An intelligent mobile robot path planning method, device and medium

Abstract

The application discloses a kind of intelligent mobile robot path planning method, simulates CAR-T targeted therapy, comprising the following steps: grid map, input robot starting point and end point position;Suppose that intelligent group for path planning is T cell, T cell uses octal neighborhood search at starting point, set the maximum iteration number of algorithm, T cell selects starting point and expands search eight sub-neighborhood, T cell in sub-neighborhood is elite to CAR-T cell, and the elite rate is directly proportional to the distance to end point;CAR-T cell is transferred to other sub-neighborhood with certain probability;Judge whether end point is found;If find, then current iteration ends, otherwise, the sub-neighborhood of selected cell concentration highest, continue to repeat the above octal neighborhood search process until end point is found.The initial T cell concentration of historical optimal path is enhanced in the next iteration.The application has the advantages that the search efficiency and accuracy of the algorithm are improved, a globally optimal task path can be planned for the robot in a short time, and the work efficiency is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent mobile robot path planning, and particularly relates to an intelligent mobile robot path planning method, device and medium simulating CAR-T targeted therapy. BACKGROUND

[0002] CAR-T therapy is a new precise targeted therapy that can effectively treat cancer and has broad medical application prospects. CAR-T therapy is divided into five steps: separating T cells from the patient's body, modifying T cells into CAR-T cells in vitro, expanding CAR-T cells in vitro, reinfusing the patient's body, and evaluating the efficacy and monitoring the response. The modified CAR-T cells have precise recognition ability and can quickly and accurately kill tumor cells. CAR-T therapy can be understood as the process of searching for tumor cell populations by T cell elite enhancement and expansion, and this search has significant targeting, which meets the characteristics of swarm intelligence optimization algorithm.

[0003] Existing swarm intelligence optimization algorithms for intelligent mobile robot path planning include ant colony algorithm, particle swarm algorithm, firework algorithm, wolf swarm algorithm, etc. These algorithms simulate the group behavior mechanism of insects, bird flocks, animal herds, etc., and abstract them to retain their swarm intelligence, which is used to solve some complex search optimization problems.

[0004] Objective shortcomings of existing swarm intelligence optimization algorithms

[0005] 1. Search efficiency and search accuracy are two major problems faced by existing swarm intelligence optimization algorithms. For example, in the prior art analysis, the ant colony algorithm has slow convergence speed and low search efficiency, and although the particle swarm algorithm improves the search efficiency, it is easy to fall into local optimal solution and has low search accuracy.

[0006] 2. Most swarm intelligence algorithms are based on probability search, and the search results have randomness, such as the roulette mechanism used in the ant colony algorithm. This leads to the possibility of losing the optimal solution found in the current search in the next search, making it difficult to guarantee the stability of the global optimal solution after multiple searches.

[0007] Prior art case analysis related to the present application

[0008] Ant colony algorithm:

[0009] The ant colony algorithm simulates the behavior of ants foraging in nature. In nature, ants can always find the shortest path from the nest to the food source through mutual cooperation in the foraging process [1] .

[0010] Disadvantages:

[0011] ①Local optimum: Ants leave pheromones on different paths they have walked, which is inversely proportional to the distance of the path. The greater the concentration of information, the greater the probability that the path will be selected again, which is a positive feedback mechanism. In the initial environment, the classic ant colony algorithm assigns the same amount of pheromone to each path, and the ants randomly select the path to be passed. This random selection inevitably leads to a suboptimal solution, and the subsequent positive feedback mechanism repeatedly strengthens this suboptimal solution, eventually falling into a local optimum and being difficult to jump out again.

[0012] ②Slow convergence speed. In the initial environment, the concentration of pheromones on each path is the same, and the ants randomly select different paths, which requires a long time to exert the effect of the positive feedback mechanism, resulting in slow convergence speed of the algorithm.

[0013] ③The roulette wheel path selection mechanism may cause the optimal solution found in the current search to be lost in the next search, making it difficult to ensure the stability of the global optimal solution after multiple searches.

[0014] Particle swarm algorithm

[0015] The particle swarm algorithm is another typical representative of swarm intelligence optimization algorithms, which simulates the swarm intelligence behavior of bird foraging. In this algorithm, each individual in the bird swarm is described as a "particle" with no mass and volume. At the beginning of iteration, each particle has its own speed and position. In each iteration, each particle finds its own optimal solution and the global best position solution of all particles through the position information sharing mechanism between particles, and the above solutions guide the particles to search for a better position. After multiple iterations, all particles converge to the global best position solution [2] .

