is greater than the avoidance threshold, the robot waits in place; otherwise, jump to S320;

[0045] S320: Retrieve the robot whose active avoidance rate exceeds the avoidance threshold and is waiting at the same place, and set the waiting time threshold;

[0046] S330: Go to steps S120-S160, obtain the second total score, and use the path corresponding to the lowest value of the second total score as the third preliminary path of the robot;

[0047] S340: Determine whether the waiting time of the robot is greater than the waiting time threshold;

[0048] If yes, it is determined whether the forward distance between the current dynamic obstacle and the forward robot exceeds the normal distance; if yes, the robot drives along the first path; if no, the current obstacle does not actively avoid the robot; the judgment result is included in the database, and the active avoidance rate of the corresponding obstacle is recalculated, and the process goes to step S350;

[0049] If not, continue to wait;

[0050] S350: assigning the third preliminary path to the third path;

[0051] The waiting time threshold is obtained by obtaining the time required for performing steps S120 to S160 for several times from a database and taking an average value, and using the average value as the waiting time threshold.

[0052] It should be noted that the forward distance is the distance between the current dynamic obstacle and the moving robot.

[0053] Preferably, the re-planning to obtain the second path includes:

[0054] Get the robot's current location and destination, re-enter step S120, obtain the sum of the total scores of each area traversed by the robot under each path and the total time the robot has traversed under each path, and use the path corresponding to the minimum value of the sum of the assigned values ​​as the second path of the robot.

[0055] To achieve the above-mentioned object, the second aspect of the present invention provides an obstacle avoidance system for a hotel intelligent robot, comprising: a database and a path selection module and an obstacle avoidance module connected thereto;

[0056] The database is used to store the types of obstacles that have been actively avoided in the past, the existence time of each obstacle, and the historical avoidance data statistically calculated based on the types of obstacles, wherein the types of obstacles include fixed obstacles and dynamic obstacles;

[0057] The path selection module: obtains the operation area of ​​the robot and obstacle information of obstacles in the operation area, and partitions the operation area according to the obstacle information and the frequency-time threshold to obtain a partition result; performs path planning according to the partition result to obtain a first path of the robot;

[0058] The obstacle avoidance module: determines whether the obstacle currently appearing on the first path is a dynamic obstacle;

[0059] If yes, then the type of the currently appearing dynamic obstacle is obtained, the active avoidance rate of the corresponding dynamic obstacle is obtained based on the type of the dynamic obstacle, a third path is obtained according to the active avoidance rate, and the robot is controlled to travel along the third path;

[0060] If no, determine whether the current fixed obstacle is surmountable; if yes, issue a continue driving instruction; if no, replan to obtain a second path, and drive along the second path.

[0061] It should be noted that the operating area is set according to the needs of the hotel; and the type of obstacle is obtained by identifying the content in the obstacle image through image recognition technology.

[0062] Preferably, the system further includes: an information linkage module: used to install a data acquisition device, collect and determine the type of obstacles through image recognition technology; and update the type of obstacles and historical avoidance data in the database; wherein the data acquisition device includes: an image acquisition device and a laser radar.

[0063] The third aspect of the present invention provides an obstacle avoidance storage medium for a hotel intelligent robot, on which a computer-readable storage medium is stored. When the computer-readable storage medium is executed by a processor, an obstacle avoidance system for a hotel intelligent robot as described in the first aspect above is implemented.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] 1. The present invention realizes refined management of the operating area by dividing the operating area into several blocks and making a detailed judgment on the obstacle situation of each block; the blocks are classified into obstacle areas and obstacle-free areas, so that the path planning algorithm can select the optimal path according to the characteristics of each block, thereby improving the flexibility of path planning; by calculating the total score and total time of each area traversed by the robot under each path, and sorting them based on these values, the process can automatically select the optimal path, thereby simplifying the avoidance algorithm, reducing the computing power of the robot during operation, enabling the robot to make decisions faster, and saving time costs.

