| DIRECT DISPLACEMENT OF PARALLEL MECHANISM WITH WAVELET NETWORK |
| |
| 作者姓名： | CHEN Weishan CHEN Hua LIU Junkao School of Mechatronics Engineering Harbin Institute of Technology Harbin China |
| |
| 作者单位： | CHEN Weishan CHEN Hua LIU Junkao School of Mechatronics Engineering，Harbin Institute of Technology，Harbin 150001，China |
| |
| 摘 要： | A new method solution for the direct displacement of parallel mechanism,wavelet network method,is proposed. Comparing with the classical analytical and numerical methods,this method can be extended to any parallel mechanism with any selected degree of freedom and configuration. A wavelet network suiting to approach multi-input and multi-output system is constructed. The network is optimized by analyzing the sparseness of input data and selecting the fitting wavelets by orthogo-nalization method according to the output data. Then it is applied to solve the direct displacement of a general six-degree-of-freedom parallel mechanism as a numerical example. For comparison purposes,a BP neural network is also used for this problem. Simulation results show that the wavelet network performs better than BP neural network. In addition,the wavelet network learns much faster than BP network.
|
| 关 键 词： | 并联机构 直接位移 小波网络 运动学 运动控制 |
DIRECT DISPLACEMENT OF PARALLEL MECHANISM WITH WAVELET NETWORK |
| |
| CHEN Weishan CHEN Hua LIU Junkao School of Mechatronics Engineering,Harbin Institute of Technology,Harbin ,China.DIRECT DISPLACEMENT OF PARALLEL MECHANISM WITH WAVELET NETWORK[J].Chinese Journal of Mechanical Engineering,2007,20(2):69-72. |
| |
| Authors: | CHEN Weishan CHEN Hua LIU Junkao |
| |
| Affiliation: | School of Mechatronics Engineering, Harbin Institute of Technology, Harbin 150001, China |
| |
| Abstract: | A new method solution for the direct displacement of parallel mechanism, wavelet network method, is proposed. Comparing with the classical analytical and numerical methods, this method can be extended to any parallel mechanism with any selected degree of freedom and configuration. A wavelet network suiting to approach multi-input and multi-output system is constructed. The network is optimized by analyzing the sparseness of input data and selecting the fitting wavelets by orthogonalization method according to the output data. Then it is applied to solve the direct displacement of a general six-degree-of-freedom parallel mechanism as a numerical example. For comparison purposes, a BP neural network is also used for this problem. Simulation results show that the wavelet network performs better than BP neural network. In addition, the wavelet network learns much faster than BP network. |
| |
| Keywords: | Direct displacement Parallel mechanism Wavelet network WAVELET NETWORK PARALLEL MECHANISM DISPLACEMENT faster BP network addition Simulation results show better BP neural network used problem comparison general numerical example analyzing input fitting |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
