Distance optimization and directional overcurrent relay coordination using edge-powered biogeography-genetic algorithms

Abstract

The effective functioning and regulation of power systems crucially rely on the coordination of distance and directional overcurrent relays. Accurate fault detection and successful clearing sequences require support for each relay and the maintenance of the coordination time interval (CTI) between major distance relays, directional overcurrent relay support, and other relay zones. Efficiently initiating relays while adhering to complex coordination limitations poses a challenging task that demands innovative solutions. This study addresses the intricate problem of relay coordination by employing heuristic methods, specifically genetic algorithms (GA) and biogeography-based optimization (BBO), in both a 9-bus and 39-bus system. The primary objective is to determine the most efficient time setting factor (TSM) that minimizes the duration of relay operation. Additionally, the intelligent features of the overcurrent relay are carefully chosen to enhance the research's results. The integration of edge computing capabilities plays a significant role in advancing this coordination method. By incorporating advanced algorithms and communication technologies at the edge, the prompt activation of relays becomes possible, thereby meeting coordination demands. This study explores the combination of edge-based servers with genetic algorithms (GA) and biogeography-based optimization (BBO) techniques to enhance relay coordination. The findings indicate a notable enhancement compared to conventional approaches. However, comparative research suggests that BBO's performance is similar to GA, without a distinct advantage in achieving higher outcomes.

Introduction

Electric power systems are currently undergoing continuous evolution, characterized by the proliferation of distributed generation units and the expansion of transmission systems [1, 2]. This evolution is driven by the need to meet escalating load demands [3, 4]. However, this ongoing transformation brings about increased complexity, leading to challenges, particularly in terms of potential errors that could result in substantial short-circuit currents, thereby posing a significant threat to the integrity of the overall system.

The integration of distributed generation sources, while advantageous in addressing energy needs, further heightens security concerns [5]. This includes the potential risk of coordination loss among network security devices. To ensure the safeguarding of these intricate power networks, distance and directional current relays are paramount due to their heightened sensitivity [6]. Nevertheless, ensuring the optimal functionality of the protection system necessitates precise coordination between these relays.

The achievement of synchronization between distance relays and directional overcurrent relays is pivotal for intelligent protection systems, requiring a meticulous and strategic approach [7, 8]. The strategies employed should not only guarantee swift responses to faults but also demonstrate adaptability to dynamic changes inherent in modern power grids. Establishing the desired coordination level entails defining stringent criteria, including the minimum allowable time interval between distance and directional current relays.

Research has underscored the paramount importance of energy optimization and enhancing power system efficiency [9]. There is a substantial emphasis on optimization strategies and the utilization of cutting-edge algorithms to elevate power system performance. Certain studies have directed their attention towards the advancement and deployment of sophisticated centralized control systems. It has been posited that the implementation of an advanced control system tailored for environmental settings holds the potential to enhance environmental quality [10].

Traditional methods, ill-equipped to address contemporary power systems' intricacies, underscore the imperative need for advanced technologies. The integration of edge-based servers, especially artificial intelligence (AI), plays a crucial role in optimizing these processes [11]. The use of edge computing technology presents an enhancement for the automatic adjustment of power system relays, and employing dual-unit digital technology further improves this process [12]. One of the solutions for improving the automatic adjustment of power system relays is the implementation of edge computing technology, which enhances system performance through dual-unit digital methods. Moreover, intelligent systems, empowered by AI, can dynamically adjust and optimize protection relay operation time and characteristics. This, in turn, reduces relay operation time and enhances coordination between primary and backup systems. The amalgamation of edge-based server technologies with AI-driven strategies promises adaptive and efficient protection in modern power systems.

One potential challenge in implementing advanced relay coordination strategies using edge processing is the limited computational power and storage of edge devices [13]. Edge-based servers facilitate real-time data processing and decision-making at the local level, but heavy computational requirements may overwhelm edge devices, leading to slower response times or potential data loss [14, 15]. Additionally, the network infrastructure may not be sufficient to handle the large volume of data generated by these advanced strategies, leading to degraded performance or connectivity issues. In addition, traditional relay coordination strategies often rely on trial-and-error methods or manual adjustments. These methods can be time-consuming, costly, and may not ensure optimal performance. They may also struggle to handle complex power grids with multiple relays and varying operating conditions.

The Biogeography Based Optimization (BBO) algorithm is inspired by the distribution of species in nature, where organisms are distributed based on their ability to adapt to their environment. In BBO, individuals (also known as particles) that represent possible solutions are distributed in a geography (a search space) based on their fitness values, and the algorithm evolves through iterations (migrations) to find the optimal solution. BBO exhibits efficient behavior in finding the global optimum and is computationally efficient, making it suitable for complex problems.

Additionally, edge processing in relay coordination strategies allows for real-time monitoring and analysis of data, leading to faster and more accurate decision-making. It also helps reduce the burden on the centralized computing system by processing and analyzing data closer to the source. This results in improved efficiency and reduced bandwidth requirements. Additionally, edge processing enables seamless connectivity and data privacy, as it allows for local data storage and transmission, eliminating the need for sensitive data to travel long distances.

