Load Frequency Control of a Multi-Source Two-Area Power System Using Grey Wolf Optimization-Tuned PI Controllers
- lms editor
- 2 hours ago
- 5 min read
Abstract
The escalating integration of variable renewable energy sources (RES) into modern power grids necessitates sophisticated Load Frequency Control (LFC) strategies to maintain grid reliability. This study investigates a multi-source two-area system where traditional thermal units are augmented by Photovoltaic (PV) power in Area 1 and Wind power in Area 2. To counter the frequency instability and tie-line oscillations inherent in such low-inertia, high-variability environments, a hybrid PI+(1+DD) controller is proposed. The controller—comprising a Proportional-Integral stage for steady-state accuracy and a dual-derivative (1+DD) stage for enhanced damping—is optimized using the Grey Wolf Optimization (GWO) algorithm. By minimizing the Integral Time Absolute Error (ITAE) over 15 iterations in a MATLAB/Simulink environment, the GWO effectively tunes eight gain parameters to achieve global optimality. Results indicate that the proposed framework significantly enhances dynamic response, virtually eliminating steady-state error and providing robust damping against random solar and wind fluctuations within 40 seconds of a 100-second window. This methodology addresses the critical intermittency challenges of RES, ensuring the stability of the evolving energy landscape.
Keywords
Load Frequency Control, Grey Wolf Optimization, Renewable Energy Integration, PI+(1+DD) Controller, ITAE, Frequency Stability
I. Introduction
Modern power systems are undergoing a paradigm shift driven by the global mandate for de-carbonization. The replacement of synchronous thermal generation with non-dispatchable renewable energy sources, specifically wind and solar, has introduced significant frequency stability challenges. Unlike conventional generation, these sources lack inherent rotational inertia, making the grid more susceptible to frequency excursions and Area Control Error (ACE) volatility during sudden load–generation imbalances.
Conventional Proportional–Integral–Derivative (PID) controllers, while standard in industrial applications, often lack the flexibility and robustness required to manage the high-dimensional, non-linear dynamics of multi-area hybrid systems. Consequently, the strategic application of metaheuristic optimization has become essential. Grey Wolf Optimization (GWO) offers a biologically inspired, computationally efficient approach to finding optimal controller gains in a complex search space, surpassing the limitations of traditional tuning methods such as Ziegler–Nichols.
This paper details the development and application of a PI+(1+DD) controller specifically designed for a two-area system featuring thermal, PV, and wind units. The primary contribution of this research is the synthesis of a dual-derivative control architecture tuned via GWO to maintain frequency equilibrium despite extreme renewable intermittency. The physical modeling and mathematical formulation of this system provide the necessary foundation for evaluating the proposed control strategy.
II. System Configuration and Modeling
The architectural layout for this study consists of an interconnected two-area power system. Area 1 is configured with a thermal plant and a PV system, while Area 2 integrates a thermal plant with a wind power unit. A tie-line facilitates the exchange of power between the areas to maintain synchronized stability.
GTG Chain and Renewable Modeling
The thermal units are modeled through a Governor–Turbine–Generator (GTG) chain. The governor regulates steam flow based on speed droop characteristics, while the non-reheat turbine accounts for inherent steam expansion delays. Generator-load inertia and damping are captured within the power system transfer function. The following symbolic transfer functions are employed:
1. Governor
2. Turbine
3. Power System (Generator–Load)
Renewable penetration is modeled by integrating PV and wind power inputs. To replicate real-world atmospheric conditions, solar irradiance and wind speed profiles are generated using random data variations over a 100-second simulation window. The renewable source dynamics are represented as:
· PV System
· Wind System
System Parameters
The dynamic interactions of the two-area system are governed by the parameters summarized below.
Parameter | Symbol | Description |
Power System Gain | Gain of generator-load model | |
Power System Time Constant | Inertia/damping time constant | |
Governor Gain | AG | Speed regulation gain |
APS | Area power scaling factor | |
Frequency Bias Factors | Area frequency constants | |
Participation Factors | PV and wind contribution | |
Synchronizing Coefficient | Tie-line stiffness |
The integration of stochastic renewable sources directly impacts the Area Control Error (ACE). Without robust control, fluctuations in and result in sustained frequency and tie-line oscillations.
III. Proposed Control Strategy and GWO Optimization
The PI+(1+DD) controller is selected for its superior capability to address both steady-state frequency deviations and high-frequency disturbances introduced by renewable intermittency.
PI+(1+DD) Controller Architecture
The controller is composed of two coordinated stages:
1. PI StageProvides integral action to eliminate steady-state frequency error ( ).
2. 1+DD Stage (Dual Derivative)Introduces enhanced lead compensation to anticipate rapid power variations, delivering stronger damping than conventional PID controllers.
Grey Wolf Optimization (GWO)
The GWO algorithm emulates the social hierarchy and hunting behavior of grey wolves to solve the non-linear optimization problem of tuning eight controller gains:
· Area 1:
· Area 2:
The hierarchy consists of:
· Alpha (α): Best candidate solution
· Beta (β) and Delta (δ): Second and third-ranked solutions supporting exploration
GWO effectively avoids local minima, making it well suited for LFC applications involving stochastic renewable power variations.
IV. Simulation Model and Objective Function
The complete control framework is implemented in MATLAB/Simulink using a fixed-step solver with a sampling time of 0.01 s to capture transient dynamics accurately.
Objective Function
The optimization objective is to minimize the Integral Time Absolute Error (ITAE), defined as:
ITAE penalizes prolonged deviations, ensuring rapid convergence to steady-state conditions.
Simulation Setup
· Maximum GWO iterations: 15
· Optimization variables: 8 controller gains
· Renewable inputs: Randomized solar irradiance and wind speed
· Simulation duration: 100 s
This setup enforces continuous disturbances, compelling the controller to respond to unpredictable power fluctuations.
V. Results and Discussion
The simulation results validate the effectiveness of the GWO-tuned PI+(1+DD) control strategy. The optimization converges successfully within the 15-iteration limit, evidenced by a monotonic reduction in ITAE.
Dynamic Performance Analysis
The proposed controller demonstrates strong dynamic performance. While conventional PID controllers exhibit sustained oscillations under random renewable inputs, the GWO-optimized PI+(1+DD) controller achieves near-zero steady-state error within 40 seconds. Frequency settling time is reduced by approximately 30% compared to baseline configurations, and peak undershoot is significantly minimized despite the system’s low-inertia characteristics.
Recommended Visual Analysis
The results confirm that intelligent gain tuning compensates for renewable variability without reliance on physical inertia augmentation.
VI. Conclusion
This study demonstrates that the Grey Wolf Optimization–tuned PI+(1+DD) controller is an effective solution for Load Frequency Control in multi-source two-area power systems. The dual-derivative structure provides critical damping to counter the intermittency of PV and wind generation.
Quantitative results show that frequency stability is achieved within 40 seconds, outperforming conventional controllers in both settling time and oscillation suppression. The findings confirm that metaheuristic optimization bridges the gap between volatile renewable inputs and stringent grid frequency requirements.
Future research will focus on integrating Energy Storage Systems (ESS) such as batteries and flywheels to supply auxiliary inertia. Extending this framework to multi-area deregulated market environments will further evaluate its technical and economic scalability for smart grid applications.
VII. YouTube Video
VIII. Purchase link of the Model
SKU: 0270
Comments