CN-121706841-B - Modularized robot allosteric planning method and system

CN121706841BCN 121706841 BCN121706841 BCN 121706841BCN-121706841-B

Abstract

The invention provides a modularized robot allosteric planning method and a modularized robot allosteric planning system, which relate to the technical field of robot control and comprise the steps of respectively solving a global optimal solution set of each optimization target aiming at three optimization targets with the least number of mobile modules, the least number of disconnected connecting mechanisms and the shortest moving distance; the method comprises the steps of respectively carrying out hierarchical sequence habitat division on a global optimal solution set of each optimization target, generating a plurality of niches for each optimization target, extracting common characteristics of an allosteric solution of each niche, encoding the common characteristics into common chromosome segments, carrying out non-dominant ordering iterative optimization in the habitat through a genetic algorithm based on the common chromosome segments until the maximum iterative times are reached, merging final populations of all the habitats, and extracting non-dominant solutions to be output as optimal solutions. According to the invention, through a genetic algorithm based on hierarchical sequence habitat division, the problem of solving three coupled optimization targets in the process of transformation is solved, and the pareto front of the optimal transformation problem is searched.

Inventors

ZHOU LELAI
ZHANG CHEN
SONG RUI
LI YIBIN

Assignees

山东大学

Dates

Publication Date: 20260512
Application Date: 20260214

Claims (8)

1. A modular robot allosteric programming method comprising: aiming at three optimization targets with the least number of mobile modules, the least number of disconnected connecting mechanisms and the shortest moving distance, respectively solving a global optimal solution set of each optimization target; respectively carrying out hierarchical sequence habitat division on the global optimal solution set of each optimization target, and generating a plurality of niches for each optimization target, wherein each niche comprises a plurality of allosteric solutions; the global optimal solution sets of all the optimization targets are respectively subjected to hierarchical sequence habitat division, wherein the hierarchical sequence habitat division comprises the steps of habitat class division and niche division; The habitat comprises four types, namely a globally optimal solution set corresponding to the least number of mobile modules The globally optimal solution set with the least number of disconnected connection mechanisms corresponds to the habitat class The habitat class corresponds to the global optimal solution set with the shortest moving distance Habitat class, non-optimal solution set correspondence A habitat class; 、 And The habitat classes each comprise a plurality of niches, The habitat class contains only one niche; extracting common characteristics of allosteric solutions of the niches aiming at each niche, and encoding the common characteristics into common chromosome fragments; the method extracts the common characteristics of the allosteric solutions and encodes the common characteristics into common chromosome fragments, and specifically comprises the following steps: the niches under the habitat class are characterized in that the shape of a static module and the serial number of the static module are taken as common characteristics, the point positions of the static module in a target configuration are determined, and the serial numbers of the static module are filled into the row vectors corresponding to the target point positions to form common chromosome segments; The method comprises the steps of extracting the largest common segment of each solution in the habitat by taking the distance between the allosteric solutions as a common characteristic of the niches in the habitat class, and forming a common chromosome segment; The niches in the habitat class are characterized in that the same alignment points are used as common features, and module numbers outside the alignment points and the maximum rectangular area are used as common chromosome fragments; The maximum rectangular area is a rectangular area formed by the point position of the mobile module and the corresponding idle position point position; based on the common chromosome segments, non-dominant ordering iterative optimization is carried out in the habitat through a genetic algorithm until the maximum iteration times are reached, the final populations of all habitats are combined, and the non-dominant solution is extracted and output as an optimal solution.
2. The modular robot allosteric planning method according to claim 1, wherein the non-dominant ordering iterative optimization is performed in the habitat by genetic algorithm, based on the initial population, chromosome crossover and mutation calculation are performed, index values of all the allosteric solutions on three optimization targets are calculated independently as ternary cost of the allosteric solutions, non-dominant ordering is performed on all the allosteric solutions, and a new population is selected for the next iteration until the maximum iteration number is reached.
3. A modular robotic allosteric programming method in accordance with claim 2 wherein each individual in the population represents a feasible solution in the genetic algorithm, the individual being represented by a chromosome code, the chromosome being one The dimension vector is composed of two parts: the first part comprises 1 st to 1 st A number of elements representing a target point location of the module; another part is an alignment point, formed by The element representation determines the relative position of the initial configuration and the target configuration in the same coordinate system.
4. A modular robot allosteric programming method according to claim 2 wherein the index values on the three optimization targets are formulated as: Wherein, the 、、 The disconnection number, the number of the mobile modules and the allosteric mobile distance are respectively, Is the total number of modules in the overall configuration, Is the number of modules contained in the static component, The number of connections to be made for the initial configuration, Is a chromosome Middle (f) The number of connections of the individual components, n being the chromosome The length () function represents the length of the path in brackets, Is the first A movement path of each module.
5. A modular robotic allosteric programming system comprising: the respective solving module is configured to respectively solve the global optimal solution sets of the optimization targets aiming at three optimization targets with the least number of the moving modules, the least number of the disconnected connecting mechanisms and the shortest moving distance; The ecological environment dividing module is configured to divide the overall optimal solution set of each optimization target into layered sequence ecological environments, generate a plurality of niches for each optimization target, and each niche comprises a plurality of allosteric solutions; the global optimal solution sets of all the optimization targets are respectively subjected to hierarchical sequence habitat division, wherein the hierarchical sequence habitat division comprises the steps of habitat class division and niche division; The habitat comprises four types, namely a globally optimal solution set corresponding to the least number of mobile modules The globally optimal solution set with the least number of disconnected connection mechanisms corresponds to the habitat class The habitat class corresponds to the global optimal solution set with the shortest moving distance Habitat class, non-optimal solution set correspondence A habitat class; 、 And The habitat classes each comprise a plurality of niches, The habitat class contains only one niche; A feature encoding module configured to extract common features of its allosteric solutions for each niche and encode as common chromosome segments; the method extracts the common characteristics of the allosteric solutions and encodes the common characteristics into common chromosome fragments, and specifically comprises the following steps: the niches under the habitat class are characterized in that the shape of a static module and the serial number of the static module are taken as common characteristics, the point positions of the static module in a target configuration are determined, and the serial numbers of the static module are filled into the row vectors corresponding to the target point positions to form common chromosome segments; The method comprises the steps of extracting the largest common segment of each solution in the habitat by taking the distance between the allosteric solutions as a common characteristic of the niches in the habitat class, and forming a common chromosome segment; The niches in the habitat class are characterized in that the same alignment points are used as common features, and module numbers outside the alignment points and the maximum rectangular area are used as common chromosome fragments; The maximum rectangular area is a rectangular area formed by the point position of the mobile module and the corresponding idle position point position; And the genetic optimization module is configured to perform non-dominant ordering iterative optimization in the habitat through a genetic algorithm based on the common chromosome segments until the maximum iteration number is reached, combine the final populations of all the habitats, and extract non-dominant solutions as optimal solutions to output.
6. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements a modular robot deformation planning method according to any of claims 1-4.
7. A non-transitory computer readable storage medium storing computer instructions which, when executed by a processor, implement a modular robot allosteric planning method according to any one of claims 1-4.
8. An electronic device comprising a processor, a memory and a computer program, wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to cause the electronic device to perform a modular robot allosteric planning method according to any one of claims 1-4.