The integration of edge computing and industrial Internet of Things (IIoT) capabilities holds significant promise for a wide range of industrial applications, including smart grid systems [16]. In the context of smart grids, the deployment of edge computing infrastructure enables real-time data processing and decision-making at the network's edge, closer to the data source. This proximity not only reduces latency and enhances responsiveness but also facilitates the implementation of intelligent relay coordination strategies. By leveraging edge-cloud capabilities, smart grid operators can optimize relay activation and coordination, ensuring timely fault detection and efficient management of power distribution networks.

Furthermore, the convergence of IIoT and edge computing technologies offers unprecedented opportunities for enhancing the resilience and reliability of smart grid systems. With the proliferation of distributed energy resources (DERs) and the increasing complexity of grid operations, there is a growing need for adaptive and scalable solutions to address emerging challenges. Edge-cloud architectures provide a scalable and flexible framework for deploying advanced monitoring, control, and optimization algorithms in smart grid environments. By harnessing the power of IIoT-enabled edge devices, smart grid operators can optimize relay coordination strategies, improve system efficiency, and enhance overall grid performance in the face of evolving operational demands and dynamic grid conditions.

In power systems, cutting-edge technologies play a crucial role in enhancing the reliability and efficiency of relay coordination strategies. One notable advancement is the development of an advanced coordination approach for Distance and Directional Overcurrent Relays. By harnessing the power of Biogeography-Based Optimization (BBO) and Genetic Algorithms (GA), in conjunction with the efficiency of edge processing, this approach offers a novel solution to the challenges encountered by traditional relay coordination methods. Several key research contributions are highlighted:

• Both BBO and GA synergistically contribute to relay coordination. BBO's capacity to efficiently explore the solution space, inspired by natural distribution patterns, is complemented by GA's evolutionary principles. This synergy creates a robust optimization framework for determining relay settings.

• Edge-based servers address conventional relay coordination limitations. By facilitating real-time monitoring and analysis at the local level, this approach ensures swift decision-making, thereby reducing response times and minimizing the risk of data loss, even in scenarios with limited computational resources.

• The proposed strategy is specifically designed to overcome complex challenges encountered in modern power grids. Traditional methods often struggle to cope with multiple relays and varying operating conditions. The utilization of BBO and GA techniques, augmented by edge processing capabilities, demonstrates adaptability to the dynamic nature of modern power systems.

• The combination of BBO and GA introduces an efficient relay coordination approach capable of seeking the global optimum. This not only enhances the overall performance of the power system but also addresses the time-consuming and costly aspects associated with traditional trial-and-error methods.

• The integration of edge-based servers contributes to resource optimization. By processing and analyzing data closer to the source, the strategy alleviates the burden on centralized computing systems. This optimization enhances efficiency and reduces bandwidth requirements, ensuring seamless connectivity and preserving data privacy.

According to the research background, the coordination process of distance relays and directional overcurrent was investigated using the theory of BBO and GA. This study aims to improve protection systems capabilities in modern networks. It does this by analyzing existing methods and providing advanced strategies so that power networks have the necessary efficiency and stability in any situation.

In this paper, we propose a systematic approach to address relay coordination challenges in power systems, focusing on distance and directional overcurrent relays. We begin with a review of related work (Section "Related work") and define the coordination problem (Section "Coordination problem of distance and directional overcurrent relays"). Our proposed methodology, using biogeography-based optimization (BBO) and genetic algorithms (GA), is detailed in Section "Simulation results". We present simulation results for 9- and 39-bus power systems (Section "Conclusion") and discuss findings and implications in Sect. 6. Finally, in Sect. 7, we offer concluding remarks and suggest future research directions.

Related work

The specialized literature extensively explores various methodologies for addressing challenges related to coordinating directional current and distance relays. These approaches leverage multiple variables, including the characterization of directional current relays, the operational time of the second zone in distance relays, the count of critical error points, and methodologies for designing objective functions.

In a study by Perez et al. [17], a linear programming method was introduced to seamlessly integrate fixed-range configurations of distance relays with directional current backup relays. This involved a well-defined operating time and targeted only critical fault locations. The programming aimed to provide flexibility by concurrently determining the time-dial settings of directional current relays, influenced by both the second zone settings and key failure considerations.

Khederzadeh et al. [18] examined distance relays and forward current relays. However, the return characteristics of directional current relays were adaptively adjusted based on the position and intensity of the fault current, rather than relying solely on the temporal conversion of the second zone. Chabanloo et al. [19] proposed a novel approach employing a Genetic Algorithm (GA) to resolve coordination challenges between forward current relays and distance relays. The enhanced target function coordinated between directional current relays and distance relays. Different relay characteristics were considered decision variables in the optimization problem for each forward current relay.

In another work by Chabanloo et al. [20], diverse reverse time characteristics were selected for forward current relays to enhance coordination between relays and distance relays, considering five critical fault scenarios. Moravej et al. [21] presented a systematic approach to optimizing the settings of the second zone in both distance and directional current relays within a series of compensated and uncompensated transmission systems. The modified adaptive particle optimization (MAPSO) method successfully addressed this nonlinear and non-convex coordination problem. Farzinfar et al. [22] proposed a novel strategy grounded in the coordination of distance relays in the presence of directional current relays as backup relays, utilizing the recently resolved multi-order embedded particle optimization algorithm (MECPSO).

