Pith. sign in

REVIEW 4 major objections 4 minor 27 references

Green Economic Load Dispatch: A Review and Implementation

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read PSO beats GA on green economic dispatch in about a third of the time.

desk verdict The review half is a clear, if dated, survey, but the headline PSO-vs-GA comparison is undermined by a load-balance violation in the paper's own 1500 MW table. read the letter →

arxiv 2506.12062 v1 pith:L7VUXZ7X submitted 2025-05-31 cs.NE

Shahbaz Hussain This is my paper
classification cs.NE
keywords economicloaddispatchcombinedemissionparticleswarmoptimizationgeneticalgorithmIEEE30-bussystempenaltyfactorMATLABsimulation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper argues that combined economic emission dispatch (CEED)—deciding how much each thermal generator should produce to minimize both fuel cost and pollutant emissions while meeting demand—can be solved well by population-based AI methods, and that particle swarm optimization (PSO) outperforms a genetic algorithm (GA) on a standard test case. On the lossless IEEE 30-bus six-generator system at 1500 MW and 2000 MW, PSO reports lower total cost (fuel plus emission-weighted cost) than GA and runs in roughly a quarter to a third of the wall-clock time. The author credits PSO's simple two-equation velocity-and-position update for faster convergence to the claimed global optimum, compared with GA's selection, crossover, and mutation layers. The paper also surveys ten AI techniques for CEED as context for the implementation.

What carries the argument

The load-bearing machinery is the price-penalty-factor transformation that collapses the two conflicting objectives—fuel cost and pollutant emission—into one scalar objective, together with the two competing search mechanisms. The penalty factor $h$ is computed per gas (for example, $h_{\mathrm{NOX}}=3.1669$, $h_{\mathrm{COX}}=0.1221$, and $h_{\mathrm{SOX}}=0.9182$ at 1500 MW) so that the combined cost is $k_1 F_T + k_2 h E_T$; both algorithms then minimize that single number. PSO carries the search with velocity and position update equations using a constriction factor (CF $=0.7298$, $c_1=c_2=2.05$), while GA carries it with binary-string selection, crossover, and mutation. The paper credits the simplicity of PSO's two equations for its faster convergence.

What would settle it

Re-run the same six-generator IEEE 30-bus CEED case at 1500 MW with a genetic algorithm that uses a larger population (for example, 100 individuals) or a convergence-based stopping rule instead of a fixed 500 iterations; if GA then matches or beats PSO's $33,948.83/h total cost in comparable time, the reported speed and cost ranking would not be a stable property of the algorithms.

Watch

Extended reading notes

Core claim

The paper's central discovery, on its own terms, is that on a lossless IEEE 30-bus system with six thermal generators, particle swarm optimization solves the combined economic emission dispatch problem better than a genetic algorithm. With all three pollutant classes (NOX, COX, SOX) folded into the fuel cost through price penalty factors, PSO reports a total cost of $33,948.83/h at 1500 MW and $56,988.30/h at 2000 MW, against GA's $34,005.52/h and $57,097.16/h, and it does so in 0.35 s and 0.36 s average operating time versus GA's 1.30 s and 1.54 s. The author attributes the gap to PSO's simpler two-equation velocity-position update compared with GA's three-layer selection, crossover, and mutation machinery, which lets PSO converge to the claimed global optimum in fewer iterations.

Load-bearing premise

The ranking depends on treating 10 particles versus 10 individuals with the listed parameter settings over 500 iterations as a fair and representative tuning of both algorithms, and on taking the best-of-50-trials results as typical performance.

