Hilbert space multi-kernel function multiplication based wind speed prediction method

Abstract

The invention provides a Hilbert space multi-kernel function multiplication based wind speed prediction method which comprises the following steps that: pulsation wind speed samples of vertical space points are generated by utilizing ARMA model simulation, and the pulsation wind speed sample of each space point is divided into two parts, namely a training set and a test set; a kernel function theory of a support vector machine is discussed in complete inner product space, both linear addition and product operation belong to closed operation in Hilbert space, and the results of linear addition and product operation still belong to the Hilbert space; and a multiplicative combination kernel function based on a global kernel function and a local kernel function is created, a model of a least-square support vector machine (LSSVM) using the multiplicative combination kernel function is established, model parameters are optimized by adopting particle swarms, and the pulsation wind speed of a single point is predicted by utilizing the model. A test sample is compared with a pulsation wind speed result predicted by the PSO-LSSVM using the multiplicative combination kernel function to ensure the accuracy for pulsation wind speed prediction.

Description

technical field [0001] The invention relates to a single-point fluctuating wind speed prediction method using a multiplication operation that satisfies the closed rule of Hilbert space to multiplicatively combine existing kernel functions, construct a kernel function based on multiplicative combination, and introduce a particle swarm optimized least squares support vector machine. Specifically, it is a wind speed prediction method based on the multiplication of multi-kernel functions in Hilbert space. Background technique [0002] When studying wind loads, the wind is usually treated as an average wind speed that does not change with time within a certain time interval and a fluctuating wind speed that varies randomly with time. The average wind speed produces a static response of the structure, while the fluctuating wind speed produces a dynamic response. When the wind acts on a high-rise structure, its positive and negative wind pressure will form a wind load on the struct...

AI Technical Summary

Problems solved by technology

However, these methods have shortcomings in theory or application. For example, although the support vector machine (SVM) transforms the input space into a high-dimensional space through the nonlinear transformation defined by the kernel function, it looks for input variables and output variables in this high-dimensional space. A nonlinear relationship between variables solves the "curse of dimensionality" problem, but the choice of kernel function determines the characteristics of the model. The local kernel function has strong learning ability and weak generalization performance, while the global kernel function has strong generalization performance. , weak learning ability

Therefore, the performance of the kernel function varies greatly in different applications, especially when the sample features contain heterogeneous information or the sample size is large, or the data is unevenly distributed in the high-dimensional feature space, the selection of the existing kernel function is effective for all samples. It is unreasonable to process

Images