Damchi et al. [23] introduced a strategy aimed at reducing the number of violations of coordination constraints in the coordination problem of directional current relays and distance relays. Singh et al. [24] aligned the settings of the second zone in distance relays with the standard time characteristics of directional current relays, accounting for diverse network topologies. Damchi et al. [25] proposed innovative operational features for the second zone of distance relays, dividing the method into two segments: the first section protecting the main line beyond the first zone with the first-time delay, and the second section safeguarding half of the adjacent lines with the second time delay. Based on the proposed operating characteristic (OPC), an innovative formulation was presented for the nonlinear coordination problem between distance and directional current relays.

Additionally, Ahmadi et al. [26] suggested measures for reducing violations of coordination limits between directional current relays and distance relays. In contrast, Yazdaninejadi et al. [27] proposed enhanced coordination between distance and directional current relays utilizing the dual time-dependent characteristics of directional current relays. This program aids in eliminating inconsistencies by simplifying the optimization domain and providing greater flexibility and scalability during the conservation coordination process, emphasizing the integration of edge computing for enhanced efficiency.

The impact of a high penetration of Renewable Energy Sources (RES) on power flow and short circuit levels necessitates reassessing and possibly updating the protection approach used in power systems relying on conventional generators. Specifically, [28,29,30,31,32] focused on modifying power system protection measures, analyzing various options for incorporating renewable energy sources (RES) while maintaining power system safety. Different types of relays were used, including distance, differential, adaptive, and voltage-based protection, as well as directional overcurrent relays. A fault current limiter was suggested in [28], and the role of RES in causing faults was discussed in [31].

The authors of [30] designed and simulated a DC distribution network using PSCAD for distance protection (Device 21). Furthermore, [33] enhanced distance protection settings (Device 21) using a fuzzy optimization method proportional to the proportion of RES in each feeder. In [33], the impact assessment and necessary updates for directional relays (Device 67) were also included. This investigation focused primarily on the settings for overcurrent protection (Device 51/50), as detailed in references [34,35,36]. The Lagrange generalized with Karush–Kuhn–Tucker optimization method was used in [34] to determine the optimal time delay setting for overcurrent protection. Based on temperature constraints for electrical equipment, the maximum allowable time delay was determined using only an IEC time delay curve with a minimum of 0.3 s.

An interior point optimization strategy was used to examine the time lag for microgrid overcurrent protection. IEEE standard curves were used for the time delay curve in the study. In a low-voltage network, the chosen time delay ranged between 0.2 and 0.3 s. Following multiple trips caused by variations in wind turbine output power, [36] used the Electrical Transient Analyzer Program (ETAP) to reevaluate the time delay for an overcurrent relay.

An attempt was made in [37] to modify the existing relay configuration in a radial system with distributed renewable energy sources. By replacing fuses with multi-function relays, the fixed curve issue associated with fuses was resolved. Additionally, [38] introduced an adaptive method for overcurrent and distance relays. By taking the equivalent circuit of the network into account, this method calculates the short circuit contribution, and the optimization process is iterative. Based on the conventional IEC curve, an optimal overcurrent relay design was found using a root tree technique [39]. As a result of time delay setting limitations, relay coordination could not be completed within the minimum allowable time.

In [40], the authors used a harmony search approach to compare their efficiency, while in [41], they used a fuzzy matching approach to compare relay coordination. The authors of [42] recommended adjusting the directional overcurrent relay to improve communication between the main and backup relays. In [43], a chaotic cuckoo search method was used to improve directional overcurrent relay settings.

A literature review indicates that the ultimate minimal time delay for an overcurrent relay has not been adequately considered. A key component of overcurrent protection is the steady-state overcurrent pick-up (Device 51), instantaneous overcurrent (Device 50), and overcurrent time delay (Device 51). To determine the optimal time delay setting, this work integrates the IEC and IEEE time–current curves (TCC). In this technique, time delay settings are adhered to base on electrical equipment withstand ratings, relay maximum time delay settings are restricted based on vendor design, and time delay settings are restricted based on IEEE Buff Book and temperature limits for electrical equipment/devices.

There are limitations and disadvantages to using artificial intelligence to coordinate distance and directional overcurrent relays, including a lack of accuracy in detecting and responding to electrical faults, a difficulty adapting to changing operating conditions, and potential security threats. The need for specialized training and expertise could hinder widespread adoption of the strategy. In addition, the strategy's reliance on AI may not be suitable for all scenarios, as AI technologies can be biased and may not always interpret real-time data accurately. As a result, these limitations and disadvantages may have a significant impact on the overall effectiveness of the strategy. System performance could be compromised, resulting in potential safety hazards, financial losses, and financial losses. Additionally, edge processing allows real-time data processing and analysis close to the source, reducing latency and improving decision-making. Additionally, advanced monitoring and control capabilities enhance system efficiency, optimize relay coordination, and improve overall system performance.

Coordination problem of distance and directional overcurrent relays

Despite modern distance relays, directional overcurrent relays are the most suitable choice for local support of distance relays from an economical and technical point of view. However, they make coordinating these relays more complicated. Three stages of coordination of overcurrent-to-overcurrent relays, distance to distance, and distance to overcurrent relays are performed for optimal coordination [44].