Editorial extensions

If this is right

  • The fixed penalty factors computed for the three gases define a reproducible single-objective dispatch problem that any optimizer can be benchmarked on.
  • For small benchmark systems, PSO's lower total cost and much shorter running time make it a practical first-choice optimizer for CEED when dispatch solutions must be recomputed as load changes.
  • GA still returns close total costs (about $57/h$ and $109/h$ more at the two loads), so both algorithms are near each other in solution quality; the clear difference is computational time.
  • Because both methods are parents of the other surveyed techniques, the paper's framing implies that hybrid algorithms borrowing PSO's velocity update and GA's recombination are a natural next step for better CEED performance.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper: the absence of a convergence criterion means the measured speed gap partly reflects the fixed 500-iteration budget; re-running with function-evaluation or convergence-based stopping would separate algorithmic speed from iteration-count artifacts.
  • Beyond the paper: the penalty-factor values depend on the load demand and the gas-specific formula, so the 'green' weighting is not a fixed constant; different penalty schedules could dispatch different generators and change the emission profile.
  • Beyond the paper: the lossless six-generator test is a smooth, small problem; on larger systems with line losses and valve-point effects, GA's population diversity may outperform PSO's faster local convergence, so the ranking should not be assumed to scale.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper is a combined review and implementation study of artificial intelligence techniques for the combined economic emission dispatch (CEED) problem. The review portion surveys ten metaheuristic methods (PSO, GA, ABC, ACSA, FA, SFLA, EP, SA, GSA, BBO) and summarizes prior CEED applications. The implementation portion formulates CEED with a weighted penalty-factor objective, applies PSO and GA to a six-generator IEEE 30-bus system for load demands of 1500 MW and 2000 MW, and compares the two algorithms in terms of total cost (fuel plus emission penalty) and wall-clock time. The central claim, stated in Section 6, is that PSO converges faster to the global optimal solution than GA, based on the reported table and convergence figures.

Significance. If the numerical comparison were sound, the paper would provide a straightforward empirical benchmark of two canonical metaheuristics on a classic CEED test case, along with a useful literature digest. The review portion is a reasonable, if introductory, catalog of ten AI methods, and the paper honestly states that transmission losses are neglected. However, the empirical contribution is weakened by the absence of any optimality baseline, statistical spread, or reproducibility details, and, most seriously, by an infeasible reported dispatch. The significance of the headline claim is therefore limited: a corrected, properly benchmarked comparison would be a modest but useful data point, but in its current form the paper does not establish the claimed PSO superiority.

major comments (4)
  1. [Section 5, Table] The PSO solution for PD=1500 MW violates the load balance constraint stated in Eq. (5). The reported P1..P6 values (195.79, 256.55, 381.25, 81.69, 381.85, 202.27) sum to 1499.40 MW, not 1500 MW, and since transmission losses are explicitly not accounted for, Eq. (5) requires exact equality. An under-generating dispatch of 0.6 MW can trivially reduce fuel and emission costs, so the PSO-vs-GA comparison at 1500 MW is invalid. The other three rows in the table sum exactly to their required loads, indicating a targeted infeasibility rather than simple rounding. This directly undermines the conclusion's strongest claim.
  2. [Section 6, Conclusion] The claim that PSO 'converges faster to global optimal solution' is unsupported because no independent optimum is provided. There is no comparison with an established optimal solution for this test system (e.g., lambda-iteration, dynamic programming, or best-known values from the literature), and no convergence criterion is defined; the fixed 500-iteration budget is a stopping rule, not evidence of convergence. The term 'global optimal' should either be replaced by 'best solution found' or justified by a reference baseline.
  3. [Section 5, Simulation Results] The results from 50 stochastic trials are summarized only by a single value for each quantity, with no standard deviation, best/worst bounds, or statistical significance test. The reported total-cost difference between PSO and GA at 1500 MW (33948.83 vs 34005.52, about 0.17%) may be smaller than the run-to-run variability of either algorithm. Report mean ± standard deviation over the 50 trials, and ideally a paired test, before claiming that PSO outperforms GA.
  4. [Section 2, Eq. (8) and following procedure] The description of the price-penalty-factor calculation is unclear: the sentence 'These obtained his are listed in increasing order and added with Pi,max of every generator one by one starting from the first hi in the list until Σ Pi,max ≥ PD' does not define a reproducible algorithm. Please rewrite this step-by-step, and show explicitly how the penalty factors hNOX, hCOX, and hSOX for PD=1500 MW and 2000 MW are computed from Eq. (8) and the coefficients taken from [3].
minor comments (4)
  1. [Section 5, Table] The sentence 'The solutions with average operating time (t) have been selected out of 50 trials' is grammatically unclear; it should state how the reported generation values were selected (e.g., from the trial with median cost or median time).
  2. [Section 3.x] There are several typographical errors: 'Biogeoraphy' should be 'Biogeography', 'offsrings' should be 'offsprings', and 'dictates' in Section 6 should be 'indicate' or 'show'.
  3. [Figures 1 and 2] The convergence figures are not described in the text; specify whether the vertical axis is total cost or fitness value, and describe the convergence behavior that the figures are meant to illustrate.
  4. [Section 1, Introduction] The claim that 'classical optimization methods lack the ability' to solve the multi-objective problem is made without a citation or a direct comparison; please support or soften this statement.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the PSO/GA comparison is an empirical benchmark, with any validity concerns being correctness issues rather than circular reasoning.

