| 3 | 0 | 20 |
| 下载次数 | 被引频次 | 阅读次数 |
变风量空调（variable air volume,VAV）系统是一个大时滞、非线性系统，常规的模型预测控制（model predictive control,MPC）算法无法同时满足高控制精度和低能耗，因此提出了一种基于双阈值事件触发机制（event-triggered mechanism,ETM）的自适应MPC算法。为了实现降低能耗的目标，首先建立系统的能耗模型，将系统的能耗纳入MPC的成本函数中，实现多目标优化；其次，将事件触发机制与MPC算法结合，提出一种新的触发条件，即双阈值触发，利用系统预测误差和输出变化率，保证输出精度的同时赋予系统预测未来趋势的能力；随后，利用ETM设置了预测时域的更新方式，根据系统当前的动态特性实现预测时域的自适应更新；最后，进行了仿真实验，结果证明基于双阈值ETM的自适应MPC算法可以有效提高系统的控制效果、降低系统的能耗以及降低计算量，避免冗余计算。
Abstract：Variable air volume(VAV) system is a large time-delay, nonlinear system for which conventional model predictive control(MPC) algorithms cannot simultaneously achieve high control precision and low energy consumption. To address this issue, an adaptive MPC algorithm based on a dual-threshold event-triggered mechanism(ETM) was proposed. To accomplish the objective of reducing energy consumption, the following steps were taken. First, an energy consumption model of the system was established and integrated into the MPC cost function to enable multi-objective optimization. Second, an event-triggered mechanism was incorporated into the MPC framework, utilizing a dual-threshold trigger condition based on both system error and its rate of change. This approach not only ensures output accuracy but also equips the system with the ability to anticipate future trends. Third, the ETM was employed to adaptively adjust the prediction horizon online based on the current dynamic characteristics of the system. Finally, simulation experiments were conducted, and the results demonstrated that the proposed dual-threshold ETM-based adaptive MPC algorithm effectively enhances control performance, reduces overall energy consumption, and decreases computational burden by avoiding redundant calculations.
[1]SHI S, MIYATA S, AKASHI Y. A hybrid multiagent distributed optimal control strategy of multizone VAV systems for edge computing in smart buildings[J]. Energy and Buildings, 2025:116089.
[2]WANG H, CHEN X W, VITAL N, et al. Energy optimization for HVAC systems in multi-VAV open offices:a deep reinforcement learning approach[J].Applied Energy, 2024, 356:122354.
[3]SHI S, MIYATA S, AKASHI Y. Event-driven model-based optimal demand-controlled ventilation for multizone VAV systems:enhancing energy efficiency and indoor environmental quality[J]. Applied Energy,2025, 377:124683.
[4]ERDEM M K, GOKALP O, CALIS G. Prediction of HVAC operational variables using recurrent neural networks for advanced control strategies[J]. Journal of Building Engineering, 2025, 115:114474.
[5]LI B X, WANG S W. Multi-objective optimal control of multi-zone VAV systems for adaptive switching between normal and pandemic modes[J]. Building and Environment, 2023, 243:110626.
[6]YAO Y, SHEKHAR D K. State of the art review on model predictive control(MPC)in heating ventilation and air-conditioning(HVAC)field[J]. Building and Environment, 2021, 200:107952.
[7]YANG S Z, LIU Y L, CAO H D. Constrained DNNbased robust model predictive control scheme with adjustable error tube[J]. Symmetry, 2023, 15(10):1845.
[8]LI T P, CAO Y, YE Q, et al. Generative adversarial networks(GAN)model for dynamically adjusted weld pool image toward human-based model predictive control(MPC)[J]. Journal of Manufacturing Processes,2025, 141:210-221.
[9]WEI Z C, TIEN P W, CALAUTIT J, et al. Investigation of a model predictive control(MPC)strategy for seasonal thermochemical energy storage systems in district heating networks[J]. Applied Energy, 2024,376:124164.
[10]KLEPIC V, WOLF M, PRÖLL T. Extension of a low-tech model predictive control(MPC)algorithm for grid-supportive heat pump operation[J]. Energy&Buildings, 2024, 323:114733.
[11]FENG N, WU D F, YU H L, et al. Event-triggered distributed disturbance rejection model predictive control for unmanned surface vehicles under lumped disturbances and input saturation[J]. Ocean Engineering,2025, 328:120949.
[12]HE N, GUO J W, LI Y X, et al. An event-triggered stochastic model predictive control of indoor thermal environment for building energy management[J]. Journal of Building Engineering, 2025, 109:113026.
[13]CHEN Y D, WAN C, CHENG P. Event-triggered stochastic model predictive control based on distributionally robust optimization approach for network control systems under DoS attacks[J]. Signal Processing,2025, 238:110182.
[14]YANG Y, YAO X M, XU H Z. Disturbanceobserver-based event-triggered model predictive control of nonlinear input-affine systems[J]. Automatica,2024, 161:111504.
[15]HU X D, YU H, HAO F, et al. Event-triggered dualmode predictive control for constrained nonlinear systems with continuous/intermittent detection[J]. Nonlinear Analysis:Hybrid Systems, 2022, 44:101149.
[16]YANG S Y, CHEN W Y, WAN M P. A machinelearning-based event-triggered model predictive control for building energy management[J]. Building and Environment, 2023, 233:110101.
[17]李凯，苏延旭.含参数不确定和输入时延的线性系统自适应Tube模型预测控制[J].控制理论与应用，2025, 42(12):2409-2418.LI K, SU Y X. Adaptive Tube model predictive control for linear systems with parameter uncertainties and input delays[J]. Control Theory&Applications,2025, 42(12):2409-2418.(in Chinese)
[18]ZHANG L, ZHANG S Z, DU Z, et al. Adaptive trajectory tracking of the unmanned surface vessel based on improved AC-MPC method[J]. Ocean Engineering, 2025, 322:120455.
[19]冯小菲，杜嘉伟，徐中显，等.基于变时域策略的多智能体系统分布式事件触发MPC[J/OL].控制工程，2025:1-11[2025-11-18]. https：∥doi. org/10.14107/j. cnki. kzgc. 20240088.FENG X F, DU J W, XU Z X, et al. Distributed event-triggered MPC for multi-agent systems based on variable time domain strategy[J/OL]. Control Engineering of China, 2025:1-11[2025-11-18]. https：∥doi. org/10. 14107/j. cnki. kzgc. 20240088.(in Chinese)
[20]ESCAÑO J M, SÁNCHEZ A J, WITHEEPHANICH K, et al. Explicit simplified MPC with an adjustment parameter adapted by a fuzzy system[J].Journal of Intelligent&Fuzzy Systems, 2019, 37:1287-1298.
[21]QIU J D, LIN D Q, TANG M N, et al. Trajectory tracking of unmanned logistics vehicle based on eventtriggered and adaptive optimization parameters MPC[J]. Processes, 2024, 12(9):1878.
[22]ZHAO Y, LI W, ZHANG J L, et al. Real-time energy consumption prediction method for air-conditioning system based on long short-term memory neural network[J]. Energy and Buildings, 2023, 298:113527.
基本信息:
中图分类号:TU831.3;TU855
引用信息:
[1]杨世忠,刘诗成,魏红超.基于双阈值事件触发机制的VAV系统自适应MPC[J].中国科技论文,2026,21(04):270-281.
2026-04-15
2026-04-15