基于模型的BP网络辨识算法代码快速生成

Model-based BP network identification algorithm fast code generation

采用MATLAB与Coder Composer Studio(CCS)编程工具联合编程,将经典BP神经网络运用于嵌入式火势辨识系统。使用模型的嵌入式代码生成技术,利用Simulink、DSP System Toolbox、Embedded Coder Support Package for Texas Instruments C2000 Processors等工具箱,在MATLAB平台上完成火势辨识系统模型的搭建以及代码生成。将生成的代码加入CCS开发平台,新建工程,修改代码,编译下载到目标板中运行。经证明,该方法可以缩短神经网络算法在嵌入式系统的开发周期同时降低开发难度。

Using MATLAB and Coder Composer Studio programming tool in conjunction with the programming,the classic BP neural network is applied in the embedded fire identification system. By utilizing the model＇s embedded code generation technology,the toolbox,such as Simulink,DSP System Toolbox,Embedded Coder Support Package for Texas Instruments C2000 Processors is used to build the fire identification system model and generate code on the MATLAB platform. The generated code is added to the CCS development platform to build the project,modify the code,compile and download it to the target board. It is proved that this method can shorten the development cycle of the identification algorithm in the embedded system and reduce the difficulty of development.