[0066] 2. The present invention obtains the operating area and obstacle information through the path selection module, and partitions the operating area according to this information. The system can plan the initial path of the robot more intelligently, which helps to reduce the risk of collision and improve the operating efficiency of the robot; the obstacle avoidance module can identify dynamic obstacles on the first path, and calculate the active avoidance rate based on historical avoidance data, which means that the system can predict and respond to the appearance of dynamic obstacles based on past experience, thereby improving the robot's obstacle avoidance ability; when encountering a dynamic obstacle, the system can re-plan a third path based on the active avoidance rate; and when encountering a fixed but insurmountable obstacle, the system can also re-plan a second path. This flexible path adjustment strategy helps to ensure that the robot can find the best driving route when facing various obstacles. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0068] Figure 1 Schematic diagram of the steps of the obstacle avoidance process of the present invention;

[0069] Figure 2 A schematic diagram of a specific process for path selection of the present invention;

[0070] Figure 3 It is a schematic diagram of a specific process of obstacle avoidance of the present invention;

[0071] Figure 4 It is a schematic diagram of the relationship between the modules included in the present invention. DETAILED DESCRIPTION

[0072] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0073] See also Figure 1 The first aspect of the present invention provides an obstacle avoidance method for a hotel intelligent robot, comprising:

[0074] Obtain the operating area of ​​the robot and obstacle information of obstacles in the operating area, and partition the operating area according to the obstacle information and the frequency-time threshold to obtain the partition result; perform path planning according to the partition result to obtain the first path of the robot;

[0075] Determine whether the obstacle currently appearing on the first path is a dynamic obstacle;

[0076] If yes, then the type of the currently appearing dynamic obstacle is obtained, the active avoidance rate of the corresponding dynamic obstacle is obtained based on the type of the dynamic obstacle, a third path is obtained according to the active avoidance rate, and the robot is controlled to travel along the third path;

[0077] If no, determine whether the current fixed obstacle is surmountable; if yes, issue a continue driving instruction; if no, replan to obtain a second path, and drive along the second path.

[0078] See also Figure 2,The specific process of path selection is to divide the operation area into several ,sub-blocks and number them based on the types of obstacles;

[0079] Determine whether there is an obstacle in the sub-block; if yes, mark the corresponding sub-block as an obstacle area; if no, mark the corresponding sub-block as an obstacle-free area; and count the activity time of obstacles in the obstacle area through the database;

[0080] Obtain several groups of historical obstacle area existence times through the database and mark them as Tr, where r is the number of days, r=1, 2, 3, ..., n; n is a positive integer;

[0081] The total duration of obstacles in the first special node is marked as Tr1, and the total duration of obstacles in the second special node is marked as Tr2; the total duration of obstacles except Tr1 and Tr2 is marked as Tr3; wherein, the first special node is the time node when the operation area is in the activity period; the second special node is the node when the operation area is in the meal time of the day except the first special node; the third special node is the time node when obstacles exist in the operation area except Tr1 and Tr2;

[0082] Construct a frequency-time threshold weight table according to the first special node, the second special node, and the third special node; obtain the frequency-time threshold weight from the frequency-time threshold weight table and mark it as Qm, where m is the number of frequency-time threshold weights, m=1, 2, 3;

[0083] Determine whether the operation area of ​​the robot is in the first special node;

[0084] Yes, then by the formula The frequency-time threshold a1 is calculated;

[0085] If not, then determine whether the robot's operating area is in the second special node; if yes, then use the formula Calculate the frequency-time threshold a1; if not, use the formula The frequency-time threshold a1 is calculated; wherein n1, n2, n3 represent the total number of statistical days under each special node; r1, r2, r3 represent the number of statistical days under each special node; n1, n2, n3, r1, r2, r3 are all positive integers greater than 1;

[0086] Retrieve the frequency-time threshold a1 to determine whether the activity time of the obstacle in the obstacle area is 0;

[0087] If yes, the corresponding sub-block is marked as a fixed obstacle area;

[0088] If not, it is determined whether the activity time is lower than the frequency threshold a1; if yes, the corresponding block is marked as a low-frequency obstacle area; if not, the corresponding block is marked as a high-frequency obstacle area; wherein 0<a1<24;