full rationale

No circular reasoning is found in the paper's derivation chain. The implementation imports generator cost and emission coefficients from an external source [3], computes price penalty factors from the stated Eq. (8) procedure for each load demand, then runs PSO and GA on the same scalarized CEED objective and reports the resulting costs, emissions, and runtimes. The penalty factor h is computed from generator limits and load demand before optimization, so the objective is not defined in terms of the output it is used to produce. The PSO-versus-GA ranking is an empirical benchmark result, not a fitted parameter renamed as a prediction, and the algorithms are not tuned to force the reported ranking. There are no self-citations by the present author and no load-bearing appeal to a prior uniqueness theorem. The skeptical concern that the 1500-MW PSO row sums to 1499.40 MW and thus violates the stated power-balance constraint Eq. (5) is a correctness and feasibility defect, not a circularity: an infeasible solution can invalidate the comparison without making the derivation tautological. Likewise, the absence of an independent optimum baseline makes the claim of reaching a 'global optimal solution' unsupported, but that is an evidential gap rather than a reduction of the result to its own inputs. Because the central claim is self-contained empirical comparison rather than a derivation equivalent to its inputs, the circularity score is 0.

Assumptions & free parameters 9 free parameters · 6 assumptions · 0 invented entities

The paper contributes no new physical or mathematical entities. Its conclusions rest on external IEEE 30-bus coefficient data from [3], on a penalty-factor scalarization from [2], and on hand-chosen hyperparameters for PSO and GA; these are the inputs the reader must accept to trust the comparison.

free parameters (9)
  • PSO population size = 10
    Set in Section 5; very small population, central to the speed comparison.
  • PSO inertia weight range = mu_max=0.9, mu_min=0.4
    Set in Section 5; no sensitivity study provided.
  • PSO acceleration constants c1, c2 = 2.05, 2.05
    Set in Section 5.
  • PSO constriction factor = 0.7298
    Set in Section 5 following reference [4].
  • GA population size = 10
    Set in Section 5; same population size as PSO.
  • GA crossover probability = 0.96
    Set in Section 5.
  • GA mutation probability = 0.033
    Set in Section 5.
  • Iteration count for both algorithms = 500
    Stopping rule for both algorithms in Section 5; no convergence criterion defined.
  • Number of trials = 50
    Section 5 says 'solutions with average operating time have been selected out of 50 trials'; no variance is reported.
assumptions (6)
  • domain assumption Generator fuel cost and emission are quadratic functions of real power (Eqs. 1 and 2).
    Standard CEED approximation used throughout, but it ignores valve-point load effects and non-smooth characteristics.
  • domain assumption Total generation equals load demand with no transmission losses (Eq. 5).
    Section 5 explicitly says losses are not accounted for, so the comparison only holds for the lossless model.
  • domain assumption Generator outputs remain within unspecified minimum and maximum limits (Eq. 6).
    The paper does not list the six generators' limits, and simply assumes the reported solutions satisfy them.
  • domain assumption The penalty factor method converts the multi-objective CEED problem into a single scalar objective (Eqs. 7 and 8).
    This scalarization from reference [2] is one of several possible ways to combine cost and emission; the chosen weights directly set the reported total cost.
  • domain assumption The IEEE 30-bus cost and emission coefficient data from reference [3] are accurate and applicable.
    All coefficients are taken from an external paper and not reproduced or validated in this work.
  • ad hoc to paper Five hundred iterations are sufficient for both algorithms to converge.
    A fixed stopping rule is chosen in Section 5 with no convergence criterion or proof that the optimum is reached.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Green Economic Load Dispatch: A Review and Implementation." pith.science (2026). https://pith.science/paper/L7VUXZ7X