Description

Modularized robot allosteric planning method and system Technical Field The invention relates to the technical field of robot control, in particular to a modularized robot allosteric planning method and system. Background The mobile modularized robot is a combination body formed by connecting a plurality of mobile modules through a docking mechanism, and each module has independent sensing, controlling and moving capabilities. The combination body changes its own configuration through an allosteric process so as to adapt to various environments and tasks. The allosteric is one of the core capabilities of a modular robot, which enables the robot to switch to a configuration adapted to traverse different terrains or transport various objects. The high-efficiency low-cost allosteric ensures the flexibility and stability of the robot when coping with dynamic environments and tasks, and improves the endurance capacity. The modularized robot allosteric planning technology still has the challenges that the moving distance and the allosteric steps are often used as allosteric cost indexes in the existing allosteric planning research, the allosteric cost covers more dimensions, the number of moving modules directly influences the moving cost, the active docking locking device is smaller in number, the energy consumption is lower, the action error rate is lower, the existing research is smaller, the two factors are considered, in addition, the coupling relation exists in multi-index optimization, the improvement of one index often causes the deterioration of the other index, the optimal solution exists in the form of the pareto front, the coupling of the multi-optimization targets increases the complexity of the allosteric planning, the number of the allosteric schemes increases exponentially with the number of modules, the solution space scale expands in steps with the number of modules, and the searching of the pareto optimal solution in a large-scale solution space faces a great challenge. In summary, the existing modularized robot allosteric programming technology has the problems of strong multi-objective coupling, space explosion solution and incomplete cost indexes. Disclosure of Invention In order to solve the problems, the invention provides a modularized robot allosteric planning method and a modularized robot allosteric planning system, which solve the problems of solving three coupled optimization targets in the allosteric process and searching the pareto front edge of the optimal allosteric problem through a genetic algorithm based on hierarchical sequence habitat division. According to some embodiments, the present invention employs the following technical solutions: a modular robot allosteric programming method comprising: aiming at three optimization targets with the least number of mobile modules, the least number of disconnected connecting mechanisms and the shortest moving distance, respectively solving a global optimal solution set of each optimization target; respectively carrying out hierarchical sequence habitat division on the global optimal solution set of each optimization target, and generating a plurality of niches for each optimization target, wherein each niche comprises a plurality of allosteric solutions; extracting common characteristics of allosteric solutions of the niches aiming at each niche, and encoding the common characteristics into common chromosome fragments; based on the common chromosome segments, non-dominant ordering iterative optimization is carried out in the habitat through a genetic algorithm until the maximum iteration times are reached, the final populations of all habitats are combined, and the non-dominant solution is extracted and output as an optimal solution. According to some embodiments, the present invention employs the following technical solutions: a modular robotic allosteric planning system comprising: the respective solving module is configured to respectively solve the global optimal solution sets of the optimization targets aiming at three optimization targets with the least number of the moving modules, the least number of the disconnected connecting mechanisms and the shortest moving distance; The ecological environment dividing module is configured to divide the overall optimal solution set of each optimization target into layered sequence ecological environments, generate a plurality of niches for each optimization target, and each niche comprises a plurality of allosteric solutions; A feature encoding module configured to extract common features of its allosteric solutions for each niche and encode as common chromosome segments; And the genetic optimization module is configured to perform non-dominant ordering iterative optimization in the habitat through a genetic algorithm based on the common chromosome segments until the maximum iteration number is reached, combine the final populations of all the habitats, and extract non-dominant solutions