带式输送机永磁驱动系统自抗扰同步控制策略

针对带式输送机永磁驱动系统采用PID控制器进行控制时电机的控制性能差和采用多电机驱动时电机间转速不同步、稳定性差等问题,提出了带式输送机多永磁电机驱动系统自抗扰同步控制策略。基于自抗扰控制(ADRC)技术设计了2种调速控制策略,并搭建仿真模型进行了对比试验;随后结合模糊PID控制技术对传统偏差耦合控制进行结构改进,并以矿用带式输送机多永磁电机驱动系统为控制对象,开展基于主从、传统偏差耦合和改进型偏差耦合同步控制结构的仿真。结果表明,基于二阶ADRC和改进型偏差耦合控制结构的带式输送机多永磁电机驱动系统同步控制策略具有更好的抗干扰能力、控制精度和同步性,有利于带式输送机平稳高效的安全运行。     

Buckling analysis of the shell of a refuge chamber in a coal mine under uniform axial compression

A stiffened cylindrical shell is normally used in refuge chambers of a coal mine.Based on the method of application and shape characteristics of a refuge chamber,we simplified its shell as an orrhotrop...     

联合支护技术在控制基础底鼓中的应用

鸣山矿在-450 m水平向下延深到-600 m时,布置有一条主井、一条副井.先施工一条副井采用1.6 m的绞车.绞车基础的岩性差,由于只在绞车基础下方平铺设了间距0.6 m×0.6 m的工字钢,连同绞车基础浇灌,支护厚度和强度不能达到足够的稳定的范围,基础浇好后,不到一年时间里绞车基础经测定底鼓量达65 mm,基础两侧高达158 mm,给提升安全带来隐患.     

神经网络用于化学知识的表达与处理

本文采用误差反传播算法,探讨了 Minsky 神经网络用于化学知识表达与处理的可行性。通过几个实例的研究表明:用神经网络方法能较好地处理一些多因子,多类和非线性问题,这种方法将知识和推理结合在一起并具有自学习能力。在化学专家系统的研究中,这将是一种有发展前途的方法。     

置信度网络在判别分析中的应用

本文叙述了置信信度网络与偏最小二乘法联用方法的建立，利用置信度网络处理信息的不完整性，利用偏最小二乘法建立预报模型，并预报结果。     

Development and applications of rescue robots for explosion accidents in coal mines

It is essential to provide disaster relief assistance after coal mine explosions. Often, it is life‐threatening for rescuers to enter an accident scene blindly; therefore, a coal mine rescue robot (CMRR) has been developed. However, the application of the CMRR has not proven satisfactory after decades of development. To solve this problem, we summarize the reasons for this disappointing state and address the technical challenges of the CMRR. Based on these reasons and the associated technical challenges, two generations of tracked robots have been developed. The China University of Mining Technology‐V (CUMT‐V) (A) robot was first developed and its walking system, body support system, communication system, environmental awareness system, and control system are described in detail. A performance test was performed on the CUMT‐V (A) robot and some problems were encountered. To address these problems, we designed the CUMT‐V (B) robot. The field test was conducted in Shanxi province, China, in August 2016. The application results show that the robot has good adaptability to complex terrain and high reliability in terms of environmental awareness and data transmission. In conclusion, the robot is nearing practical applications.     

AOAC 991.14 PetrifilmTM大肠菌群测试片法与GB4789.3大肠菌群计数法的比较研究

将AOAC 991.14 PetrifilmTM大肠菌群测试片与我国国家标准GB4789.3进行比较并做验证试验。在20个国家和食品企业实验室对生肉、熟肉、水产品、调味品、冰品、原奶和奶粉7大类食品进行试验与分析。试验结果表明,用PetrifilmTM方法与国标GB4789.3法检测这7类食品样品中的大肠菌群,所得结果高度一致,说明使用PetrifilmTM大肠菌群测试片法检测食品中的大肠菌群等效于国标GB4789.3方法。     

正交递归选择法及其在转炉炉龄研究中的应用

本文提出了一种新的变量筛选法—正交递归选择法,该法可以得到预报能力较强的模型,即PRESS(预报残差平方和)值较低的模型。转炉炉龄问题作为研究实例,并与逐步回归正向选择法及偏最小二乘回归法进行了比较。     

Feature extraction under bounded noise background and its application in low speed bearing fault diagnosis

We investigated the low speed bearing fault diagnosis under bounded noise background in three typical conditions. These situations included when the frequency of bounded noise was much larger, a little larger, and a little smaller than the driving frequency. Through the investigation of a harmonic signal in these three situations, we found the effect of stochastic resonance method was the best in the first situation, while it was invalid in the third situation. For the third situation, we introduced the vibrational resonance method by adding an auxiliary signal to extract the character frequency successfully. Then we also studied the low speed experimental bearing fault signal under bounded noise background, and it showed the same results as those of the simulation harmonic signal case. The methods and results of this paper might be useful in practical engineering, especially in extracting the weak character signal in the bounded noise background.

     

利用偏最小二乘法的一种变量筛选法

根据偏最小二乘法(PLS)建模中的回归系数等一些信息,筛选原始自变量,在不损失模型预报能力的前提下,除去冗余的或影响不大的一些原始自变量,使模型更简单。本研究中找到了用于删除变量的一种新判据,计算简单,使用效果好。研究结果表明,利用PL3法得到的删除变量的新判据筛选变量是一种非常实用和有效的变量筛选方法,该法非常适合处理海量数据或变量数很大的建模问题,可使最终所得的模型中变量数大大减少,使模型大大简化,因而便于分析和解决实际问题。在处理中药指纹图谱数据时,与传统的算法比较,模型得到了大大简化。