[0016] Disadvantages:

[0017] This algorithm improves the convergence speed, but is prone to fall into a local optimal solution and is difficult to obtain a global optimal solution.

[0018] References

[0019] [1] Zhang Songcan. Mobile Robot Path Planning Based on Ant Colony Algorithm[D]. Henan University of Science and Technology, 2021. DOI:10.27115 / d.cnki.glygc.2021.000001.

[0020] [2] Wang Chunbo. Optimization Operation of Combined Cooling Heating and Power Microgrid Based on Improved Particle Swarm Algorithm[D]. China University of Mining and Technology, 2022. DOI:10.27623 / d.cnki.gzkyu.2022.001361. Summary

[0021] The present application provides a kind of intelligent mobile robot path planning method, equipment and medium for the defects of prior art.

[0022] To achieve the above object of the application, the technical scheme adopted by the present application is as follows:

[0023] A path planning method for an intelligent mobile robot, comprising the following steps:

[0024] S1: rasterizing a map, inputting a starting point and an ending point of the intelligent mobile robot;

[0025] S2: assuming that the intelligent group for path planning is T cells, the T cells adopt eight-neighborhood search at the starting point, setting the maximum number of iterations of the algorithm, initializing the concentration of T cells, i.e., the number of cells, establishing tables T1 and T2 to record the child neighborhoods and parent neighborhoods in the eight-neighborhood search respectively, and establishing table D to record the path length of each iteration.

[0026] S3: the T cells select the starting point as the first parent neighborhood, expand the search to eight child neighborhoods, and record them in tables T1 and T2 respectively. The T cells in the child neighborhoods in table T1 are elitized into CAR-T cells, and the elitism rate is proportional to the distance to the ending point, i.e., the closer the child neighborhood is to the ending point, the higher the concentration of the elitized CAR-T cells in the child neighborhood is. The Euclidean distance can be used for measurement. The CAR-T cells are transferred to other child neighborhoods at a certain probability.

[0027] S4: the CAR-T cells continue to expand, and it is judged whether the ending point is found. If yes, the current iteration ends, otherwise, the child neighborhood with the highest cell concentration in T1 is selected as the parent neighborhood for the next step of search, and the above eight-neighborhood search process is repeatedly continued until the ending point is found.

[0028] Further, the specific steps of S3 are as follows:

[0029] A: the T cells adopt eight-neighborhood search from the starting point to search for the ending point.

[0030] B: the T cells are elitized into CAR-T cells, and the elitism rate is proportional to the distance of the child neighborhood to the ending point.

[0031] C: the CAR-T cells continue to expand, and the neighborhood with the highest cell concentration is updated to achieve targeted search for the ending point.

[0032] D: the targeted search in step C falls into a local optimum, and the elitized CAR-T cells need to be transferred to other child neighborhoods at a certain probability.

[0033] E: after the probability deviates, the distance in the current step and the concentration of the CAR-T cells are comprehensively measured, the cells are expanded, and then the neighborhood with the highest cell concentration is updated again to complete the final targeted search.

[0034] Further, the cell elitism rate in S3 is calculated as follows:

[0035]

[0036] where (x k ,y k ) is the position of the above-mentioned sub-neighborhood in the grid map, (x e ,y e ) is the position of the end point, H k is the Euclidean distance between the sub-neighborhood and the end point, β k is the elite rate, and b is the scaling factor.

[0037] As preferred, in S3, in order to ensure global optimal search, the CAR-T cells will be transferred to other sub-neighborhoods with a certain probability, that is, the sub-neighborhoods far from the end point also have the opportunity to obtain the CAR-T cells.

[0038] As preferred, at the end of each iteration, the table D records the corresponding search path. From the table D, the historical optimal path solution can be found, and in the next iteration, the initial T cell concentration of the historical optimal path is additionally enhanced to accelerate the convergence of the algorithm to the global optimal path.

[0039] The application further discloses a computer device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the intelligent mobile robot path planning method when executing the program.

[0040] The application further discloses a computer readable storage medium, which stores a computer program, and the program is executable on the processor to implement the intelligent mobile robot path planning method.

[0041] Compared with the prior art, the application has the following advantages:

[0042] 1. The application simulates the CAR-T targeted treatment process and has the characteristics of elite enhancement, targeted search, probability deviation, and self-guidance. Compared with the existing swarm intelligence optimization algorithm, the algorithm of the application improves the search efficiency of the algorithm through cell elitization and targeted search. The probability deviation mechanism of the search is introduced to ensure the acquisition of the global optimal solution. In addition, the self-guidance of the historical optimal search solution enables the subsequent iteration to further explore on the basis of the historical optimal solution, and the final iteration result maintains the historical optimal solution or obtains a better solution, further improving the search stability and precision of the algorithm.

