---

The main aim of the present study is to propose a modified harmony search (MHS) algorithm for size and shape optimization of structures. The standard harmony search (HS) algorithm is conceptualized using the musical process of searching for a perfect state of the harmony. It uses a stochastic random search instead of a gradient search. The proposed MHS algorithm is designed based on elitism. In fact the MHS is a multi-staged version of the HS and in each stage a new harmony memory is created using the information of the previous stages. Numerical results reveal that the proposed algorithm is a powerful optimization technique with improved exploitation characteristics compared with the standard HS.

---

Many optimization techniques have been proposed since the inception of engineering optimization in 1960s. Traditional mathematical modeling-based approaches are incompetent to solve the engineering optimization problems, as these problems have complex system that involves large number of design variables as well as equality or inequality constraints. In order to overcome the various difficulties encountered in obtaining the solution of these problems, new techniques called metaheuristic algorithms are suggested. These techniques are numerical optimization algorithms that are based on a natural phenomenon. In this study, a state-of-art improved harmony search method with a new adaptive error strategy is proposed to handle the design constraints. Number of numerical examples is presented to demonstrate the efficiency of the proposed algorithm in solving engineering optimization problems.

---

Vulnerability assessment of structures encounter many uncertainties like seismic excitations intensity and response of structures. The most common approach adopted to deal with these uncertainties is vulnerability assessment through fragility functions. Fragility functions exhibit the probability of exceeding a state namely performance-level as a function of seismic intensity. A common approach is finding some response points of the fragility function and then fitting a typical probability distribution like lognormal through curve fitting estimation techniques. Maximum-likelihood approach is a fitting method to find the probability distribution parameters. Performing this approach for distributions like lognormal which is defined by just two parameters are straight forward while for more complicated distribution which are based on additional characterizing parameters is not feasible, since this approach is based on minimizing an error function through classic mathematical approaches like calculating partial derivations. An applicable modification is to add an efficient optimization approach to determine maximum-likelihood function. In this article, an optimization algorithm is proposed with maximum-likelihood-estimation and the results indicate the efficiency and feasibility of future developments in finding the most appropriate fragility function.