针对未考虑抗震设防的旧有房屋存在的抗震问题 ,在常用抗震鉴定方法的基础上 ,应用神经网络理论 ,提出了一种新的房屋两级抗震鉴定法。其中一级鉴定主要对房屋的抗震性能作出鉴定 ,二级鉴定以抗震验算为主、结合构造影响进行综合评价。鉴定实例表明 ,这种鉴定方法是行之有效的。 Aiming at the aseismic problems existing in the old buildings without considering the seismic fortification, a new method of seismic appraisal of two levels of buildings is put forward based on the commonly used methods of seismic appraisal and neural network theory. One of the appraisals mainly identifies the seismic performance of the houses, and the second appraisal is based on the checking of the earthquake resistance, and carries out comprehensive appraisal in combination with the influence of the structure. Identification example shows that this method of identification is effective.