[0043] 2. The application is applied to the path planning of an intelligent mobile robot, and compared with the existing swarm intelligence optimization algorithm, the method can plan a global optimal path in a shorter time. The robot can quickly obtain the shortest task route, effectively improving the work efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1is a flow chart of the intelligent path planning method of the embodiment of the present application;

[0045] Figure 2 is a schematic diagram of the eight-neighbor search method of the embodiment of the present application;

[0046] Figure 3 is a schematic diagram of the intelligent path planning of the embodiment of the present application. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical scheme and advantages of the present application more clear and explicit, the present application is further described in detail below according to the drawings and examples.

[0048] Inspired by CAR-T therapy, the present application proposes a new swarm intelligence optimization algorithm for solving path planning problems. The algorithm simulates the CAR-T targeted therapy process, assuming that the T cells in the human body are intelligent swarms, which will search for target cells in a limited search space. The above is a blind large-scale search with low efficiency. Some T cells are genetically modified into CAR-T cells. CAR-T cells have target information (distance information) of target cells, and can achieve rapid targeted search. In order to avoid falling into local optimal solution, the search of CAR-T cells will deviate to non-target direction with a certain probability, ensuring effective search for global optimal solution. And after completing multiple iterations, the proposed algorithm will use the historical optimal search solution to guide itself in subsequent iterations. In summary, the proposed CAR-T algorithm has the characteristics of elite enhancement, targeted search, probability deviation, and self-guidance. The targeted search of the elite CAR-T cells effectively improves the search efficiency of the algorithm, the probability deviation of the search ensures the acquisition of the global optimal solution, and the self-guidance using the historical optimal search solution enables the subsequent iteration to further explore on the basis of the historical optimal solution. The final iteration result maintains the historical optimal solution or obtains a better solution, effectively improving the search stability and precision of the algorithm. The algorithm flow is shown in Figure 1 .

[0049] Step 1: Algorithm initialization. Grid the map, input the search starting point and target cell position (end point), and use eight-neighbor search for T cells at the starting point, see Figure 2 . Set the maximum number of iterations of the algorithm (100 times), initialize the concentration of T cells, establish tables T1 and T2 to record the sub-neighborhood and parent neighborhood in the eight-neighbor search, and establish table D to record the search path length of each iteration.

[0050] Step2: Algorithm search. T cells select the starting point as the first parent neighborhood, and search eight child neighborhoods outwardly, which are recorded in tables T1 and T2. The T cell elites in the table T1 child neighborhood are elitized as CAR-T cells, and the elitization rate is proportional to the distance to the terminal point, that is, the closer the child neighborhood is to the terminal point, the higher the concentration of the elitized CAR-T cells in the child neighborhood. The Euclidean distance can be used for measurement. In order to ensure global optimal search, the CAR-T cells will transfer to other child neighborhoods with a certain probability, that is, the child neighborhood far from the terminal point also has the opportunity to obtain high concentration of CAR-T cells. The CAR-T cells continue to expand, and it is judged whether the target cell is found. If yes, the current iteration ends, otherwise, the child neighborhood with the highest cell concentration in T1 is selected as the parent neighborhood for the next search, and the above eight-neighborhood search process is repeated until the target cell is found.

[0051] Step3: Self-guiding. Once the iteration is completed, the corresponding search path is recorded in table D. From table D, the historical optimal path solution can be found, and the initial T cell concentration of the historical optimal path is additionally enhanced in the next iteration to accelerate the algorithm to converge to the global optimal path.

[0052] The specific description of Step2 is shown in Figure 3

[0053] A: T cells search the target cell in an eight-neighborhood search manner from the starting point. As can be seen from the figure, the T cells are blind and large-scale search in eight directions, which is low in efficiency. B: The T cells are elitized as CAR-T cells, and the elitization rate is proportional to the distance of the child neighborhood to the target cell. As can be seen from the figure, the rightmost neighborhood is closest to the target cell, and the concentration of the elitized CAR-T cells is the highest. C: The CAR-T cells continue to expand, and the cell concentration of the rightmost neighborhood is the highest. It can be seen that the higher the CAR-T cell concentration of a child neighborhood, the higher the cell concentration of the child neighborhood after cell expansion under the condition of the same search path distance (referring to the current step distance from the parent neighborhood to the child neighborhood), and the target search of the target cell is realized along the neighborhood with the highest concentration. D: The target search in C falls into a local optimum, and the elitized CAR-T cells need to transfer to other child neighborhoods with a certain probability. E: After the probability deviates, the CAR-T cell concentration of the lower right child neighborhood is the highest, and the cell concentration of the child neighborhood is the highest after cell expansion by comprehensively measuring the current step distance and the CAR-T cell concentration, and the final target search is completed along the neighborhood with high cell concentration.