Coordination of overcurrent–overcurrent

The problem of optimal coordination of overcurrent relays in the power network can be expressed as an optimization problem with linear or non-linear variables defined as follows [23].

$$OF=\text{min}{\sum }_{i=1}^{n} {T}_{i}$$

(1)

$$\text{Subject to: }{T}_{b}-{T}_{m}\ge CTI$$

(2)

$${T}_{i}=f\left(TMS,{I}_{pi},{I}_{sc}\right)$$

(3)

$$TM{S}_{\text{min}}\le TMS\le TM{S}_{\text{max}}$$

(4)

$${I}_{pi(\text{min})}\le {I}_{pi}\le {I}_{pi(\text{max})}$$

(5)

where, n shows the number of relays and T _i presents the operating time of the named relay. This is a function of the TMS and I _pi setting parameters and the short circuit current passing through the relay. In addition, T _b is the operation time of the backup relay, and T _m is the operation time of the primary relay at the critical points of the error. Equation (2) requires coordinating the two main and backup relays for the error at two critical points at the beginning and end of the main line. CTI is the coordination time interval, which is usually between 0.2 and 0.5 s. Solving the above problem is about minimizing the objective function. This is so that the coordination constraints for all pairs of primary and backup relays and the constraints of relation setting parameters (4 and 5) are satisfied for all overcurrent paths.

Distance-distance coordination

The purpose of coordinating the distance ways is to determine the impedances of the three zones and the operation time of each zone. These areas should cover a greater percentage of the front line of the relay and the adjacent lines under their protective cover, with the condition that there is no interference in the operation of the main and backup relay areas. Distance relays should be coordinated in all lines before starting for optimal coordination of distance relays and directional overcurrent relays. In this article, the settings of distance relays have been done using the methods presented in references [45,46,47,48] for the sample network. The third zone adjustment impedance in reference [45] and the second zone adjustment impedance of distance relays in study [48] were calculated by considering the uncertainty in the input and output network structure and system operating conditions.

Distance-directional overcurrent coordination

When distance and overcurrent relays are combined in the protective procedure, the coordination constraints (6) and (7) are shown at the critical points F4 at the beginning of the second zone of the main distance relay and F5 at the end of the backup distance relay, which is added to the coordination problem of overcurrent relays in Fig. 1 [46, 47].

Fig. 1

Coordination of distance and overcurrent relays [46, 47]

$$T_{\text{z}2(\text{backup})}-T_{oc(\text{main})}\geq CTI'$$

(6)

$$T_{oc(\text{backup})}-T_{\text{z}2(\text{main})}\geq CTI'$$

(7)

${T}_{z2(\text{ main })}$ and ${T}_{\text{z}2(\text{ backup })}$ in the above equations of the operation of the second zone of the main and backup distance relays, are the main and backup directional overcurrent relays of${T}_{oc(\text{ backup })},{T}_{\text{oc( main )}}$, respectively.

The number of coordination constraints is three times more than that of the overcurrent relay coordination problem, considering the simultaneous coordination of overcurrent and distance relays. Thus, a new coordination time interval (CTI) between the distance and overcurrent relays must be defined, which can be considered equal to the CTI used in coordinating the pair of overcurrent relays [21]. An instantaneous operation unit and an overcurrent relay have been used to support distance protection. The instantaneous operation unit is used in cases with the overcurrent relay when it is necessary to interrupt the faults near the instantaneous overcurrent relay. Instantaneous operation units do not have a time setting and are only set with a certain current, as soon as the current exceeds this current, the relay operates instantaneously. The setting criteria of these relays depends on the location and type of system elements under control. The instantaneous performance unit settings for lines between substations, distribution lines, and transformers are as follows:

• A) Lines between substations: Instantaneous operation relay settings are made for 125% of the symmetrical effective current of the maximum fault of the next substation.

• B) Distribution lines: One of the two values of 50% of the maximum short circuit current at the CT connection point of the relay and between 6 and 10 times the rated load of the line can be used for adjustment.

• C) Transformers: between 125 and 150% of the short circuit current in the low-pressure load bus that is transferred to the high-pressure side is adjusted for the high-pressure side relay.

The instantaneous operation unit is set with a current in the proposed method that instantaneously trips up to 70% of the line in front of the line. The distance and directional overcurrent are necessary to determine the critical points of the line in the coordination problem to determine the coordination constraints. The critical point is a point on the line that if a fault occurs at this point, the time interval between the main and backup relays' operation time is minimal, which is shown for the proposed method in Fig. 2 [49].

Fig. 2

Critical points in the coordination of distance and directional overcurrent relays [49]

As shown in Fig. 2, it is assumed that overcurrent and distance relays are installed on all buses. Points F1 and F3 are the beginning and the end of the main line, respectively. F2 is the limited protection end of the instantaneous function unit. F is the end of the first zone of the main distance relay. In Fig. 2, point F4 is the end of the second zone of the backup distance relay, which has been chosen as a critical point in most references, and is the end of the second zone of the backup distance relay with an operation time of 0.3 s.

