一种动态自适应最近邻聚类学习算法在工业污水处理中的应用

为了建立工业污水pH值中和系统的正模型,研究了具有大滞后非线性特性的加药中和过程。利用一种动态自适应最近邻聚类(DANNC)学习算法,全面调整网络参数完成了污水pH值加药中和控制系统网络的学习和训练。采用中和过程神经网络内模控制系统的逆模型充当控制器,进行了各种工业条件下污水中和的仿真实验。结果表明,该系统实现了△pH≤0.2的工业污水的控制精度目标,系统实时跟踪和抗干扰性良好。     

A Single-Mode,Self-Adapting,and Self-Powered Mechanoreceptor Based on a Potentiometric–Triboelectric Hybridized Sensing Mechanism for Resolving Complex Stimuli

Human skin is equipped with slow adapting (SA) and fast adapting (FA) capabilities simultaneously. To mimic such functionalities, elaborately designed devices have been explored by integrating multiple sensing elements or adopting multimode sensing principles. However, the complicated fabrication, signal mismatch of different modules, complex operation, and high power-consumption hinder their widespread applications. Here, a new type of single-mode and self-powered mechanoreceptor that can mimic both SA and FA via seamless fusion of complementary while compatible potentiometric and triboelectric sensing principles is reported. The resultant potentiometric–triboelectric hybridized mechanoreceptor exhibits distinctive features that are hard to achieve via currently existing methods, including single-mode output (only voltage signal), greatly simplified operation (single-measurement setup), ultralow power-consumption (<1 nW), self-adaptive response behavior, and good capability for resolving complex stimuli. Diverse mechanical characteristics, including magnitude, duration, frequency, applying and releasing speed, can be well interpreted with this single-mode and self-powered mechanoreceptor. Its promising application for monitoring object manipulations with a soft robotic gripper is explored. Furthermore, the versatility of the mechanoreceptor for resolving complex stimuli in diverse daily scenarios is demonstrated. This work presents a new design that will significantly simplify the fabrication/operation and meanwhile boost the functionality/energy-efficiency of future electronic devices and smart systems.     

模糊控制在永磁同步电机调速系统中的应用

传统PI控制器基于线性系统设计且参数固定,不能满足高性能控制的要求。而模糊控制不依赖于控制对象的数学模型,能够根据模糊信息实行精确控制。为了结合PI控制和模糊控制的优点,设计了一种模糊自适应PI控制器,并应用到永磁同步电机矢量控制调速系统中。该控制器能根据外部和内部参数的变化自动调节PI控制器的参数。仿真结果验证了,采用模糊自适应PI控制器的永磁同步电机矢量控制调速系统具有更好的动态和鲁棒性能。     

LNB温度PID-NN控制系统设计

针对低噪声功率放大器（LNB）温度控制系统的非线性,时变性,大滞后等特点,采用PID．NN控制策略来实现温度控制,并且给出了单片机控制实现的具体方案。实验结果表明,该电路简单可靠,运行稳定;对于低噪声功率放大器的温度控制,与普通PID控制器相比,PID-NN参数自适应算法具有较好的控制性能和较强的鲁棒性．     

神经网络动量—自适应学习率BP算法与BP算法的性能比较及其应用

动量－自适应学习率BP算法是对标准BP算法的改进,本文对这两种算法进行了分析,并利用计算机程序对其性能进行了比较测试,利用VISUAL C＋＋和MATLAB开发了仿真测试程序。     

Neural network adaptation process effectiveness dependent of constant training data availability

The paper discusses and compares two different ways of adapting artificial intelligence systems. One is founded on a well known biological mechanism of gradual training of neurons or other parameters. The second one uses a significant extra feature of training data that ably makes us possible to adapt the artificial intelligence system in more effective way than nature does in biological systems. This extra feature is availability of all training data before the adaptation process begins till an end of which all these data have to be constant. This feature provides an ability to analyze training data globally and very quickly tune an artificial intelligence system with them. The paper focus the attention on this important difference between biological and artificial intelligence problems because in most cases of artificial intelligence problems training data are gathered, available and constant during the training process. On the other hand, the biological nervous systems gather training data during the whole life, have to change the inner model, so training is a very good solution for them because it makes them possible to tune with changing training data. Artificial intelligence systems can also use training inherent in biological systems but in most cases it is possible to find more quickly and effectively the solution if only the mentioned feature is met. The above thesis is illustrated by means of some examples.     

单片机型GPS车载监控仪设计与软件优化

对使用MSP430F低功耗单片机开发车载GPS智能监控终端和系统的软、硬件的设计方案及实现过程进行了阐述.在商品级的终端和系统中,设计并实现了采样频率自适应模型和椭圆曲线密码体制(ECC),对终端与中心的通讯层面从现行通用方案进行了优化,达到提高通信效率、数据安全和降低运行费用的效果.     

基于变指数趋近律的起重机自适应滑模控制

以起重机升降系统位置控制为目标,针对传统滑模控制中要求系统不确定因素的界为已知的缺点,将自适应控制引入传统滑模变结构控制中,通过设计自适应律对起重机升降系统参数的变化及外部扰动进行辨识,同时采用变指数趋近律进一步克服传统滑模方法中存在的抖振问题。仿真结果表明,该控制算法不仅保持了传统滑模方法良好的鲁棒性,而且提高了系统的控制精度,减小了系统的抖振。     

一种竞争型量子进化算法

针对连续空间数值优化问题,提出了一种竞争型量子进化算法。粒子每次向全局最优和种群均值两个方向分别进化,从而得到两个子粒子。根据“优胜劣汰”原则选择适应度较高者作为下一代。同时,为了保证粒子的多样性,引入了一种自适应变异机制：对适应度较低的粒子以较高概率进行变异,而对适应度较高粒子以较低概率进行扰动。通过5个标准测试函数验证了算法的性能。仿真结果表明,与PAQEA及NVCQEA相比,该算法收敛速度快,收敛精度高,稳定性好。