[0054] ​In the intelligent mobile robot path planning application, the target cell position in the above algorithm process is the end position to be reached by the robot, the T cell is the intelligent group used for path planning of the algorithm (the basic feature of the group intelligence optimization algorithm, similar to the ant group used for path planning of the ant colony algorithm, the bird group used for path planning of the particle swarm algorithm), and the process (Step2) that the T cell searches for the target cell is the process that the mobile robot searches for the path to the end point, that is, the T cell searches for the target cell, which represents that the mobile robot reaches the end position. It should be noted that the T cell cannot be equal to the mobile robot, the T cell is the bionic object of the algorithm, the mobile robot is the application object of the algorithm, and the algorithm has the biological characteristics of the bionic object and is not affected by the difference of the application object. Compared with other group intelligence optimization algorithms, the algorithm can find a global shortest path for the mobile robot in a shorter time.

[0055] In another embodiment of the present application, a terminal device is provided, which comprises a processor and a memory, the memory is used to store a computer program, the computer program comprises program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc., which are the computing core and control core of the terminal, and are suitable for implementing one or more instructions, and are particularly suitable for loading and executing one or more instructions to implement a corresponding method flow or a corresponding function; the processor in the embodiment of the present application can be used for the operation of the intelligent mobile robot path planning method, comprising the following steps:

[0056] S1: rasterizing a map, inputting a search starting point and an end position;

[0057] S2: assuming that the intelligent group is a T cell, the T cell adopts eight-neighborhood search at the starting point, setting the maximum number of iterations of the algorithm, initializing the concentration of the T cell, that is, the number of cells, establishing tables T1 and T2 to record the child neighborhood and the parent neighborhood in the eight-neighborhood search, and establishing table D to record the search path length of each iteration.

[0058] S3: The T cell selection starting point is the first parent neighborhood, and eight child neighborhoods are searched outwardly, which are recorded in tables T1 and T2 respectively. The T cells in the table T1 child neighborhood are elitistized as CAR-T cells, and the elitist rate is proportional to the distance to the terminal point, that is, the closer the child neighborhood is to the terminal point, the higher the concentration of the elitistized CAR-T cells in the child neighborhood. The Euclidean distance can be used for measurement. The CAR-T cells are transferred to other child neighborhoods at a certain probability.

[0059] S4: The CAR-T cells continue to expand, and it is judged whether the terminal point is found. If yes, the current iteration ends, otherwise, the child neighborhood with the highest cell concentration in T1 is selected as the parent neighborhood for the next search, and the above eight-neighborhood search process is repeated until the terminal point is found.

[0060] In another embodiment of the present application, the present application also provides a storage medium, specifically a computer readable storage medium (Memory), which is a memory device in a terminal device, used for storing programs and data. It can be understood that the computer readable storage medium herein can include an internal storage medium in the terminal device, and of course can also include an expansion storage medium supported by the terminal device. The computer readable storage medium provides a storage space, which stores an operating system of the terminal. Moreover, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer readable storage medium herein can be a high-speed RAM memory, or a non-volatile memory such as at least one disk memory.

[0061] The one or more instructions stored in the computer readable storage medium can be loaded and executed by the processor to realize the corresponding steps of the intelligent path planning method in the above embodiments; the one or more instructions in the computer readable storage medium are loaded and executed by the processor as follows:

[0062] S1: rasterize the map, input the search starting point and terminal position;

[0063] S2: assuming that the intelligent group is T cells, the T cells adopt eight-neighborhood search at the starting point, the maximum iteration number of the algorithm is set, the T cell concentration, that is, the number of cells, is initialized, tables T1 and T2 are established to record the child neighborhoods and parent neighborhoods in the eight-neighborhood search respectively, and table D is established to record the search path length of each iteration.

[0064] S3: The T cell selection starts from the first parent neighborhood, and expands to search eight child neighborhoods, which are recorded in tables T1 and T2. The T cells in the child neighborhoods in table T1 are elitist to CAR-T cells, and the elitist rate is proportional to the distance to the end point, that is, the closer the child neighborhood to the end point, the higher the concentration of the elitist CAR-T cells in the child neighborhood. Euclidean distance can be used for measurement. CAR-T cells migrate to other child neighborhoods with a certain probability.

