A Study the Eho-FF Algorithms for Optimal Distribution energy resource Sizing
Sachin Suke1*, Dr. D. S. Bhangari2
1 Research Scholar, Shridhar University, Pillani, Rajasthan, India
2 Professor, Shridhar University,Pillani, Rajasthan,India
Abstract- The advent of the Smart Grid idea has presented a significant obstacle in the power industry. The smart grid now mostly consists of distributed power that integrates renewable energy resources. Demand Side Management (DSM) adapts to fluctuating demand by letting customers choose their own power generation options. By limiting demand & maximising resource allocation for power generation, DSM enables customers to access power at a reduced cost. Developing a hybrid evolutionary algorithm for DSM that accurately monitors smart grid real power loss is the main goal of this study. This method will be useful for reactive power optimisation, or RPO. Improving the power grid network's energy efficiency and voltage profile under various load situations is the goal of the proposed effort.
Keywords- Smart Grid, Demand Side Management, Distributed Energy Resources, Reactive Power Optimisation, hybrid evolutionary algorithm, PSO
INTRODUCTION
The smart grid environment has seen a significant growth in the penetration levels of Distributed Energy Resources (DERs) into the existing system. A system's DG units can be located and sized with the help of optimisation techniques, which allow for their best utilisation within predetermined boundaries and constraints. The main goal of this research is to develop a DSM hybrid evolutionary algorithm that can precisely monitor actual smart grid power loss; this will help with Reactive Power Optimisation (RPO). Improving the power grid network's energy efficiency and voltage profile under various load situations is the goal of the proposed effort.
PROBLEM FORMULATION
The RPO takes into account operational and physical limitations in addition to the formulae for reactive and active power balance at each bus. Minimising power loss (Ploss), as shown in Equation (1), is its goal function.
The magnitude of bus i's voltage is Vi;, bus j's voltage is Vj, & conductance between buses i & j is Gij. The total number of branches in the network is denoted by S, and the voltage angle between buse si and j is θij. As shown in Equations (2) and (3), the power balancing equation limitations at each bus i are demonstrated.
Reactive power generation output limitations, static VAR compensator capacity limits, allowable voltage magnitude operating ranges (excluding slack bus), and transformer tap position constraints are detailed in Equation (4).
Where,
In this case, Bij represents the mutual susceptibility of buses i and j. Bus angles, load bus voltage magnitude, generator bus reactive power, & slack bus real power generation are all examples of state variables in an RPO, while generator bus voltage magnitude, shunt capacitor/reactor output, and transformer ratios are examples of control variables. This classification is standard in optimisation problems. The system's buses are supposed to be in the following order: slack, generator, & load (with load buses containing shunt capacitors/reactors coming previous in the sequence for easier expression). Equation (6) gives the state variable vector (X) based on this order,
Another thing that Equation (7) describes is the control variables, which are U.,
With the RPO's focus on optimising the voltage profile & reactive power, certain limitations that were formerly part of the ideal power flow have been eliminated. These constraints pertain to the actual power line flow & slack bus's real power generation. We know how to generate active power. In practical terms, the transformer and shunt capacitor/reactor ratios are discrete variables, and the penalty function is a good way to deal with these variables.
It has been noted that the RPO becomes an uncertain nonlinear programming issue due to the presence of multiple uncertain components in its input data. An interval, fuzzy integer, or even a random number can be used to depict this uncertainty. Because the RPO model is self-validated, the intervals used to characterise its uncertainties are selected for this study. It is easy to get the bound information on the uncertainty using engineering approaches. Transmission line parameter uncertainties, for example, pale in comparison to the uncertainty connected with active & reactive power changes. This study does not consider any other uncertainty besides active power generation & load demand (Cetai er al. 209). Equation (8) describes active power generation, and Equation (9) describes active power demand, for the sake of clarity.
The reactive power demand is evaluated 85 given by Equation (10),
The slack & generator buses are thought to have predictable active and reactive power loads because they are mostly generated by power plants. Equations (11) through (13) restructure the RPO model following the development of the equilibrium equations including interval uncertainty.
Here the interval functions are defined as f(X,U),h(X,U) & g (X, U), the interval of the real power losses is & interval vector of the power flow variation is represented as . X & U, the state & control variable vectors, respectively, correspond to the RPO model's X and U. U is a real-valued vector in the suggested model since the excitation mechanism fixes the magnitude of the generators' voltages. Additionally, the output of the shunt capacitors/reactors must be coordinated manually with the transformer ratios. Given that its unregulated nature is present in interval input data, U is formed of interval values, unlike X. In addition, every time U is determined by combining the interval power flow equations, X is also determined by U; that is, for every U, there is an associated X. Naturally, the issue is resolved by applying the techniques of nonlinear programming issues including interval uncertainty. Deterministic RPO models utilise the linear approximation method with internal control. The suggested hybrid EHO-FF evolutionary algorithm is an improved way to get the intervals of the unknown power flow equations, which further improves its accuracy.