In the operation zone of the instantaneous operation unit of the main high current relay, the operation time is assumed to be about 0.02 s. Therefore, the time interval at this point is approximately equal to 0.3. There is no need to check the performance interval and enter it into the optimization problem. In this article, the constraints of the coordination problem of distance and directional overcurrent are defined as follows, according to Fig. 2.

$${t}_{b}\left({F}_{1}\right)-0.02\ge CTI$$

(8)

$${t}_{b}\left({F}_{2}\right)-{t}_{m}\left({F}_{2}\right)\ge CTI$$

(9)

$${t}_{b}\left({F}_{3}\right)-{t}_{m}\left({F}_{2}\right)\ge CTI$$

(10)

$$t_b\left(F_4\right)-0.3\geq CTI'$$

(11)

where, t_m(F_i) and t_b(F_i) are the operation time of the main and backup overcurrent relays per fault at the critical point Fi(i = 1,2,3,4). The number 0.02 in Eq. (8) is related to the operation time of the instantaneous operation unit of the main overcurrent relay. In Eq. (11), the number 0.3 indicates the operation time of the second zone of the main distance relay. CTI and CTI values are assumed equal to 0.3. The objective function of the optimal coordination problem is formulated as Eq. (14) with the sum of Eqs. (12) and (13).

$$O{F}_{Dac-Doc}={\alpha }_{1}{\sum }_{i}^{n} {t}_{i}{ }^{2}+{\alpha }_{2}{\sum }_{i=1}^{3} {\sum }_{{k}_{i}=1}^{{P}_{i}} \left[\Delta {t}_{mb}\left({F}_{i},{k}_{1}\right)\right.$$

(12)

$${\left.-\beta \left(\Delta {t}_{mb}\left({F}_{i},{k}_{1}\right)-\left|\Delta {t}_{mb}\left({F}_{i},{k}_{1}\right)\right|\right)\right]}^{2}$$

$$O{F}_{Di\text{s}-D\text{oc}}={\alpha }_{3}{\sum }_{{k}_{1}=1}^{{P}_{2}} \left[\Delta {t}_{\text{mbDISDOC }}\left({F}_{4},{k}_{2}\right)\right.$$

(13)

$${\left.-\beta \left(\Delta {t}_{\text{mbDISDOC }}\left({F}_{4},{k}_{2}\right)-\left|\Delta {t}_{\text{mbDISDOC }}\left({F}_{4},{k}_{2}\right)\right|\right)\right]}^{2}$$

$$OF=O{F}_{Doc-Doc}+O{F}_{Dis-Doc}$$

(14)

where, α₃, α₂, α₁, and β are the weighting coefficients of the objective function P1, the number of pairs of main and backup overcurrent relays. P2 is the number of pairs of main distance and backup overcurrent relays. t_i is the operation time of the i-th main relay. Moreover, $\Delta {t}_{mb}\left({F}_{i},{k}_{1}\right)$,$\Delta {t}_{\text{mbDIsDOC }}\left({F}_{4},{k}_{2}\right)$ for the above objective function is defined as follows:

$$\Delta t_{mb}\left(F_i,k\right)=t_b\left(F_i,k\right)-t_m\left(F_i,k\right)-CTI$$

(15)

$$\Delta t_{\text{mbDISDOC}}\left(F_4,k\right)=t_b\left(F_4,k\right)-t_{z2}-CTI'$$

(16)

where, t_m(F_i,k) and t_b(F_i,k) are, respectively, the operation time of the main relay and the backup relay pair for a symmetrical three-phase fault at the critical points Fi (i = 1,2,3,4) and tz2 is the operation time of the second zone of the distance relay. In this article, overcurrent relays are constant, and operation time is set to 0.3 s per fault current. The operating time of the overcurrent relays for the fault current I_sc is calculated from the following common equation:

$$t=\text{TMS}\left(\frac{k}{{M}^{a}}+L\right) M=\frac{{I}_{sc}}{{I}_{pi}}$$

(17)

where, M is the ratio of short circuit current to regulation current. L, α, and k are numerical values in Table 1 for different characteristics of overcurrent relays. In this problem, TMS is considered a variable with a constant regulation current and 1/3 times the load current. In the power system, digital overcurrent relays have various characteristics [24]. In this method, the characteristics of the relays are included as variables in the optimal problem, and the algorithm is designed to select the best characteristic for optimal coordination.

Table 1 Specifications of different overcurrent relays [25]

In Table 1, factor L, α, and K, respectively, indicate the setting time, time setting factor, and resistance setting factor, which are used to fine-tune the performance of protective relays against overcurrent’s and gaps. These settings are based on various standards such as AREVA, IEC, and ANSI/IEEE, and each has different reverse characteristics designed to meet specific power system conditions.

The coordination of distance and overcurrent relays is non-linear, considering TMS and overcurrent relays' characteristics as problem variables, which requires intelligent optimization algorithms to find the optimal answer. In this article, biogeography-based optimization algorithm (BBO) and genetic algorithm (GA) are used to solve the coordination optimization problem of distance and overcurrent relays.

Proposed method

BBO and GA, as tools for optimizing the coordination of distance and overcurrent relays, can be used to set these parameters for optimal coordination between relays. These algorithms seek to find the configuration that provides the best performance by presenting multiple solutions and evaluating them against certain criteria. Moreover, in our study, we have indeed integrated edge computing capabilities into the relay coordination process to enable real-time relay activation. By deploying edge-based servers at strategic locations within the power grid infrastructure, we can perform data processing and decision-making tasks closer to the data source, minimizing latency and enhancing responsiveness. The integration of edge computing technology facilitates swift relay activation by reducing the communication overhead and processing delays associated with centralized computing systems. This approach enhances the overall efficiency and reliability of the relay coordination process, particularly in dynamic and time-critical situations. By utilizing edge computing for real-time relay activation, we can achieve faster fault detection, more accurate decision-making, and improved system resilience. This not only enhances the reliability of power system operations but also enables adaptive and proactive responses to evolving grid conditions.

