新型电力系统背景下，通过虚拟电厂对需求侧灵活性资源进行高效聚合与调控是提升新型电力系统灵活性的有利措施。空调负荷是最具潜力的需求侧资源之一，对其精准建模可为控制策略的设计奠定基础。针对当前空调系统参数辨识拟合精度有限、忽略环境变化因素以及求解效率低下等问题，本文提出了一种基于改进的牛顿拉夫逊优化算法(improved Newton-Raphson-based optimizer, INRBO)的空调二阶等效热力学参数(equivalent thermal parameters, ETP)解析解动态参数辨识方法。首先，为减少由于传统二阶递推形式参数辨识方法存在的固有累积误差，引入基于解析解的二阶等效热参数模型。其次，考虑开窗通风、温度骤变等因素导致的多参数动态特性，以仿真系统与实际系统的动态误差最小为优化目标，兼顾室内初始固体温度，建立计及多参数动态特性的空调二阶ETP参数辨识模型。最后，为提高求解效率，提出了基于Chebyshev混沌映射的INRBO算法求解上述模型，算例结果表明，本文所建立的动态参数辨识模型能够兼顾拟合精度与求解效率。

Under the background of the new power system, the efficient aggregation and deployment of demand-side flexibility resources through virtual power plants is a beneficial measure to improve the flexibility of the new power system. Air-conditioning load is one of the most potential demand-side resources, and its accurate modeling can lay the foundation for the design of control strategy. Aiming at the problems of limited fitting accuracy, ignoring environmental factors and low efficiency of current air-conditioning system parameter identification, in order to further accurately quantify the scheduling potential of air-conditioning load in engineering applications, this paper proposes an improved Newton-Raphson-based optimizer ( INRBO ) based air-conditioning second-order equivalent thermal parameters ( ETP ) analytical solution dynamic parameter identification method. Firstly, in order to reduce the inherent cumulative error due to the traditional second-order recursive form parameter identification method, a second-order equivalent thermal parameter model based on analytical solution is introduced. Secondly, considering the multi-parameter dynamic characteristics caused by variable factors such as window ventilation, personnel flow and temperature sudden change, the second-order ETP parameter identification model of air conditioning considering multi-parameter dynamic characteristics is established with the minimum dynamic error between the simulation system and the actual system as the optimization goal, taking into account the initial solid temperature in the room. Finally, in order to improve the solution efficiency, an improved Newton-Raphson optimization algorithm based on Chebyshev chaotic mapping is proposed to solve the above model. The results of the example show that the dynamic parameter identification model established in this paper can take into account the fitting accuracy and solution efficiency.