top of page
Writer's pictureLMS RS

MATLAB Implementation of GA Tuned PI Controller for a Two-Area Power System

Introduction to Genetic Algorithm and PI Controller

Genetic Algorithm (GA) is a powerful optimization technique inspired by natural selection and evolution. In the context of control systems, GA is used to optimize the parameters of controllers like Proportional-Integral (PI) controllers. The PI controller is commonly used to regulate various processes, and its tuning plays a critical role in achieving desired performance in dynamic systems.

In this example, we focus on tuning a PI controller for a two-area power system. GA helps us find the optimal values of the proportional (Kp) and integral (Ki) gains, improving the stability and response of the system.

Overview of the Two-Area Power System

The two-area power system consists of two interconnected areas that exchange power based on demand. Each area includes a governor, turbine, and generator load model. These components are essential for regulating the frequency and maintaining the power balance in the system.

In the simulation, each area is connected via a transmission line (Ti-line), which allows for power transfer between the areas. The system behavior is studied with and without controllers to evaluate the effects of tuning.

System Behavior Without a Controller

Before implementing any control strategies, the system is analyzed in its uncontrolled state. In this initial scenario, the response of the system is observed by looking at parameters like the change in angular frequency (delta omega) and power transfer between the areas.

Without any controller, the system exhibits unregulated behavior, with both areas experiencing similar frequency deviations and power transfer. The performance is far from ideal, which emphasizes the need for a well-tuned controller to maintain stability.

Implementing the Integral Controller

To improve the system's behavior, an integral controller is introduced. The integral controller is designed to minimize steady-state error by adjusting the control signal based on accumulated error over time.

For the simulation, a gain of 0.3 is used for the integral controller in both areas. The results from this implementation show a slight improvement in system performance, but it still does not reach the desired level of stability and responsiveness.

Tuning the PI Controller Using Genetic Algorithm

Next, we move on to a more sophisticated approach by tuning the PI controller using GA. In this case, the genetic algorithm optimizes the proportional (Kp) and integral (Ki) gains to achieve the best possible system performance.

The GA operates through several iterations, adjusting the controller parameters at each step. With a population size of 10 and 10 iterations, the GA continuously evaluates the fitness of different controller parameter combinations using an objective function. The fitness function measures the system's performance based on criteria such as Integral of Squared Error (ISE), Integral of Time-weighted Squared Error (ITSE), and others.

Throughout the iterations, the fitness function improves, with the parameters of the controller being adjusted dynamically to minimize the error. For instance, after a few iterations, the fitness function value improved significantly, leading to more optimal controller settings.

Results After GA Tuning

After completing the GA optimization process, the PI controller's tuned parameters (Kp and Ki) are used to simulate the two-area power system. The results show a marked improvement in system performance. The oscillations in the frequency deviation (delta omega) are significantly reduced, and the settling time for the system response improves.

The optimized controller ensures that both areas achieve better synchronization, with smoother power transfer between the areas. This demonstrates the effectiveness of using Genetic Algorithms to fine-tune controller parameters, resulting in a more stable and efficient system.

Conclusion

In this blog post, we demonstrated how to use a Genetic Algorithm to tune the parameters of a PI controller for a two-area power system. By leveraging GA’s optimization capabilities, we were able to achieve improved system stability and response, compared to traditional methods like using an integral controller alone.

GA provides an effective and powerful tool for controller tuning in complex systems, ensuring optimal performance and reliability. This technique can be applied to various other control systems, making it an invaluable method in the field of power system automation.

0 views0 comments

Comentarios


bottom of page