基于LS的人眼血流自动调节系统模型研究

发布时间：2018-02-24 14:38

本文关键词： 眼血流自动调节系统辨识线性模型非线性模型　出处：《哈尔滨商业大学》2016年硕士论文　论文类型：学位论文

【摘要】：目前,眼血流循环失调已被证实是许多眼部疾病的独立危险因素之一,但是血流自动调节系统受损的直接原因并没有明确定论。而且有关眼血流自动调节的研究并未深入探讨眼血流系统的生理机制对眼血流自动调节系统进行建模研究。研究眼血流自动调节系统,可以了解眼睛组织的供血机制,为眼部疾病的诊断、治疗和预防提供依据。本研究的实验数据是在cuff(下肢束带法)条件下用LSFG(激光散斑血流成像)仪器采取眼血流信号,与此同时,用Finometer仪器对眼血压信号进行同步采集。眼血流、眼血压是能量集中在低频的周期非平稳随机信号,在采集的过程中易受仪器、呼吸、肢体动作等干扰,通常带有较强的噪声。本文采用小波去噪等方法滤除信号的基线漂移、工频干扰、肌电干扰等噪声。眼血流自动调节机制是指在一定眼血压范围内,眼血流量保持相对恒定的生理机能。目前研究眼血流自动调节机制最新的方向是研究系统的线性或非线性模型。本文主要利用系统辨识技术建立眼血流自动调节系统系统模型。该技术中的模型参数辨识选用LSM算法。LSM算法是目前解决一些实际问题中使用最广泛的经典数据处理方法之一。因为在0.05～0.3Hz频率范围内,眼血压与眼血流之间的关系更多地表现为线性关系。所以本文首先选用了线性ARX模型建立系统模型,然后转化系统的传递函数参数模型并分析了模型的稳定性。为了能够全面的了解眼血流自动调节系统的线性特性与非线性特性,本文也尝试建立系统的非线性模型。鉴于非线性模型可重复性差的特性,本文提出采用将线性和非线性相结合的Hammerstain模型建立系统模型。该模型的预测数据与实际数据拟合率为86.81%,而ARX模型的拟合率为83.92%。当使用另一组样本验证这两个模型时,Hammerstain模型的拟合率为82.07%,而ARX模型的拟合率为63.7%。Hammerstain模型的可重复性优于ARX模型,适用性更强。
[Abstract]:At present, ocular blood flow disorder has been proved to be one of the independent risk factors for many eye diseases. However, the direct cause of the damage of the automatic blood flow regulation system is not clear. Moreover, the research on the automatic regulation of the eye blood flow has not deeply discussed the physiological mechanism of the eye blood flow system, and the modeling of the automatic eye blood flow regulation system has been carried out. To study the automatic regulation system of eye blood flow, It can be used to understand the blood supply mechanism of eye tissue and to provide evidence for the diagnosis, treatment and prevention of ocular diseases. The experimental data of this study are to use the LSFG (laser speckle flow imaging) instrument to take eye blood flow signals under the condition of cuff (lower extremity banding method). At the same time, the Finometer instrument is used to synchronously collect blood pressure signals. Eye blood flow and eye blood pressure are non-stationary random signals whose energy is concentrated in low frequency. They are easily disturbed by instruments, breathing and limb movements. In this paper, wavelet denoising method is used to filter the baseline drift, power frequency interference, myoelectric interference, etc. The automatic regulation mechanism of eye blood flow is within a certain range of eye blood pressure. At present, the newest research direction of eye blood flow automatic regulation mechanism is to study the linear or nonlinear model of the system. This paper mainly uses the system identification technology to establish the eye blood flow automatic regulation. In this technique, LSM algorithm. LSM algorithm is one of the most widely used classical data processing methods to solve some practical problems, because in the frequency range of 0.05 ~ 0.3Hz, The relationship between eye blood pressure and eye blood flow is more linear. Therefore, the linear ARX model is used to establish the system model in this paper. Then the transfer function parameter model of the system is transformed and the stability of the model is analyzed. This paper also attempts to establish a nonlinear model of the system. In view of the poor repeatability of the nonlinear model, This paper presents a system model based on linear and nonlinear Hammerstain model. The fitting rate of the predicted data and the actual data of the model is 86.81, while the fitting rate of the ARX model is 83.92. When another set of samples is used to verify the two models, The fitting rate of Hammerstain model was 82.07, while that of ARX model was 63.7. The reproducibility of Hammerstain model was better than that of ARX model. The applicability is stronger.
【学位授予单位】：哈尔滨商业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：R77

【相似文献】