关于时滞复杂网络的精密温度自动控制研究

Research on Precise Temperature Automatic Control of Complex Network with Time Delay

由于控制系统复杂度越来越高,传统控制器在很多场合下无法满足精密温度控制要求。为此提出关于时滞复杂网络的精密温度自动控制系统。针对温度控制过程中的时变时滞和输入时滞等因素,建立了由多个耦合节点组成的复杂网络模型,基于Lyapunov稳定性定理,引入环函数使采样点在温度点之外无约束,有效降低系统的保守性。为了保证系统的稳定性,由Schur引理以及导数负定性对非线性项进行线性化处理。采用由电桥和运算放大器等构成的温度采集系统,为了能显示出对应温度下的电阻值,对采集的数据进行最小二乘法拟合,并对得出的温度与电阻关系模型进行线性化处理,经过多组实验得出数据值的均值拟合函数。在双闭环函数中构造出更多温度状态信息的增广向量,进一步对泛函数优化处理,并过MATLAB中的工具箱求解出温度采样控制器。实验结果表明,构建的时滞复杂网络模型具有更好的保守性,所设计的控制器是可行的,能够明显降低温度采集的标准差,在精密温度自动控制过程中表现出良好的优越性。

Due to the increasing complexity of the control system, the traditional controller can not meet the requirements of precision temperature control in many cases. Therefore, a precision temperature automatic control system based on time-delay complex network is proposed. Aiming at the time-varying delay and input delay in the process of temperature control, a time-varying delay complex network model composed of multiple coupling nodes was established. Based on the stability theorem, a loop function was introduced to make the sampling point unconstrained outside the temperature point, which effectively reduced the conservatism of the system. In order to ensure the stability of the system, the nonlinear term was linearized by Lemma and derivative negativity. In order to display the resistance value under the corresponding temperature, a temperature acquisition system composed of electric bridge and operational amplifier was used to fit the collected data with the least square method, and the relationship model between temperature and resistance was linearized. The mean value fitting function of the data value was obtained through multiple groups of experiments. In order to construct more augmented vectors of temperature state information in the double closed-loop function, the pan function was further optimized, and the temperature sampling controller was solved through the toolbox in MATLAB. The experimental results show that the time-delay complex network model has better conservatism, and the designed controller is feasible, which can significantly reduce the standard deviation of temperature acquisition, showing good advantages in the process of precision temperature automatic control.