RESULTS AND DISCUSSION
The results of the simulation work are described in this section. On a platform consisting of an Intel (R) Core (TM) two CPU, 4GB RAM, and MATLAB 7.10.0 (R2014a), the code has been built to optimise reactive power in the smart grid by positioning DG units. The suggested work is tested against the benchmark systems of the IEEE 30 bus & the IEEE 57 bus to ensure its efficacy. The appendix provides a full single-line schematic of the IEEE 30 Bus system, including load and branch data. The 283.4 MW and 26.2 MVAr total loads of the six producing units & eight transmission networks that make up the IEEE 30 bus test system's 41 branches.
Voltage, power loss, active power, & reactive power tests are conducted on all transmission lines initially. DGs operate well at generator buses that employ a hybrid algorithm to boost distribution network voltage and decrease reactive power compensation loss. A number of meta-heuristic algorithms, including Bat & PSO, are employed to verify that the proposed hybrid method is effective.
A number of optimisation procedures were modelled in an effort to resolve the RPO issue; Table 1 displays the usual or selected values for the different parameters.
Table 1: Optimisation Algorithm Parameters
Base Case Solution of IEEE 30 Bus Test System
The study's investigated IEEE 30 bus test system follows a typical 24-hour load pattern, as shown in Figure 1. Using the base case load data & Newton-Raphson approach, the load flow program is initially executed for the IEEE 30 bus system. Table 2 shows the load flow solution for each bus used in the test system. It contains the magnitude of the bus voltage, real power, & reactive power generation/load.
Figure 1: Daily load curve for IEEE 30 bus system
Table 2: Optimisation of an IEEE 30 bus system's load flow utilizing Newton-Raphson
Table 3 displays the power flow diagram and reactive and actual power losses for each transmission line in the IEEE 30 bus system. With base case load conditions, the reactive power line losses for the IEEE 30 bus system were 45.58MVAr & real power line losses were 11.463MW, as determined by the evaluation utilising the Newton-Raphson technique. Using the load flow solution, we can calculate the line flows & losses in each of the branches. Reactive power generation is to 53.944MVAr, while total active power generation is estimated to be 297.663MW. Tables 2 and 3 display the outcomes of the load generation balance equation for active & reactive power, total load, and line losses, respectively.
Table 3: Power flow & losses in an IEEE 30 bus system utilizing the Newton-Raphson model
Optimising Efficiency utilising PSO for the IEEE 30 Bus System
In the current investigation of the 30 bus test system for IEEE compatibility, DG units are installed on generator buses. Specifically, DGI is located on bus 1, DG2 on bus 2, DG3 on bus 6, DG 4 on bus 3, DG5 on bus 22, & DG6 on bus 27. Figure 1 shows the varied load pattern that may be achieved by applying the PSO algorithm to determine the ideal size of distributed generation units (DG) installed at specific places. The PSO algorithm's evaluation of the base case line flows and losses is displayed in Table 4. Analysis of the PSO algorithm yielded real power line losses of 0.886 MW and reactive power line losses of 43.074 MV Ar for the IEEE 30 bus system under base case load conditions.
Table 4: PSO algorithm power loss & flow evaluation in IEEE 30 Bus system
In Table 5, we can see the total real power loss that was computed with and without the deployment of DG utilising the PSO algorithm. Assessing reactive & real power losses for every load scenario follows the optimal sizing of DG units. Real power losses average 11.85 MW when DG is not implemented and 9.66 MW when it is. The results indicate that after DGs are placed in the designated places with the right size, real power loss is decreased.
Table 5: Evaluate power losses with DG in an IEEE 30 Bus system with variable load demand using the PSO technique.
Bat Algorithm for Optimal DG Unit Sizing in IEEE 30 Bus Systems
Similarly to how the Bat algorithm was used for the PSO method, this one is used to find the ideal size of DG units to be put at specific locations in order to accommodate different load patterns, as seen in Figure 1. According to Table 6, the line losses for the base case load scenario of the IEEE 30 bus system, as determined by the Bat method assessments, are 34.389 MVAr and 8.556 MW. Table 7 displays the outcomes of calculating power losses for every load scenario following the installation of appropriately sized DG units. Average real power losses are 11.4592 MW when DG is not used, and they decrease to 7.3971 MW when DG is used. Using this method for optimal DG unit sizing led to the conclusion that DG put with suitable sizing at the specified sites reduces average real power loss. Moreover, when all potential loads are considered, the Bat algorithm is found to reduce reactive & actual power losses on average compared to the PSO algorithm's results.
Table 6: IEEE 30 bus power flow and losses under bat algorithm
Table 7: Bat algorithm investigation of IEEE 30 Bus variable load demand power losses with DG.
Hybrid EHO-FF Method for the Sizing of DG Units in an IEEE 30 Bus System
The proposed hybrid evolutionary algorithm integrates the EHO and FF algorithms to mimic the PSO & Bat algorithms. It is now being used in the IEEE 30 bus test system that accounts for DG placement for various load patterns. Total line losses assessed using the suggested hybrid algorithm are 7.60 MW and 30.674 MV Ar, as shown in Table 8, which displays the power flow & losses in the transmission lines utilising the proposed method.