[0089] S100: retrieve obstacle-free areas, fixed obstacle areas, low-frequency obstacle areas, and high-frequency obstacle areas; mark the obstacle-free areas and fixed obstacle areas as A1x, the low-frequency obstacle areas as A2d, and the high-frequency obstacle areas as A3g; where x, d, and g represent the number of corresponding areas, respectively; x≥0, d≥0, g≥0, and x+d+g≠0;

[0090] S120: Obtain the starting point and destination of the robot, obtain the number of all traversable paths through the path planning algorithm, and record it as u; count the type of area and the corresponding number of areas passed by each traversable path;

[0091] S130: Assign scores to the regions corresponding to A1x, A2d, and A3g as qf1, qf2, and qf3, respectively; wherein qf1 is the score of the first region, qf2 is the score of the second region, and qf3 is the score of the third region; and 0≤qf1<qf2<qf3≤100;

[0092] The total score of each area that the robot has gone through under each path is calculated by the calculation formula ZQFu=x×qf1+d×qf2+g×qf3;

[0093] S140: Obtain the time required for the robot to pass through the corresponding area and record it as ST1, ST2, and ST3 respectively;

[0094] The total time the robot takes on each path is calculated using the formula ZSCu=x×ST1+d×ST2+g×ST3;

[0095] S150: assign the total duration of the interval [(ZSCmax+ZSCmin) / 2, ZSCmax] to sf1;

[0096] Assign the total duration in the interval [(ZSCmax+ZSCmin) / 3, (ZSCmax+ZSCmin) / 2) to sf2;

[0097] Assign the total duration in the interval [0, (ZSCmax+ZSCmin) / 3) to sf3; where ZSCmax represents the total duration of the robot corresponding to the longest path under all paths; ZSCmin represents the total duration of the robot corresponding to the shortest path under all paths; where sf1 is the first duration value, sf2 is the second duration value, and sf3 is the third duration value; and 0≤sf1<sf2<sf3≤100;

[0098] S160: Mark the sum of the total scores of each area traversed by the robot under each path and the total time the robot has traversed under each path as the first total score, and mark the path corresponding to the lowest first total score as the first path of the robot; sort the paths except the first path from low to high according to the total scores; wherein p represents the total number of areas traversed by the robot under each path, and p=x+d+g.

[0099] For example, an intelligent robot Z in a hotel J needs to go from point MI to point ZD. The robot's operating area is obtained from the database, and the operating area is divided into 20 blocks based on the obstacle situation and numbered; there are active obstacles in 15 of the 20 blocks, no obstacles in three blocks, and fixed obstacles in two blocks;

[0100] Obtain 8 sets of historical obstacle area existence times through the database;

[0101] Among them, at the first special node, the total duration of obstacles in the three groups of historical obstacle areas corresponding to the obstacle zone was 8 hours, 7.5 hours, and 6.9 hours respectively; at the second special node, the total duration of obstacles in the two groups of historical obstacle areas corresponding to the obstacle zone was 3.6 hours, 2.8 hours, and 3.2 hours respectively; at the third special node, the total duration of obstacles in the two groups of historical obstacle areas corresponding to the obstacle zone was 1.5 hours and 2.3 hours respectively; a frequency-time threshold weight table is constructed according to the special nodes, as shown in Table 1;

[0102]

[0103] Now the robot is at the second special node, through the formula The calculated frequency-time threshold a1=(3.6+2.8+3.2) / 3×0.35=1.12h;

[0104] Among the 15 blocks with active obstacles, if the activity time of 6 blocks exceeds 1.12h, these 6 corresponding areas will be marked as high-frequency obstacle areas; if the activity time of 9 blocks does not exceed 1.12h, these 9 corresponding areas will be marked as low-frequency obstacle areas;

[0105] Get the robot's starting point and destination from the database, and get the number of all traversable paths, which are u1, u2, and u3;

[0106] Among them, u1 passes through two high-frequency obstacle areas marked as A32 from point MI to point ZD, passes through three low-frequency obstacle areas A23, passes through one obstacle-free area and one fixed obstacle area marked as A12;

