Takemasa Arakawa

Trajectory generation for biped locomotion robot

The purpose of this research is to generate natural motion for a biped robot. It is difficult to generate stable and natural walking motion in various environments. In this paper, we propose the hierarchical evolutionary algorithm to generate a walking motion through energy optimization and aim to generate a natural motion by considering ZMP. The hierarchical evolutionary algorithm is a co-evolution algorithm of two optimization algorithms, one is the GA that minimizes the total energy of all actuators, and the other is the EP that optimizes interpolated configuration of biped locomotion robot. This hierarchical evolutionary algorithm can find the solution from a large numerical combination area, effectively. We obtain natural walking motions on some slopes using this algorithm, and confirm that the acquired trajectories can be applied successfully to the practical biped locomotion robot.

Stabilization control of biped locomotion robot based learning with GAs having self-adaptive mutation and recurrent neural networks

The purpose of this research is to generate natural motion of the biped locomotion robot such as the walking of a human in various environments. In this paper, we propose a method of stable motion generation of a biped locomotion robot. We apply the control of this proposed method with eight force sensors at the soles of the biped locomotion robot. The zero moment point (ZMP) is a well known index of stability in walking robots. ZMP is determined by the configuration of the robots. However, there are many configurations against the ZMP. Because of that, when we use ZMP as the stabilization index, we must select the best configuration in many stability configurations. Then it is a problem of which configurations are selected. In this paper, we solve the problem with recurrent neural networks. In both the single support and double support periods, we calculate the position of ZMP by using values from four force sensors at each sole, and actuation joints and the angles can be determined by recurrent neural networks without ZMP moving out from the supporting area of sole. We employ the recurrent neural networks with genetic algorithms for learning capability and self-adaptive mutation operator. Further, we build a biped locomotion robot in trial, which has 13 joints and verified that the calculated stable motion trajectory can be successfully applied to the practical biped locomotion.

Trajectory generation for redundant manipulator using virus evolutionary genetic algorithm

This paper deals with an application of a virus-evolutionary genetic algorithm (VEGA) to hierarchical trajectory planning of a redundant manipulator. The hierarchical trajectory planning is composed of a trajectory generator and position generator. The position generator generates collision-free intermediate positions of the redundant manipulator. The trajectory generator generates a collision-free trajectory based on some intermediate positions sent from the position generator. To generate a collision-free trajectory of the redundant manipulator, the VEGA is applied to the hierarchical trajectory planning only based on forward kinematics. The VEGA realizes horizontal propagation and vertical inheritance of genetic information in a population of candidate solutions. The main operator of the VEGA is a reverse transcription operator, which plays the roles of a crossover and selection simultaneously. In this paper, self-adaptive mutation is applied to the VEGA for local search of trajectory planning to obtain higher performance and the quick solution. Simulation results of the hierarchical trajectory planning show that the VEGA can generate a collision-free trajectory.

Natural motion trajectory generation of biped locomotion robot using genetic algorithm through energy optimization

The purpose of this research is to study the natural motion of a biped locomotion robot walking like a human in various environments. In this paper, we report about the natural motion trajectory generation of biped locomotion robot. We apply the genetic algorithm to off-line trajectory generation of biped locomotion robot through energy optimization and aim to generate more natural motion by considering the dynamic effect. Furthermore, we formulate the trajectory generation problem as a minimizing problem of energy so as to generate natural motion. To apply this calculated trajectory to a practical robot, we build a biped locomotion robot in trial, which has 13 joints and made of aluminum materials. Finally, we confirm that the calculated natural motion trajectory can be applied to practical biped locomotion.

