Optimization of Two-Stage Cooling and Dehumidification System by Multi-Strategy Improved Parrot Optimizer Algorithm

To further explore the energy-saving performance of liquid desiccant dehumidification air conditioning systems, this paper constructs a performance optimization model based on the established air temperature and humidity prediction model and the actual physical constraints of the system’s operation. The total system energy consumption and cooling/dehumidification performance are adopted as indicators in the optimization model. Furthermore, to address the nonlinearity, multivariate and constraint in the optimization model, a multi-strategy improved parrot optimizer algorithm (MSPOA) is proposed, incorporating Cauchy inverse mapping initialization, adaptive t-distribution mutation, and random walk strategies into the standard parrot optimizer algorithm. To evaluate the performance of the proposed MSPOA, it is compared with other four algorithms using six CEC benchmark test functions. The results show that MSPOA has higher optimization accuracy and faster convergence speed. In addition, a MSPOA-based optimization control strategy is implemented on a two-stage cooling and dehumidification system experimental platform, and its energy consumption is compared with that of the traditional control strategy. The results show that, on the premise of meeting the requirements of cooling and dehumidification performance, the total system energy consumption under the optimized control strategy is reduced by 6.57 kWh compared with the traditional control strategy, and the energy-saving rate reaches 27.91%, verifying the effectiveness and engineering application value of the proposed method.