[0107] u2 passes through three high-frequency obstacle areas marked as A33 from point MI to point ZD, passes through three low-frequency obstacle areas marked as A23, passes through an obstacle-free area and does not pass through a fixed obstacle area marked as A11;

[0108] u3 passes through a high-frequency obstacle area marked as A31 from point MI to point ZD, three low-frequency obstacle areas A23, two obstacle-free areas, and one fixed obstacle area marked as A13;

[0109] The A1x, A2d, and A3g areas are divided into 10, 30, and 60 respectively;

[0110] The total score of each area that the robot goes through under the u1 path is calculated by the formula ZQFu=x×qf1+d×qf2+g×qf3, which is ZQFu1=2×60+3×30+2×10=230;

[0111] Through calculation, we can get the total score of each area that the robot goes through under the u2 path ZQFu2=3×60+3×30+1×10=280;

[0112] Through calculation, we can get the total score of each area that the robot goes through under the u3 path ZQFu3=1×60+3×30+3×10=180;

[0113] From the database, it is found that the time required for the robot to pass through the high-frequency obstacle area is 10 minutes, the time required to pass through the low-frequency obstacle area is 6 minutes, and the time required to pass through the obstacle-free area and the fixed obstacle area is 3 minutes;

[0114] The total time the robot takes on path u1 is calculated by the formula ZSCu1=x×ST1+d×ST2+g×ST3, which is ZSCu1=2×10+3×6+2×3=44min.

[0115] By calculating ZSCu2=3×10+3×6+1×3=51min, we can get the total time the robot takes on the u2 path;

[0116] By calculating ZSCu3=1×10+3×6+3×3=37min, we can get the total time the robot takes on the u3 path;

[0117] The total time taken by the robot for the longest of the three paths is 51 minutes, and the total time taken by the robot for the shortest path is 37 minutes. The total time taken for the path between [44 minutes, 51 minutes] is assigned a value of 50.

[0118] Assign the total duration of the period [29.33min, 44min) to 30;

[0119] Assign the total duration in [0, 29.33min) to 20;

[0120] The sum of the total score of each area that the robot has gone through on each path and the total time that the robot has gone through on each path is taken as the first total score; then the first total score corresponding to the u1 path is 230+50=280 points; the first total score corresponding to the u2 path is 280+50=330 points; the first total score corresponding to the u3 path is 180+30=210 points;

[0121] Use the u3 path as the first path.

[0122] See also Figure 3 , the specific process of obstacle avoidance:

[0123] S210: Retrieving historical avoidance data from a database;

[0124] S220: Counting the number of times that the dynamic obstacles ahead appear and the number of times that the obstacles are actively avoided, and marking the ratio of the number of times that the dynamic obstacles appear to the number of times that the obstacles are actively avoided as the active avoidance rate;

[0125] S310: extract the active avoidance rate, and compare the active avoidance rate with the avoidance threshold; when the active avoidance rate is greater than the avoidance threshold, the robot waits in place; otherwise, jump to S320;

[0126] S320: Retrieve the robot whose active avoidance rate exceeds the avoidance threshold and is waiting at the same place, and set the waiting time threshold;

[0127] S330: Go to step S120, obtain the second total score, and use the path corresponding to the lowest value of the second total score as the third preliminary path of the robot;

[0128] S340: Determine whether the waiting time of the robot is greater than the waiting time threshold;

[0129] If yes, it is determined whether the forward distance between the current dynamic obstacle and the robot in motion exceeds the normal distance; if yes, the robot drives along the first path; if no, the robot does not actively avoid the current obstacle; the determination result is included in the database, and the active avoidance rate of the corresponding obstacle is recalculated, and the process proceeds to step S350;

[0130] If not, continue to wait;

[0131] S350: assigning the third preliminary path to the third path;

[0132] The waiting time threshold is obtained by obtaining the time required for performing steps S120 to S160 for several times from a database and taking an average value, and using the average value as the waiting time threshold.