Motion planning for a robotic system with structured intelligence

This paper deals with a robotic system with structured intelligence. Recently, behavior engineering has been discussed as a new technological discipline. The intelligence of a robot depends on the structure of hardware and software for processing information, that is, the structure determines the potentiality of intelligence. Based on perceptual information, a structured intelligent robot makes action from four levels in parallel. In addition, the robot generates its motion through the interaction with environment and at the same time gradually acquires its skill based on the generated motion. To acquire skill and motion, the robot requires internal and external evaluations at least. Actually, a virus-evolutionary genetic algorithm is applied to motion planning for a redundant manipulator and discuss its effectiveness through computer simulation results.

GA algorithms in intelligent robots

This chapter presents the role of genetic algorithms in intelligent robots. In general, the motion planning problems in intelligent robots can be fundamentally split into path planning problems, trajectory planning problems, and task planning problems. These planning faculties have many constraints concerning kinematics and dynamics of the robot and therefore it is very difficult to solve these planning problems. This chapter presents the general application of genetic algorithms to these planning tasks. Furthermore, the chapter discusses a trajectory planning problem for redundant manipulators and a motion planning problem for biped locomotion robots.

Evolutionary transition on Virus-Evolutionary Genetic Algorithm

The paper deals with a genetic algorithm (GA) based on the virus theory of evolution (VEGA) and evolutionary transition of a population. VEGA can self adaptively change the searching ratio between local search and global search according to the current state of population of candidate solutions. In addition, various types of evolutionary optimization methods have been proposed and successfully applied to many optimization problems. However, it is difficult to determine the coding method, genetic operators and selection scheme. To analyze the behavior of GAs, Markov chain analysis, deceptive problems and schema analysis have been discussed. We discuss evolutionary transition concerning fitness improvement through numerical simulation of the traveling salesman problem. The simulation results indicate that particular genetic operators give a population different potentialities for generating candidate solutions and that virus infection operators can generate effective schemata and propagate them to a population evolved with any genetic operators.

Fuzzy manufacturing scheduling by virus-evolutionary genetic algorithm in self-organizing manufacturing system

This paper deals with a fuzzy manufacturing scheduling problem in the self-organizing manufacturing system (SOMS), in which modules self-organize effectively according to other modules. A module decides its outputs through the interaction with other modules, but the module does not share all information of other modules. In addition, the information received from other modules often includes ambiguous and incomplete information. We therefore apply fuzzy theory to represent incomplete information of other modules. Furthermore, we apply a virus-evolutionary genetic algorithm (VEGA) to a fuzzy flow shop scheduling problem with fuzzy transportation time. The VEGA is a stochastic optimization method simulating coevolution of host population and virus population. The simulation results indicate that the fuzzified information is effective when a module has incomplete information in the SOMS.

Trajectory Planning and Learning of A Redundant Manipulator with Structured Intelligence

Abstract This paper deals with trajectory planning and motion learning for a redundant manipulator. Recently, behavior engineering for robotic systems has been discussed as a new technological discipline. A robotic system requires the whole structure of intelligence, and acquires skill and knowledge through interaction with a dynamic environment. Consequently, the whole structure determines the potentiality of intelligence. This paper proposes a robotic system with structured intelligence. Based on perceptual information, a robot with structured intelligence makes decision and action from four levels in parallel. In addition, the robot generates its motion through interaction with the environment, and at the same time, gradually acquires its skill based on the generated motion. To acquire skill and motion, the robot requires internal and external evaluations at least. This paper applies a virus-evolutionary genetic algorithm for trajectory planning and applies neural network for motion learning. Furthermore, we discuss its effectiveness through computer simulation results.

Virus-evolutionary genetic algorithm with subpopulations: application to trajectory generation of redundant manipulator through energy optimization

In this paper, we apply virus-evolutionary genetic algorithm with subpopulations (VEGAS) to trajectory generation of redundant manipulator through energy optimization. VEGAS is based on the virus theory of evolution, it has subpopulations that usually evolve independently. A virus infects host individuals in the same subpopulation. And a virus sometimes immigrates from a subpopulation to another subpopulation. The energy-optimized trajectory has successfully been found by VEGAS.