[0065] S4: The CAR-T cells continue to expand, and it is determined whether the end point is found. If yes, the current iteration ends, otherwise, the child neighborhood with the highest cell concentration in T1 is selected as the parent neighborhood for the next search, and the above eight-neighborhood search process is repeated until the end point is found.

[0066] Those skilled in the art will understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0067] The present application is described with reference to flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus produce an apparatus that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in a flow or multiple flows and / or blocks Figure 1 The functions specified in a flow or multiple flows and / or blocks

[0068] These computer program instructions can also be stored in a computer-readable memory that can direct the computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including instruction apparatus, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in a flow or multiple flows and / or blocks Figure 1 The functions specified in a flow or multiple flows and / or blocks

[0069] These computer program instructions can also be loaded into a computer or other programmable data processing devices, so that a series of operational steps are generated to realize the computer-implemented processes in the computer or other programmable devices, and the instructions executed in the computer or other programmable devices provide the processes for implementing the functions specified in the flowchart Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0070] Those skilled in the art will realize that the embodiments described herein are for illustrative purposes only and that various modifications and combinations of elements can be made without departing from the scope of the application, which is set forth in the following claims.

Claims

1. A method for path planning of an intelligent mobile robot, characterized in that, The method comprises the following steps: S1: rasterizing a map, inputting a start point and an end point of an intelligent mobile robot; S2: assuming that the intelligent group for path planning is T cell, T cell adopts octagonal neighborhood search at the starting point, sets the maximum iteration number of algorithm, initializes T cell concentration, i.e. the number of cells, establishes table and Record the child neighborhood and parent neighborhood in the octagonal neighborhood search respectively, establish table Record the search path length of each iteration; S3: The T cell selection starts from the first parent neighborhood, and expands outward to search eight sub-neighborhoods, which are recorded in Tables and ; Tables The T cell elites in the sub-neighborhoods are elitized into CAR-T cells, and the elitization rate is proportional to the distance to the end point, that is, the closer the sub-neighborhood is to the end point, the higher the concentration of the elitized CAR-T cells in the sub-neighborhood; the Euclidean distance can be used for measurement; the CAR-T cells can be transferred to other sub-neighborhoods at a certain probability; The elite rate is calculated as follows: ； wherein, is the position of the above sub-neighborhood in the grid map, is the end position, is the Euclidean distance of the sub-neighborhood to the end position, is the elite rate, is the scaling factor; S4: CAR-T cell continuous expansion, determine whether the end point is found; If found, the current iteration ends, otherwise, the selected The sub-neighborhood with the highest cell concentration is selected as the parent neighborhood for the next iteration, and the eight-neighborhood search process is repeated until the end point is found. 2.The intelligent mobile robot path planning method of claim 1, wherein: The specific steps of S3 are as follows: A: T cells search for the end point from the start point in an eight-neighborhood search mode; B: T cells are elitized into CAR-T cells, and the elite rate is proportional to the distance from the end point to the sub-neighborhood; C: CAR-T cells continuously expand, and target search is achieved along the neighborhood with the highest cell update concentration; D: The target search in step C falls into a local optimum, and the elitized CAR-T cells need to be transferred to other sub-neighborhoods at a certain probability; E: After the probability deviates, the current step distance and the CAR-T cell concentration are comprehensively measured, the cells are expanded, and the final target search is completed along the neighborhood with the highest cell update concentration. 3.The intelligent mobile robot path planning method of claim 1, wherein: In S3, in order to ensure global optimal search, CAR-T cells will be transferred to other sub-neighborhoods at a certain probability, that is, sub-neighborhoods far from the end point also have the opportunity to obtain CAR-T cells.

4. The intelligent mobile robot path planning method of claim 1, wherein: At the end of each iteration, the table The corresponding search path is recorded; from the table The historical optimal path solution can be found in the table, and in the next iteration, the initial T cell concentration of the historical optimal path is additionally enhanced to accelerate the algorithm to converge to the global optimal path.

5. A computer device, characterized by: The computer readable storage medium stores a computer program, and the program is executed by the processor to realize the intelligent mobile robot path planning method in any one of claims 1 to 4.

6. A computer-readable storage medium, characterized in that: The computer readable storage medium stores a computer program, and the program is executed by the processor to realize the intelligent mobile robot path planning method in any one of claims 1 to 4.

Images