[0133] For example, the robot is currently executing a task from the first path, and encounters two obstacles in front of it, marked as ZA1 and ZA2 respectively; ZA1 exists in the historical avoidance data, and ZA1 appears 7 times in total, of which 6 times the robot actively avoids the obstacle, so the active avoidance rate of ZA1 is 85.7%; ZA2 does not appear in the historical avoidance data, so the active avoidance rate of ZA2 is 0;

[0134] The avoidance threshold is set to 80%. Since 85.7%>80%, the robot waits in place.

[0135] The waiting time threshold is obtained as 5S, the second total score is obtained again, and the path corresponding to the lowest value of the second total score is used as the third preliminary path of the robot;

[0136] The current waiting time of the robot has exceeded the waiting time threshold, and the forward distance between the current dynamic obstacle and the moving robot does not exceed the normal distance, indicating that the current obstacle does not actively avoid the robot this time. The third preliminary path is assigned to the third path, and the robot drives along the third path.

[0137] See also Figure 4 ,The second aspect of the present invention provides an obstacle avoidance system for a hotel intelligent robot, comprising: a database and a path selection module, an obstacle avoidance module and an information linkage module connected thereto;

[0138] The database is used to store the types of obstacles that have been actively avoided in the past, the existence time of each obstacle, and the historical avoidance data statistically calculated based on the types of obstacles, wherein the types of obstacles include fixed obstacles and dynamic obstacles;

[0139] Path selection module: Obtain the robot's operating area and obstacle information in the operating area, and partition the operating area according to the obstacle information and frequency threshold to obtain the partition result; perform path planning based on the partition result to obtain the robot's first path;

[0140] Obstacle avoidance module: determines whether the obstacle currently appearing on the first path is a dynamic obstacle;

[0141] If yes, then the type of the currently appearing dynamic obstacle is obtained, the active avoidance rate of the corresponding dynamic obstacle is obtained based on the type of the dynamic obstacle, a third path is obtained according to the active avoidance rate, and the robot is controlled to travel along the third path;

[0142] If no, determine whether the current fixed obstacle is surmountable; if yes, issue a continue driving instruction; if no, replan to obtain a second path and drive along the second path;

[0143] Information linkage module: used to install data acquisition devices, collect and determine the type of obstacles through image recognition technology; and update the type of obstacles and historical avoidance data in the database; wherein the data acquisition device includes: image acquisition equipment and laser radar.

[0144] The third aspect of the present invention provides an obstacle avoidance storage medium for a hotel intelligent robot, on which a computer-readable storage medium is stored. When the computer-readable storage medium is executed by a processor, an obstacle avoidance method for a hotel intelligent robot as described in the first aspect above is implemented.

[0145] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0146] The working principle of the present invention is as follows: the present invention obtains obstacle information of the robot's operating area and obstacles in the operating area; and partitions the operating area according to the obstacle information and the frequency-time threshold to obtain a partition result; performs path planning according to the partition result to obtain a first path of the robot; determines whether there are dynamic obstacles among the obstacles appearing on the first path; if yes, determines the corresponding dynamic obstacle through image recognition technology, calculates the active avoidance rate of the corresponding dynamic obstacle based on historical avoidance data, obtains a third path according to the active avoidance rate, and drives along the third path; if no, sets the type of surmountable obstacles, and determines whether the current obstacle is a surmountable obstacle; if yes, issues a continue driving instruction; if no, replans to obtain a second path, and drives along the second path.

[0147] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An obstacle avoidance method for a hotel intelligent robot, characterized in that: include: Obtain the robot's operating area and obstacle information in the operating area, and partition the operating area according to the obstacle information and the frequency-time threshold to obtain the partition result; Perform path planning based on the partitioning results to obtain the first path of the robot; Determine whether the obstacle currently appearing on the first path is a dynamic obstacle; If yes, then the type of the currently appearing dynamic obstacle is obtained, the active avoidance rate of the corresponding dynamic obstacle is obtained based on the type of the dynamic obstacle, a third path is obtained according to the active avoidance rate, and the robot is controlled to travel along the third path; If no, determine whether the current fixed obstacle is surmountable; if yes, issue a continue driving instruction; If no, re-plan to obtain a second path and drive along the second path; The method for obtaining the frequency-time threshold includes: Obtain several groups of historical obstacle area existence times through the database; A frequency-time threshold weight table is constructed according to the first special node, the second special node and the third special node; the first special node is a time node when the operation area is in an active period; the second special node is a node when the operation area is in a meal time of a day other than the first special node; the third special node is a time node when there are obstacles in the operation area other than the first special node and the second special node; The frequency-time threshold is calculated based on the robot's operating area and the location of the special node.

