<p>Gas turbine is a kind of dual-purpose rotating thermal machinery, widely used in power generation, shipbuilding and aviation power, etc., with the advantages of high efficiency, fast start and low emissions. In this study, an improved Circle-SCA-BSO algorithm (IC-SCA-BSO) is proposed to optimize PID parameters to ease the complicated parameter setting of gas turbine controller. Optimization of beetle swarm optimization algorithm (BSO) usually comes with slow convergence speed, low accuracy and prone to fall into local optimum, hereby our optimization is carried out from three aspects: population initialization, optimization weight and learning factors. First, a uniformly distributed circle mapping is utilized for population initialization. Second, the nonlinear decreasing idea is employed for weight optimization. Considering characteristics of global search in the early stage and local development in the later stage of algorithm optimization, the nonlinear decreasing function expression is designed. Third, combined with the sine cosine algorithm (SCA), the sine and cosine factors are introduced into the learning factors and combined with the nonlinear decreasing coefficient to make the learning factors show a trend of oscillatory attenuation in the set interval. According to probability <i>p</i>, sine or cosine factor is switched as the learning factors. The optimized PID controller and other four controllers are compared by tracking test and anti-interference test. The test results show that IC-SCA-BSO-PID yields faster response, higher steady-state accuracy and stronger anti-interference control effect, which is significantly outperforming the other four controllers. The proposed IC-SCA-BSO-PID tuning framework offers plant engineers an automated, low-cost alternative to labor-intensive manual calibration, enabling faster commissioning, reduced fuel consumption, and lower emissions for gas turbines in power-generation, marine, and aero-derivative applications.</p>