扩散硅压力传感器迟滞建模与补偿研究硕士.doc

硕 士 学 位 论 文题目扩散硅压力传感器迟滞建模与补偿研究〔英文〕Study on Modeling and Compensation for Hysteresis of Diffused Silicon Pressure Sensor / 摘 要扩散硅压力传感器是目前用于测量压力的一种重要传感器,被广泛应用于工业、农业、气象部门、航天航空等领域随着半导体技术的发展,扩散硅压力传感器的各项技术指标不断提高,但扩散硅压力传感器本身固有的非常规、多值对应的迟滞非线性特性,限制了其测量精度目前国外针对传感器迟滞特性的研究还比较少,已有文献证明不能直接使用神经网络对其进行建模,大部分学者的工作是把迟滞非线性当作一般非线性处理,其结果不能满足补偿精度的要求因此本文引入迟滞算子的概念,建立T-S模糊动态迟滞模型,实现对迟滞的智能补偿,并在此基础上设计了基于ARM的嵌入式智能扩散硅压力传感器补偿系统,本文主要获得如下成果：1. 借鉴提升输入维数的思想,引入一个迟滞算子,将迟滞非线性的多值映射转化为单值映射,建立一个基于算子的T-S模糊动态迟滞模型,该模型具有结构简单、易于辨识算法等优点。