2. The obstacle avoidance method for a hotel intelligent robot according to claim 1, characterized in that: The partitioning of the operation area according to the obstacle information and the frequency-time threshold includes: Divide the operating area into several sub-blocks based on the type of obstacles and number them; Determine whether there is an obstacle in the sub-block; if yes, mark the corresponding sub-block as an obstacle area; if no, mark the corresponding sub-block as an obstacle-free area; and count the activity time of obstacles in the obstacle area through the database; Determine whether the activity time of the obstacle in the obstacle area is 0; If yes, the corresponding sub-block is marked as a fixed obstacle area; If not, determine whether the activity time is lower than the frequency-time threshold a1; if yes, mark the corresponding block as a low-frequency obstacle area; if not, mark the corresponding block as a high-frequency obstacle area; wherein, 0<a1<24.

3. The obstacle avoidance method for a hotel intelligent robot according to claim 1, characterized in that: The calculating of the frequency-time threshold according to the operation area where the robot is located and the position of the special node includes: Determine whether the operation area of ​​the robot is in the first special node; Yes, then by the formula The frequency-time threshold a1 is calculated; If not, then determine whether the robot's operating area is in the second special node; if yes, then use the formula Calculate the frequency-time threshold a1; if not, use the formula The frequency-time threshold a1 is calculated; Wherein, Tr1 is the total duration of obstacles existing in the first special node, Tr2 is the total duration of obstacles existing in the second special node, Tr3 is the total duration of obstacles existing except Tr1 and Tr2, Qm is the frequency-time threshold weight of the mth special node obtained from the frequency-time threshold weight table, m=1, 2, 3; n1, n2, n3 represent the total number of statistical days under each special node; r1, r2, r3 represent the number of statistical days under each special node; n1, n2, n3, r1, r2, r3 are all positive integers greater than 1.

4. The obstacle avoidance method for a hotel intelligent robot according to claim 1, characterized in that: The step of performing path planning according to the partitioning result to obtain the first path of the robot includes: S100: retrieve obstacle-free areas, fixed obstacle areas, low-frequency obstacle areas, and high-frequency obstacle areas; mark the obstacle-free areas and fixed obstacle areas as A1x, the low-frequency obstacle areas as A2d, and the high-frequency obstacle areas as A3g; where x, d, and g represent the number of corresponding areas, respectively; x≥0, d≥0, g≥0, and x+d+g≠0; S120: Obtain the starting point and destination of the robot, obtain the number of all traversable paths through the path planning algorithm, and record it as u; count the type of area and the corresponding number of areas passed by each traversable path; S130: Assign scores to the regions corresponding to A1x, A2d, and A3g as qf1, qf2, and qf3, respectively; wherein qf1 is the score of the first region, qf2 is the score of the second region, and qf3 is the score of the third region; and 0≤qf1<qf2<qf3≤100; The total score of each area that the robot has gone through under each path is calculated by the calculation formula ZQFu=x×qf1+d×qf2+g×qf3; S140: Obtain the time required for the robot to pass through the corresponding area and record it as ST1, ST2, and ST3 respectively; The total time the robot takes on each path is calculated using the formula ZSCu=x×ST1+d×ST2+g×ST3; S150: assign the total duration of the interval [(ZSCmax+ZSCmin) / 2, ZSCmax] to sf1; Assign the total duration in the interval [(ZSCmax+ZSCmin) / 3, (ZSCmax+ZSCmin) / 2) to sf2; Assign the total duration in the interval [0, (ZSCmax+ZSCmin) / 3) to sf3; where ZSCmax represents the total duration of the robot corresponding to the longest path under all paths; ZSCmin represents the total duration of the robot corresponding to the shortest path under all paths; where sf1 is the first duration value, sf2 is the second duration value, and sf3 is the third duration value; and 0≤sf1<sf2<sf3≤100; S160: Mark the sum of the total scores of each area traversed by the robot under each path and the total time the robot has traversed under each path as the first total score, and mark the path corresponding to the lowest first total score as the first path of the robot; sort the paths except the first path from low to high according to the total scores; wherein p represents the total number of areas traversed by the robot under each path, and p=x+d+g.

