Modelling of latent heat thermal storage systems is mainly achieved with an analytical formulation describing the thermodynamic state of the phase change material (PCM). In literature, different models exist to describe the evolution of the liquid fraction, the effective heat capacity and the enthalpy as a function of temperature. The main issue is to correctly establish these behaviour laws and to identify the thermophysical properties of each PCM. Analytical models proposed in literature often have obvious limitations, such as the presence of discontinuities or the impossibility to represent asymmetric phase change dynamics. Such approaches a priori do not allow to model accurately the thermodynamic behaviour of the PCM. Given the limitations identified on existing models, a new analytical model is proposed in this paper. Moreover, it is not uncommon to observe the presence of two heat flux peaks during the phase change process of a PCM, as with the studied PCM (RT58). Therefore, we propose a method to model the two heat flux peaks, which are rarely considered in literature. PCM properties are determined thanks to a four-step identification process, using experimental data and the different analytical models studied. Based on the numerical modelling of the rectangular container used for the experiments, and on the identified thermodynamic properties of the enclosed PCM, a comparative study between the new analytical model and existing models is carried out. It appears that modelling only a single heat flux peak is not satisfactory for the studied PCM, and taking into account a second heat flux peak decreases the error on the exchanged energy by a factor of 2.5 to 6.5 depending on models. The new model leads to an accuracy gain of a factor of 1.5 to 3.3 compared with existing models. We do not recommend the model based on a linear evolution of the liquid fraction as the resulting error is between 1.6 and 3.3 times greater than for the other models. The study carried out is based on both melting and solidification processes, and concerns a PCM which does not undergo supercooling.

潜热蓄热系统的建模主要通过描述相变材料（PCM）的热力学状态的分析公式来实现。在文献中，存在不同的模型来描述液体分数，有效热容量和焓随温度的变化。主要问题是正确建立这些行为定律并确定每个PCM的热物理性质。文献中提出的分析模型通常具有明显的局限性，例如存在不连续性或无法表示不对称相变动力学。这种方法先验地不允许精确地建模PCM的热力学行为。考虑到现有模型中存在的局限性，本文提出了一种新的分析模型。此外，像研究的PCM（RT58）一样，在PCM相变过程中观察到两个热通量峰值并不罕见。因此，我们提出了一种对两个热通量峰值建模的方法，这在文献中很少考虑。通过使用实验数据和研究的不同分析模型，通过四步识别过程确定了PCM属性。基于用于实验的矩形容器的数值模型，并基于所识别的封闭式PCM的热力学性质，对新的分析模型与现有模型进行了比较研究。似乎仅对单个热通量峰进行建模对于所研究的PCM是不令人满意的，并且考虑到第二热通量峰将交换能量的误差降低了2.5到6倍。5取决于型号。与现有模型相比，新模型可将精度提高1.5到3.3倍。我们不建议基于液体分数的线性演化的模型，因为由此产生的误差是其他模型的1.6到3.3倍。进行的研究是基于熔化和凝固过程，并且涉及不进行过冷的PCM。