2. 引入与迟滞算子作用类似的迟滞逆算子,将多值映射转化为单值映射,建立一个基于逆算子的T-S模糊动态迟滞逆模型,将该逆模型置于被补偿对象后面进行迟滞非线性补偿,实现了对扩散硅压力传感器的智能补偿3. 在32位ARM微处理器的EM9000嵌入式模块平台上,设计了Windows CE嵌入式扩散硅压力传感器补偿系统,完成了嵌入式扩散硅压力传感器系统的设计与开发,并结合迟滞智能补偿原理,实现了对扩散硅压力传感器高精度测量的实时性要求,与未补偿的扩散硅压力传感器迟滞相比精度提高了约四倍,证明了建立的T-S模型是非常有效性的关键词：迟滞算子,T-S模糊模型,嵌入式系统,Windows CE,ARMAbstractThe diffused silicon pressure sensor is an important device to measure the pressure,which has been widely used in the field of industry, agriculture, meteorology, aviationand so on. With the development of semiconductor technology, the performance of diffused silicon pressure sensor is improving. However, The hysteresis characteristic of diffused silicon sensor is a kind of non-smooth non-linearity with mult-ivalued mapping, whichreduces the precision of measurement. Up till now, few authors study the hysteresis of sensors and published literature proved that neural networks can’t approximate the hysteresis non-linearity directly. The compensation for the hysteresis of the sensor usually employs the conventional methods,but the methods can’t totally meet the requirment of compensation precision. Therefore, a hysteretic operator of simple structure is introduced to establish the hysteresis model and put forward to make intelligent compensation to the hysteresis. Hence, an embedded smart diffused silicon pressure sensor system based on ARM is designed, and the main results are as follows:1. Thehysteretic operator based on the input extended space method is proposed to transformthe multi-valued mapping into a one-to-one mapping. In this method, a novel hystereticoperator is constructed to extend the input space to establish T-S fuzzy dynamic hysteresis model. It has a rather simple structure,simplifies identification procedure. 2. Hysteretic inverse operator which has the simlar functions with hysteresis operator, i.e. transforming the multi-valued mapping into a one-to-one mapping, is proposed. Then, T-S fuzzy dynamic hysteresis inversemodel is presented, and nonlinear hysteresis characteristics are compensated by the inverse modelbehind, which realizes intelligent compensation for the hysteresis.3. On the 32-bits ARM microprocessor EM9000 embedded module, the Windows CE embedded compensation system about the diffused silicon pressure sensor is designed. Then, the system about the diffused silicon pressure sensor is developed, which combines with principle of the compensation. Such reaches the real-time demand of the diffused silicon pressure sensor, four timeshigher than other method without compensationin precision. Result shows that T-S model is extraordinarily effective.Key words：hysteresis operator,T-S fuzzy model,embedded system,Windows CE,ARM目 录摘要IABSTRACTII目录III第一章绪论1§1.1扩散硅压力传感器基本结构和原理介绍1§1.1.1 扩散硅压力传感器基本结构1§1.1.2 扩散硅压力传感器基本原理2§1.1.3 影响扩散硅压力传感器迟滞特性的因素2§1.2 扩散硅压力传感器国外发展状况3§1.3 课题来源和主要研究容4§1.4 论文结构安排4第二章迟滞的定义及数学模型介绍6§2.1 迟滞的定义6§2.2 迟滞数学模型7§2.2.1 基于微分方程的模型8§2.2.1.1 Duhem模型8§2.2.1.2 Jiles-Atherton模型8§2.2.1.3 Chua-Stromsmoe模型9§2.2.1.4 Bouc-Wen模型9§2.2.2 基于算子类的模型10§2.2.2.1 Preisach模型10§2.2.2.2 KP 模型11§2.2.2.3 PI 模型12§2.2.2.4 TK模型13§2.2.3 基于神经网络的模型14§2.3 迟滞系统的补偿方案15§2.4 本章小结16第三章基于迟滞算子的迟滞非线性建模与补偿17§3.1 引言17§3.2 T-S模糊模型的数学描述18§3.2.1 基于T-S模糊模型的ANFIS结构19§3.2.2 T-S模糊模型的辨识20§3.3 动态T-S模糊迟滞模型22§3.4 扩散硅压力传感器迟滞非线性神经网络建模23§3.4.1 迟滞算子23§3.4.2 基于迟滞算子的模糊神经网络模型26§3.4.3 仿真与比较26§3.5 基于迟滞逆算子的T-S模糊逆模型建模与补偿30§3.5.1 迟滞逆算子31§3.5.2 基于迟滞逆算子空间维数的提升33§3.5.3 仿真与比较33§3.6 扩散硅压力传感器迟滞非线性的智能补偿36§3.7 本章小结38第四章嵌入式系统概述及设计开发39§4.1 嵌入式系统39§4.1.1 嵌入式系统的特点39§4.1.2 嵌入式系统的硬件和软件40§4.2 嵌入式微处理器40§4.3 Windows CE嵌入式操作系统41§4.3.1嵌入式操作系统41§4.3.2 Windows CE嵌入式操作系统42§4.3.2.1 Windows CE系统简介42§4.3.2.2 Windows CE的特点43§4.3.2.3 Windows CE体系结构和功能44§4.4 基于Windows CE的嵌入式系统开发流程45§4.5本章小结47第五章嵌入式智能扩散硅压力传感器补偿系统设计与实现48§5.1 嵌入式扩散硅压力传感器系统的总体设计48§5.2 压力测量信号的的采集49§5.3 采样信号的滤波处理51§5.4 迟滞补偿算法52§5.5 LCD显示53§5.6 调试与运行55§5.7 本章小结60第六章结论与展望61§6.1 结论61§6.2 进一步需要完善的工作62致63参考文献64作者读研期间已发表的论文69第一章 绪论随着人类探知领域的拓展和空间的延伸,人们越来越依赖于获取外界信息的采集技术。

传感器作为人类探知自然界信息的触角,可将人们需要探知的各种非电量信息转化为电量信息,提供认识和控制相应对象的条件和依据近年来,随着对压阻效应非线性的深入研究及半导体技术的发展,扩散硅压力传感器因具有灵敏度高、性能稳定、量程围广和大批量生产等特点,被广泛应用于工业、农业、气象部门、航天航空等领域然而,存在于这些半导体材料中的迟滞非线性不但会限制了传感器测量精度的进一步提高,甚至会导致测量出现严重偏差传感器的迟滞特性是一种非常规非平滑的非线性,它的复杂性表现在：<1> 多映射性：在相。