5. The obstacle avoidance method for a hotel intelligent robot according to claim 1, characterized in that: The obtaining the active avoidance rate of the corresponding dynamic obstacle based on the type of the dynamic obstacle includes: S210: Retrieving historical avoidance data from a database; S220: Count the number of times the dynamic obstacle appears ahead and the number of times the obstacle is actively avoided, and mark the ratio of the number of times the dynamic obstacle appears to the number of times the obstacle is actively avoided as the active avoidance rate.

6. The obstacle avoidance method for a hotel intelligent robot according to claim 4, characterized in that: The acquiring the third path according to the active avoidance rate includes: S310: extract the active avoidance rate, and compare the active avoidance rate with the avoidance threshold; when the active avoidance rate is greater than the avoidance threshold, the robot waits in place; otherwise, jump to S320; S320: Retrieve the robot whose active avoidance rate exceeds the avoidance threshold and is waiting at the same place, and set the waiting time threshold; S330: Go to steps S120-S160, obtain the second total score, and use the path corresponding to the lowest value of the second total score as the third preliminary path of the robot; S340: Determine whether the waiting time of the robot is greater than the waiting time threshold; If yes, it is determined whether the forward distance between the current dynamic obstacle and the forward robot exceeds the normal distance; if yes, the robot drives along the first path; if no, the current obstacle does not actively avoid the robot; the judgment result is included in the database, and the active avoidance rate of the corresponding obstacle is recalculated, and the process goes to step S350; If not, continue to wait; S350: assigning the third preliminary path to the third path.

7. The obstacle avoidance method for a hotel intelligent robot according to claim 4, characterized in that: The re-planning to obtain the second path includes: Get the robot's current location and destination, re-enter step S120, obtain the sum of the total scores of each area the robot has gone through under each path and the total time the robot has gone through under each path, and take the path corresponding to the minimum value of the sum of each assignment as the robot's second path.

8. An obstacle avoidance system for a hotel intelligent robot, adapted to the obstacle avoidance method for a hotel intelligent robot according to any one of claims 1 to 7, characterized in that: include: A database, a path selection module and an obstacle avoidance module connected to said database; The database is used to store the types of obstacles that were actively avoided in the past, the existence time of each obstacle, and the historical avoidance data calculated based on the types of obstacles; The path selection module: obtains the operation area of ​​the robot and obstacle information of obstacles in the operation area, and partitions the operation area according to the obstacle information and the frequency-time threshold to obtain a partition result; Perform path planning based on the partitioning results to obtain the first path of the robot; The obstacle avoidance module: determines whether the obstacle currently appearing on the first path is a dynamic obstacle; If yes, then the type of the currently appearing dynamic obstacle is obtained, the active avoidance rate of the corresponding dynamic obstacle is obtained based on the type of the dynamic obstacle, a third path is obtained according to the active avoidance rate, and the robot is controlled to travel along the third path; If no, determine whether the current fixed obstacle is surmountable; if yes, issue a continue driving instruction; No, re-plan to obtain a second path, and drive along the second path.

9. The obstacle avoidance system of a hotel intelligent robot according to claim 8, characterized in that: Also includes: Information linkage module; The information linkage module is used to install a data acquisition device to collect and determine the type of obstacles through image recognition technology; And, updating the obstacle type and historical avoidance data in the database; wherein the data acquisition device includes: image acquisition equipment and laser radar.

10. An obstacle avoidance storage medium for a hotel intelligent robot, on which a computer-readable storage medium is stored, characterized in that: When the computer-readable storage medium is executed by a processor, the obstacle avoidance method for a hotel intelligent robot as described in any one of claims 1 to 7 is implemented.

Images