Internal Space Optimization Of Intelligent Voice Robot Based On SVR-PSO

With the advancement of human-computer interaction technology and the development of social economy,intelligent robots are playing an increasingly important role in our daily life.Among them,voice robots,which can directly interact with people and serve human beings,are widely used in intelligent home,smart transportation and intelligent medical treatment.For some specific application scenarios,such as family companionship,the voice robot is required to be small and exquisite in appearance,not too large in size,and must have a certain load-bearing capacity to ensure that the interior of the fuselage can accommodate necessary parts such as motors and batteries.Because the outer shell of a robot reflects its internal space,the smaller the outer shell of the robot is under the constraints of accommodating internal parts,the smaller the volume of its inner space.As a result,this research explores an effective method for increasing the internal space utilization of the intelligent voice robot by adjusting the shape parameters of the robot.Focusing on this goal,the main research carried out in this paper is as follows:Firstly,the overall parameters optimization design framework of the robot is established,and the shell of the voice recognition robot is 3D modeled by Solidworks software,then we derive the 3D coordinate points of the shape surface.Next,polynomial fitting was carried out for the coordinate points to obtain the fitting equation,and the parameterized representation of the shape shell was realized to prepare for the following optimization.Secondly,the components inside the robot,such as lidar,audio and so on are modeled and combined assembly,and then the component assembly is represented by a function.Finally,the internal space optimization of the robot is carried out with the volume minimization as the optimization objective.In view of the problems that traditional optimization algorithms in robot optimization applications,such as have high requirements of mathematical properties of the model,slow running speed,and difficulty in accurately obtaining the global optimal solution,this paper considers introducing particle swarm algorithm to optimize the robot.Particle swarm optimization(PSO)is a swarm intelligent optimization algorithm that mimics the behavior of individuals and groups searching for food when birds are foraging.The algorithm is simple to implement,does not depend on mathematical information such as the gradient of the objective function,and can satisfy the complex design variable types.It is an effective tool for solving nonlinear optimization problems and combinatorial optimization problems.However,when we tried to introduce the particle swarm algorithm into the shape design of the voice robot,we found that the algorithm still has the problem that the convergence speed is slow in the late iteration and it is easy to converge to the local optimal solution prematurely.Therefore,in this paper,we propose three improved algorithms for particle swarm,namely,chaotic particle swarm optimization algorithm,genetic particle swarm optimization algorithm based on crossover and mutation operation,and support vector regression-based particle swarm optimization algorithm,then we compare the convergence speed and calculation accuracy of the three improved algorithms with the basic particle swarm algorithm through experiments.Since the fitness functions in high-dimensional multimodal optimization problems are extremely complex,for example,there are 56 design variables in robot optimization of this paper,resulting in high calculation cost.Therefore,when analyzing the experimental results,we replace the actual running time of the algorithm with the number of calculations of the fitness function.The experimental results show that the support vector regression-based particle swarm optimization algorithm converges faster,and has higher stability and calculation accuracy among the four algorithms.So,particle swarm algorithm based on support vector regression is finally selected in this paper to adjust the shape parameters of the voice robot.Under the constraints of accommodating the robot's internal parts,the volume of the robot is reduced to realize the internal space optimization of the robot.