Optimizing Solar-Powered Hybrid Systems with Grasshopper Optimization Algorithm

By Samudrapom DamFeb 25 2024Reviewed by Susha Cheriyedath, M.Sc.

In an article recently published in the journal Scientific Reports, researchers proposed an energy management scheme (EMS) for a solar-powered battery-ultracapacitor (UC) hybrid system based on a rule-based grasshopper optimization algorithm (RB-GOA).

Importance of EMS

Harvesting energy from the sun has become popular compared to other renewable energy sources owing to low maintenance and operating costs. However, output fluctuations caused by ecological factors like partial shading conditions (PSCs), dust, clouds, temperature, and solar irradiance, and high response time increase the challenges to run instantaneous pulsed loads (PL) like electric vehicles (EVs), ships, electromagnetic recovery and launch systems, and high-power radars.

Photovoltaic (PV) energy can be integrated with high-power-density UCs and high-energy-density batteries to address the issue of output fluctuations. However, an EMS is crucial to regulate the energy flow among UCs, battery bank (BB), PV, and PL to realize the distinct benefits of every component in a BB-UC hybrid system powered by solar energy.

Limitations of EMS

EMS can be categorized into metaheuristic optimization approaches, intelligent control methods, and classical techniques. Classical techniques like filtration-based control and deadbeat control need an accurate system model for functioning efficiently and are sensitive to model parameter variations.

Although artificial intelligence (AI) methods such as machine learning and neural networks are effective for slow-dynamic applications when handling future load pattern uncertainties, the need for expert experience and knowledge for developing their rules is a significant challenge. Similarly, the effectiveness of optimization-based strategies depends on sophisticated computational setup and precise modeling.

Metaheuristic optimization techniques like particle swarm optimization (PSO) and white shark optimization (WSO) have been used in different EMS. Although the PSO technique has been more effective than other techniques, premature and slow convergence is a major drawback of this technique.

The proposed approach

In this study, researchers proposed an EMS based on an RB-GOA for a solar-powered battery-UC hybrid system. The objective of the study was to meet the PL demands efficiently and extract maximum energy from the PV array. The proposed approach initially establishes a simple IF-THEN set of rules for defining the search space, including PV, BB, and UC constraints. Then, GOA dynamically allocates the power shares among UC, BB, and PV to meet PL demand based on these IF-THEN rules and search space.

The presented RB-GOA-based EMS focused on two key aspects, including supporting PL demands and ensuring UC charging by BB and developing a maximum power point tracking (MPPT) scheme under PSC to exploit the maximum available PV energy. The first aspect was realized by creating two autonomous management layers, including a short-term power management layer (PML) and a long-term energy management layer (EML).

EML shrinks the search space by considering the components' operating limits. This layer activates different operational modes through the IF-THEN rules based on various scenarios. PML utilizes a GOA for establishing real-time optimal power sharing among UC, BB, and PV within the EML-defined search space.

RB-GOA generates reference signals and optimizes the objective function to control two bidirectional converters (BDCs) connected between the DC link and the BB and between the DC link and the UC. Additionally, the second aspect was achieved by implementing a GOA that regulates the duty ratio of a boost converter connected to the common DC link.

Evaluation and findings

Researchers comprehensively evaluated and compared the proposed technique’s performance with other popular swarm intelligence techniques (SITs), including the salp swarm algorithm (SSA), gray wolf optimization (GWO), and cuckoo search algorithm (CSA). The evaluation was performed for different cases, including PV with MPPT, variable PV with PL cases, variable PV with a constant load, and PV without any energy storage device.

The comparative analysis demonstrated that the proposed technique outperformed other SITs based on MPPT speed, oscillation mitigation, and reduction in power surge during load transition. Specifically, the technique reduced the power surge by 8%, 22%, and 26% compared to SSA, GWO, and CSA, respectively, for the variable PV with constant load case.

Additionally, the technique mitigated oscillations over three times as fast as SSA and twice as fast as GWO and CSA. The technique also reduced the power surge by six times compared to SSA and by nine times compared to GWO and CSA in variable PV with the PL case. Moreover, the proposed technique's MPPT speed was almost 29% to 61% faster compared to its alternatives irrespective of weather conditions.

To summarize, the findings of this study demonstrated that the proposed EMS is more effective than other SITs in minimizing oscillations while maximizing PV energy usage, reducing power surges, and maintaining a stable output across PL demand.