基于PSO优化Elman神经网络的恒温恒湿控制策略研究

Research on Constant Temperature and Humidity Control Strategy Based on PSO-Optimized Elman Neural Network

摘要: 医药洁净车间恒温恒湿空调系统能提供最佳的温湿度环境保障方案，有效确保药品生产质量合规并节省系统运行能耗。武汉地区亚热带季风气候全年剧烈波动，使常规PID控制方法在实际应用中暴露出抗干扰能力弱、大时滞响应迟缓、强耦合难以解耦等突出缺点。为此，针对武汉光谷某医药企业片剂生产洁净车间恒温恒湿空调系统，提出了一种基于粒子群优化(PSO)算法改进Elman递归神经网络的控制策略。以该洁净车间温湿度双回路控制策略为依据构建仿真模型，将基于PSO算法优化的Elman神经网络控制器与基础Elman控制、传统PID控制进行MATLAB仿真对比，实验表明，该方法可有效减小系统超调、缩短动态调节时间，并能以更高效率完成神经网络权值参数的全局寻优，为武汉地区医药洁净车间恒温恒湿系统智能化升级提供了切实可行的技术方案。

Abstract: The constant temperature and humidity air conditioning system in pharmaceutical cleanrooms can provide the optimal temperature and humidity environment protection scheme, effectively ensuring compliance with drug production quality while saving system operating energy consumption. The subtropical monsoon climate in Wuhan experiences drastic fluctuations throughout the year, which exposes significant shortcomings of conventional PID control methods in practical applications, such as weak anti-interference capability, slow response due to large time delays, and difficulty in decoupling strong coupling. To address this, for the constant temperature and humidity air conditioning system of a tablet production cleanroom in Wuhan’s Optics Valley pharmaceutical enterprise, a control strategy based on the particle swarm optimization (PSO) algorithm improved Elman recurrent neural network is proposed. Based on the temperature and humidity dual-loop control strategy of the cleanroom, a simulation model is constructed, and MATLAB simulations are carried out to compare the PSO-optimized Elman neural network controller with the basic Elman control and traditional PID control. The results show that the proposed method has lower overshoot and shorter regulation time, can more efficiently search for the optimal neural network weight parameters, and provides a practical and feasible technical solution for the intelligent upgrading of constant temperature and humidity systems in pharmaceutical cleanrooms in Wuhan.