Table 8: Power consumption and losses in an EHO-FF algorithm-based IEEE 30 bus system
In Table 9, we can see the results of simulating the ideal DG unit sizing for each load pattern using the suggested EHO-FF method. We also analyse the line losses that occur after DG placement at the selected sites (generator bus). The suggested hybrid EHO-FF algorithm decreases average power loss to 11.4592 MW without DG placement & 7.3971 MW after DG placement, in comparison to PSO and Bat algorithms.
Table 9: A proposed algorithm for DG power loss analysis in IEEE 30 bus systems with variable load demand.
EHO-FF Algorithm with a Hybrid Approach for Optimal DG Unit Sizing in IEEE 57 Bus Systems
The proposed hybrid approach, which integrates the EHO and FF algorithms, is now being used by the IEEE 57 bus test system that includes DG placement for different load patterns (as shown in Figure). Table 10 shows the results of the evaluation of the appropriate DG unit sizing for each load pattern and real power losses. Results from the suggested hybrid EHO-FF algorithm outperform those from the PSO and bat algorithms.
Table 10: Power losses with DG for variable load demand in IEEE 57 bus system using proposed algorithm
CONCLUSION
The suggested hybrid method outperforms the state-of-the-art methods and is supported by appropriate outcomes in a comparison of optimisation algorithms used for DG unit optimal sizing in IEEE 30 bus & IEEE 57 bus test systems. Because of this, we can say that the suggested hybrid algorithm improves performance for different types of power sector loads and is more effective at reactive power optimisation for optimally sizing distributed generation in smart grid environments. A novel hybrid EHO-FF method for optimising DG units deployed at indicated locations was described & developed by Muthukumaran et al. (2021). We discuss the test systems that incorporate all operative limitations and use a 24-hour load pattern to solve the objective function that has been formulated.
REFERENCES
- Calafate, C.T. 2009, ‘Qos support in MANET: a modular architecture based on the IEEE 802.11e technology’. Circuits and Systems for Video Technology, IEEE Transactions on Volume:19 , Issue: 5 pp.678-692.
- Bose, 2010, ‘ Smart transmission grid application and their supporting infrastructure’. A Smart Grid, IEEE Transactions on Volume: 1, Issue: 1. pp.11-19.
- Arif, S. M., Hussain, A., Lie, T. T., Ahsan, S. M., & Khan, H. A. (2020). Analytical hybrid particle swarm optimization algorithm for optimal siting and sizing of distributed generation in smart grid. Journal of Modern Power Systems and Clean Energy, 8(6), 1221-1230.
- Duman, S, Rivera, S, Li, J & Wu, L 2020, ‘Optimal power flow of power systems with controllable wind-photovoltaic energy systems via differential evolutionary particle swarm optimization’, International Transactions on Electrical Energy Systems, vol. 30, no. 4, pp. 1-28.
- Ghasemi, M, Ghavidel, S, Ghanbarian, MM, Gharibzadeh, M & Vahed, AA 2014c, ‘Multi-objective optimal power flow considering the cost, emission, voltage deviation and power losses using multi-objective modified imperialist competitive algorithm’, Energy, vol. 78, pp. 276-289
- Ghasemi, M, Ghavidel, S, Rahmani, S, Roosta, A & Falah, H 2014a, ‘A novel hybrid algorithm of imperialist competitive algorithm and teaching learning algorithm for optimal power flow problem with non-smooth cost functions’, Engineering Applications of Artificial Intelligence, vol. 29, pp. 54-69.
- Kannayeram, G. P., Muniraj, R., Prakash, N. B., Jarin, T., & Boselin Prabhu, S. R. (2021). An elitist control scheme for power flow management in smart grid system: a hybrid optimization scheme. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 1-22.
- Kessel, P. and Glavitsch, H. (1986). Estimating the voltage stability of a power system. IEEE Transactions on Power Delivery., 1(3):346-354.
- Kumar, AR & Premalatha, L 2015, ‘Optimal power flow for a deregulated power system using adaptive real coded biogeography-based optimization’, International Journal of Electrical Power & Energy Systems, vol. 73, pp. 393-399.
- Kumar, S, Tyagi, B, Kumar, V & Chohan, S, 2018, ‗Incremental PMU placement considering reliability of power system network using analytical hierarchical process‘, IET Generation, Transmission & Distribution, vol. 12, no. 16, pp. 3900-3909.
- Zhao, Y, Cai, Y & Cheng, D 2017, ‗A novel local exploitation scheme for conditionally breeding real-coded genetic algorithm‘, Multimedia Tools and Applications, vol. 76, no. 17, pp.17955-17969.
- Zehar, K & Sayah, S 2008, ‘Optimal power flow with environmental constraint using a fast successive linear programming algorithm: Application to the Algerian power system’, Energy Conversion and Management, vol. 49, no. 11, pp. 3362-3366.