Using BBO to coordinate relays

BBO is an optimization algorithm that uses biogeographic concepts to find optimal solutions. Each island represents a possible TMS setting for distance and directional overcurrent relays. These configurations are considered as potential solutions to the synchronization problem. The goal of optimization is to find settings that minimize the cost function.

Problem formulation

The relay coordination problem is defined as an optimization problem where the objective is to minimize the cost function J(θ), where θ represents the parameters of the relays. The cost function can include measures such as relay response time (t_response), error detection accuracy (accuracy), and the number of false operations (n_false):

$$\text{J}(\uptheta )\hspace{0.17em}=\hspace{0.17em}\text{w}1\cdot \text{tresponse }(\uptheta )\hspace{0.17em}+\hspace{0.17em}\text{w}2\cdot (1\hspace{0.17em}-\hspace{0.17em}\text{accuracy }(\uptheta ))\hspace{0.17em}+\hspace{0.17em}\text{w}3\cdot \text{nfalse }(\uptheta )$$

(18)

where, w₁, w₂, and w₃ are the weights that show the importance of each criterion. BBO starts with an initial population of islands (TMS relays) in the algorithm and uses processes of migration and variation to improve the solutions. Migration refers to the transfer of successful traits from one island to another, while variation refers to creating biodiversity and generating new solutions through random variation.

In the optimization process with BBO, migration, and changes are the two main mechanisms that help to improve the solutions. As depicted in Fig. 3, the flowchart illustrates the process of the BBO algorithm.

Fig. 3

General flowchart of BBO algorithm

Migration allows successful traits from one setting to be transferred to another setting, while variation helps create biodiversity and generate new solutions. If the relay settings on island A led to a lower response time and the relay settings on island B have higher detection accuracy, migration can combine the successful features of these two islands to create a new island with a lower response time and higher detection accuracy. This process is done iteratively until the optimal settings for the relays are achieved, which helps to increase the reliability and efficiency of the power grid protection system.

Using GA to coordinate relays

GA uses mutation, crossover, and selection processes to evolve solutions. In this method, relay settings are considered chromosomes, and the best ones are selected to produce the next generations. Crossovers and mutations help create diversity in the population and allow new and improved settings to emerge. Chromosomes are the key variables of the GA algorithm, which include the TMS of relays and are characteristic of all overcurrent relays. Initially, several chromosomes are randomly assigned, and the initial population is generated to continue the algorithm. Chromosome structure considering TMS and characteristics of relays is shown in Fig. 4. TMS1, TMS2, …, TMS in this structure are the categories of setting CHAR₁, CHAR₂, …, CHAR_ng time and characteristics of relays R₁, R₂,… R, respectively.

Fig. 4

Chromosome structure of GA algorithm

Simulation results

Using a nine-bus system, we simulate the integration of Distributed Generation (DG) at bus 5. Installation of distance and directional overcurrent relays at each end of the transmission lines is assumed. DG alters short-circuit current levels, potentially leading to protection inconsistencies. Relay operation time and coordination distance are calculated, considering a maximum allowable time difference of 0.2 s when DG is connected. This simulation assesses the impact of DG integration on relay coordination.

Nine-bus system

A sample network consisting of nine buses was used to implement the proposed method, which is shown in Fig. 5 and includes six transmission lines, three transformers, and three generators. Bus number 5 is considered as one of the candidate busses for DG placement.

Fig. 5

Nine-bus system

As stated in the proposed method, it is assumed that a distance relay and a directional overcurrent relay are installed at both ends of the lines. The short-circuit current levels at different points of the grid are changed by installing Distributed Generation (DG) units in the grid, which can lead to protection inconsistencies in the network. In the studied network, it is assumed that a DG unit is installed in bus number 5, which is a synchronous type with a reactance of 0.25 per unit and is connected to the network by a transformer with a reactance of 0.05 per unit. The presence of this distributed generation unit may lead to a mismatch between the overcurrent and distance relays.

The operation time of the relays and the coordination distance between each pair of main and backup relays are calculated using the previous TSM adjustment values obtained for the relays without the presence of DG in the network to investigate its effect on the coordination of the relays. The allowed time difference between relays in case DG is connected to the network is 0.2 s.

Table 2 shows three-phase symmetrical short-circuit currents passing through the main and backup relays for faults at critical points. Isem and Isch are the short-circuit current passing through the main and backup relays per fault in the indicated locations. System fault currents change with the presence of DG. Therefore, a new protection structure needs explanation and simple solutions to deal with the protection issues caused by the use of DG.

Table 2 Short circuit current passing through relays at critical points

The fault currents of this new system should be accurately calculated and used in the design of the new protection system. Throughput fault currents can be programmed using BBO and GA optimization algorithm techniques, which achieves more accurate and efficient coordination of distance and directional overcurrent relays and ensures increased reliability and efficiency of the network protection system.

Table 3 presents the results of optimization, TMS, and operation time of main overcurrent relays per fault in three critical points: F2, F1, and F4. The operation time is equal to 0.02 s, not only at point F1 but also at 70% of the front line of the main relay. Increasing the operating speed of the overcurrent relay in this range makes the overcurrent relay like a fast backup distance relay for distance relays, and the equipment is less stressed when a fault occurs. Table 3 represents the difference between the operation time of the backup and main relays and the coordination time interval (∆tmb) for the pair of overcurrent and distance-directional overcurrent relays.

