4 results about "Polynomial method" patented technology

The invention discloses a method for optimizing operation time and impact of a Niryo One robot based on a multi-strategy improved dung beetle optimization algorithm, and belongs to the technical field of multi-degree-of-freedom industrial robot control. Four interpolation points are set for the motion trail of the Niryo One robot for optimization; on the basis of three sections of motion tracks of Niryo One, six joint motion tracks of each section are described through cubic polynomial interpolation, quintic polynomial interpolation and cubic polynomial interpolation methods, and a 3-5-3 mixed polynomial method is constrained; after six joint coefficients of different polynomials of Niryo One are solved, the influence of multiple constraints such as optimal time and impact and the maximum speed and acceleration of a motor of each joint is synthesized, and a fitness function which finally needs to be optimized is defined; and optimizing the trajectory of the Nryo One by using a multi-strategy improved dung beetle optimization algorithm, and returning the optimal operation time and fitness value of each section of trajectory after optimization. Compared with an original dung beetle optimization algorithm and other algorithms, the method has the advantages that the iteration speed is higher, a better solution can be found, and the working efficiency of the Niryo One robot in use is improved.

The invention belongs to the technical field of electronic information and wind tunnel test measurement, and particularly relates to a wind tunnel six-component balance dynamic decoupling method and application based on multi-mode transfer learning and physical constraint, and the method comprises the following steps: feature extraction and data preprocessing; the preprocessed data features are input into a pre-trained convolutional neural network ResNet-18 model, and hierarchical data feature extraction is carried out; the processed image feature data, strain feature data and vibration and temperature feature data are input to a full splicing layer for heterogeneous fusion splicing, the spliced data are input to a pre-trained machine learning regression XGBoost model for feature fusion regression and mapping, and finally six aerodynamic force / torque components acting on the aircraft are output. Through fusion of multi-modal transfer learning and physical constraint, in actual measurement of subsonic and hypersonic wind tunnels, the dynamic decoupling precision is improved by more than 50% and 65% compared with a traditional least square method and a polynomial method.

This invention discloses a method for optimizing the dynamic transmission error distribution characteristics of a harmonic reducer, comprising: establishing a static transmission error probability model to obtain the probability distribution of the overall static transmission error; constructing a dynamic transmission error mathematical model considering static transmission error and dynamic parameters; constructing a high-precision surrogate model of dynamic transmission error including the probability distribution of static transmission error and the range of dynamic parameters; obtaining the probability distribution of dynamic transmission error; and dynamically adjusting the range of dynamic parameters based on a particle swarm optimization strategy to find the parameter range that optimizes the dynamic transmission error distribution. By introducing static transmission error into the dynamic model of the harmonic reducer, considering the influence of the processing and assembly of the harmonic reducer on the transmission error, a probability-range hybrid uncertainty model is used to mathematically describe the probability distribution characteristics of static transmission error and the range of dynamic parameters, and a Chebyshev polynomial method is used to construct an approximate model of dynamic transmission error.

This invention relates to the field of geophysical exploration technology, specifically to a high-precision three-dimensional forward modeling method for airborne electromagnetic spectral elements considering induced polarization effects. Based on the Cole-Cole model, a complex conductivity expression considering induced polarization effects is established, and the frequency domain Helmholtz equation is introduced to construct the corresponding governing equations. A regular hexahedral mesh is generated, and reference domain vector basis functions are constructed using the Gauss-Lobatto-Legendre orthogonal polynomial method according to the spectral element method. The mapping relationship between these functions and physical mesh elements is established. The Galerkin weighted residual method is applied to discretize the governing equations into element system matrices, which are then assembled to form the overall system matrix and the corresponding linear equation set. A direct solver is used to obtain the frequency domain three-dimensional electromagnetic response, and the time domain three-dimensional forward modeling result of the computational region is obtained through Hankel transformation. This invention achieves rapid and high-precision three-dimensional forward modeling simulation of complex geological models containing induced polarization effects.