Abstarct: In this thesis, an optimized method for algorithmic trading in the forex financial market is introduced, utilizing Bollinger Bands indicators and genetic algorithms and particle swarm optimization algorithms. Bollinger Bands indicator is widely recognized as a crucial tool in technical analysis. Many traders and researchers have utilized this indicator in their strategies; however, most of them have employed it for timing buy and sell orders, such as placing a sell order when the price reaches the upper band of the Bollinger Bands indicator and vice versa when the price hits the lower band. Alternatively, they use this indicator to confirm buy and sell signals, paying less attention to capital management and trading constraints. This paper presents a novel approach that employs Bollinger Bands indicators in a different manner for trading strategies. This approach statistically examines the average price movement above the upper band and below the lower band of Bollinger Bands when the price crosses these bands. This information is then utilized for timing buy and sell orders. To manage capital and determine optimal order sizes, standard deviation, as well as profit and loss limits in trades, genetic algorithms and particle swarm optimization algorithms are employed. This innovative approach utilizing Bollinger Bands has led to a trading system that has been tested for the initial ten months of 2023 on four currency pairs in the forex market and across four different time periods. Observing the performance of this trading system reveals a significant improvement in trade results compared to the baxseline system using Bollinger Bands. The profit growth from trades using the optimized genetic algorithm method has mostly been positive, with the lowest reported figure being 1.99% and the highest being 55.96%. For the particle swarm optimization algorithm, the lowest reported figure was 2.79%, and the highest was 336.83%. Furthermore, the crucial metric of maximum drawdown in both methods has also shown improvement. For the genetic algorithm, the minimum improvement in maximum drawdown was 6.51%, and the maximum was 71.44%. For the particle swarm optimization algorithm, the minimum improvement in maximum drawdown was 19.5%, and the maximum was 74.22%.