Table 3 Performance of backup and main relays using BBO and GA optimization algorithms for 9-bus system

Analyzing the coordination performance of the distance relay to the directional flow relay of the 9-bus system shows that the BBO algorithm typically provides lower coordination times compared to GA, which indicates the higher efficiency of BBO in optimizing the operation times of the relays. These findings can be effective in choosing the right algorithm to improve protection systems.

39-bus system

The proposed method was implemented in the 39-bus system to show efficiency. The 39-bus system includes 34 lines, 12 transformers, and ten generators, as shown in Fig. 6. Table 4 summarizes the differentiation time between the main and supporting relay pairs, for example, the number of 10 relay pairs for different cases resulting from the BBO and GA algorithms. As shown in Table 4, the simultaneous examination of TSM and characteristics of overcurrent relays as optimization variables lead to reduced differentiation times and the rate of non-coordination.

Fig. 6

39-bus system [46, 47]

Table 4 Performance of backup and main relays using BBO and GA optimization algorithms for 39-bus system

Therefore, the effectiveness of BBO and GA has been proven to solve the optimal coordination problem for the combination of distance and overcurrent relays in practical power systems. As shown in Table 4, using the protection coordination index to set the overcurrent relays in coordination with the distance relays has a positive effect on solving the protection problem. Table 4 presents the coordination results of distance and directional overcurrent relays for the 39-bus system using two different optimization approaches: the GA algorithm and the BBO algorithm. Optimization algorithms can be effectively used to fine-tune the coordination times between different relays and thus help improve the performance of the protection system. Comparing the results of GA and BBO can also guide choosing the best algorithm for specific applications. In this case, it is observed that BBO generally provides lower coordination times than GA, which can be taken as an indicator of BBO's higher efficiency in this context. Table 4 can be used to analyze and compare the performance of optimization algorithms in real conditions.

Power systems operate and control efficiently through the precise coordination of distance and directional overcurrent relays. This coordination, crucial for accurate fault detection and swift clearing sequence, is addressed in this paper through the application of heuristic techniques such as genetic algorithm (GA) and biogeography-based optimization (BBO) in a 9- and 39-bus system. The focus is on determining the optimal time setting factor (TSM) to minimize relay operating time, thereby enhancing the overall performance of the power system. Notably, the integration of intelligent features into the overcurrent relay contributes to the efficacy of the proposed coordination strategy.

One area of research ripe for exploration is the integration of advanced optimization algorithms with relay coordination techniques. While linear programming, genetic algorithms, and particle optimization methods have shown promise in optimizing relay settings, there is potential to explore more sophisticated algorithms, such as machine learning-based approaches, to adaptively optimize relay settings in real-time based on evolving system conditions and operational requirements. By leveraging historical operational data and advanced analytics, machine learning algorithms could offer enhanced predictive capabilities, enabling proactive relay coordination strategies that anticipate and mitigate potential system vulnerabilities before they escalate into critical faults.

Furthermore, the impact of high-RES penetration on power system dynamics necessitates a holistic reassessment of protection strategies to ensure the resilience and stability of the grid. With the increasing variability and intermittency of renewable generation, traditional protection schemes designed for conventional generation sources may prove inadequate in effectively safeguarding against emerging risks, such as voltage fluctuations, frequency deviations, and islanding events. Thus, there is a pressing need to develop adaptive protection schemes that dynamically adjust relay settings and coordination logic to accommodate the unique operating characteristics of renewable energy resources and mitigate their potential destabilizing effects on the grid.

Moreover, the proliferation of distributed energy resources (DERs), including rooftop solar panels, wind turbines, and energy storage systems, introduces additional complexities to relay coordination efforts. Traditional centralized protection schemes may struggle to effectively coordinate relay operations across distributed assets, leading to suboptimal fault detection and isolation performance. To address this challenge, researchers could explore decentralized coordination approaches that leverage distributed intelligence and communication protocols to enable seamless coordination among interconnected DERs. By decentralizing control logic and leveraging local sensing and decision-making capabilities, decentralized protection schemes could enhance the scalability, resilience, and adaptability of power system protection architectures in the era of distributed energy.

Another area deserving of attention is the development of integrated protection and control strategies that leverage advanced sensing, communication, and control technologies to enhance situational awareness and system resilience. By integrating protection functions with wide-area monitoring systems (WAMS) and synchrophasor technologies, utilities can gain real-time insights into grid conditions and rapidly respond to emerging threats, such as cascading failures and transient disturbances. Furthermore, the deployment of intelligent electronic devices (IEDs) and digital substations enables advanced fault detection and localization capabilities, allowing for faster fault isolation and restoration, thereby minimizing outage durations and enhancing grid reliability.

In the context of advancing relay coordination, the paper highlights the pivotal role of edge-based server capabilities. By leveraging sophisticated algorithms and communication technologies at the edge, relays can be activated promptly, meeting stringent coordination requirements. This integration of edge computing, GA, and BBO techniques aims to increase the efficiency and effectiveness of relay coordination. In edge computing, terminal users are connected to edge servers via the wireless network, where various channels exist within each wireless link [50]. This architecture facilitates efficient communication between users and edge servers, enabling seamless data processing and delivery at the network's edge.

The results demonstrate a significant improvement over traditional method. This demonstrates the potential of this hybrid approach for enhancing power system reliability and efficiency. However, the comparative analysis suggests that while both GA and BBO contribute positively, BBO does not exhibit clear superiority over GA in achieving better results. This underlines the nuanced considerations in selecting optimization techniques for specific applications.

Limitations and outlooks

In this subsection, we provide a detailed analysis of the potential limitations associated with the use of GA and BBO in the context of relay coordination. We address issues related to solution quality, convergence speed, and sensitivity to parameter settings.

While GA and BBO are powerful heuristic optimization techniques, they may not always guarantee optimal solutions due to the stochastic nature of their search processes. Suboptimal solutions may arise from premature convergence to local optima or inadequate exploration of the solution space.

The convergence speed of GA and BBO can vary depending on factors such as population size, crossover and mutation rates, and the complexity of the optimization problem. In some cases, these algorithms may require a large number of iterations to converge to satisfactory solutions, leading to longer computational times.

GA and BBO performance can be sensitive to the selection of algorithmic parameters, such as mutation rates, migration rates, and selection mechanisms. Poorly chosen parameter settings may hinder convergence, result in premature convergence, or lead to suboptimal solutions.

Addressing these limitations requires careful consideration and fine-tuning of algorithmic parameters, as well as the exploration of alternative optimization strategies. Hybrid approaches that combine GA or BBO with other optimization techniques or problem-specific knowledge may enhance solution quality, convergence speed, and robustness.

Moreover, while our study explores the innovative optimization of distance and directional overcurrent relay coordination using heuristic methods, it is essential to acknowledge certain limitations and assumptions associated with these approaches.

• 1. Complexity of Power Systems: Heuristic methods, such as genetic algorithms and biogeography-based optimization, offer powerful tools for solving complex optimization problems. However, the inherent complexity of power systems presents challenges in accurately modeling all system dynamics and constraints. The simplifications and abstractions required to formulate the optimization problem may introduce limitations in capturing the full intricacies of real-world scenarios.

• 2. Computational Resources: Heuristic methods often require significant computational resources, especially when applied to large-scale power systems with numerous relays and complex coordination requirements. While edge computing can alleviate some computational burden by distributing processing tasks closer to the data source, resource constraints on edge devices may still limit the scalability and efficiency of heuristic optimization algorithms.

• 3. Sensitivity to Initial Conditions: Heuristic methods, including genetic algorithms and biogeography-based optimization, are sensitive to initial parameter settings and population configurations. Suboptimal initial conditions or parameter choices may lead to premature convergence or suboptimal solutions. Careful tuning and validation of algorithmic parameters are necessary to ensure robust and reliable performance.

• 4. Trade-off Between Exploration and Exploitation: Heuristic optimization algorithms must strike a balance between exploring the solution space to discover new promising regions and exploiting known solutions to refine and improve performance. However, finding the optimal balance between exploration and exploitation can be challenging, particularly in dynamic and uncertain environments characteristic of power systems operation.

• 5. Generalizability and Transferability: While heuristic methods have shown promising results in optimizing relay coordination for specific power system configurations and operating conditions, their generalizability and transferability to diverse settings may be limited. Extrapolating findings from one case study to different contexts or scales requires careful consideration of underlying assumptions and potential differences in system characteristics.

Besides, while our paper demonstrates the feasibility and benefits of leveraging edge computing for relay activation, further research is warranted to explore optimization opportunities, scalability challenges, and integration complexities associated with large-scale deployment in real-world power systems.

In conclusion, the incorporation of edge computing technologies into our relay coordination framework represents a significant advancement in enhancing the efficiency and responsiveness of power system operations. By harnessing the computational capabilities of edge devices, we can achieve real-time relay activation, thereby strengthening the resilience and reliability of modern power grids.

Conclusion

This study systematically assessed the optimal coordination of distance and directional overcurrent relays within 9- and 39-bus systems. The investigation focused on determining the optimal operating time, time setting factor (TSM), and harvest currents of relays. Two distinct optimization techniques, namely genetic algorithm (GA) and biogeography-based optimization (BBO), were employed to perform these calculations. The study constraints were meticulously adhered to. This was done with a keen emphasis on defining the smart overcurrent features necessary to attain the desired outcomes. Comparative analysis between GA and BBO results revealed that, on the whole, BBO outperformed GA, providing superior results. This finding positions BBO as a more pragmatic approach to relay coordination in power systems. Furthermore, the proposed protection settings for the discussed power network were deemed satisfactory. This underlines the considerable potential of these techniques to enhance power system performance. To consolidate these positive outcomes, relay coordination strategies are recommended for larger test systems. This expansion will enable a comprehensive evaluation of their capabilities on a more extensive scale. In the context of an evolving power landscape characterized by increasing complexity, this study represents a crucial step towards innovative solutions. As power networks grow in complexity, the importance of edge calculations becomes paramount. Future research should focus on refining and expanding relay coordination strategies. This should emphasize their adaptability and efficacy to larger and more intricate power systems. Additionally, exploring the integration of advanced technologies such as machine learning or artificial intelligence into relay coordination could pave the way for even more sophisticated and adaptive solutions in the realm of power system protection.