@misc{pith2026250612062,
  author       = {Pith},
  title        = {Pith review of: Green Economic Load Dispatch: A Review and Implementation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L7VUXZ7X}},
  note         = {Machine review of arXiv:2506.12062}
}
read the original abstract

The economic dispatch of generators is a major concern in thermal power plants that governs the share of each generating unit with an objective of minimizing fuel cost by fulfilling load demand. This problem is not as simple as it looks because of system constraints that cannot be neglected practically. Moreover, increased awareness of clean technology imposes another important limit on the emission of pollutants obtained from burning of fossil fuels. Classical optimization methods lack the ability of solving such a complex and multi-objective problem. Hence, various modern artificial intelligence (AI) techniques based on evolution and social behaviour of organisms are being used to solve such problems because they are easier to implement, give accurate results and take less computational time. In this work, a study is done on most of the contemporary basic AI techniques being used in literature for power systems in general and combined economic emission dispatch (CEED) in particular. The dispatch problem is implemented on IEEE 30-bus benchmarked system in MATLAB for different load demands considering all gases (COX, NOX and SOX) using particle swarm optimization (PSO) and genetic algorithm (GA) and their results are compared with each other.

Figures

Figures reproduced from arXiv: 2506.12062 by the authors.

Figure 1
Figure 1. Convergence characteristics of PSO and GA for IEEE 30 bus system with PD = 1500 MW [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Convergence characteristics of PSO and GA for IEEE 30 [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

27 extracted references · 27 canonical work pages

  1. [3]

    ∑ (3) where PG is total real power of N generators, PD is total load demand and PL represents total transmission losses that can be calculated using B -coefficient matrix ( Bmn) by the following relation: ∑ ∑ (4) If losses are ignored, then ∑ (5)

    work page

  2. [5]

    Particle Swarm Optimization In 1995 , particle swarm optimization was invented by two scientists Kennedy and Eberhart

    AI TECHNIQUES A brief introduction of AI techniques reported in literature for multi-objective optimization problems in power systems is given below: i. Particle Swarm Optimization In 1995 , particle swarm optimization was invented by two scientists Kennedy and Eberhart. They were actually studying the patterns of social interaction within swarms of fishe...

    work page 1995

  3. [1]

    The total generation should meet the load demand of consumers and also the line losses if they are considered

    INTRODUCTION The goal of generators dispatch in a thermal power plant is to deliver at optimal point by satisfying system constraints for economy saving. The total generation should meet the load demand of consumers and also the line losses if they are considered. To achieve this objective, generators f uel cost which constitutes mainly the generation cos...

    work page

  4. [2]

    Both fuel cost and pollutants emission can be equated as quadratic functions of generator real power

    PROBLEM FORMULATION Combined economic emission dispatch in a thermal power plant is to minimize both fuel cost and pollutants emission simultaneously such that generation equals load demand with transmission losses and no unit violates its generation limits. Both fuel cost and pollutants emission can be equated as quadratic functions of generator real pow...

    work page

  5. [4]

    All units should generate power within their minimum and maximum limits i.e. Sci.Int.(Lahore),28(2),1401-1406,2016 ISSN 1013-5316; CODEN: SINTE 8 1402 March-April (6) The objectives of mi nimizing fuel cost and emission can be made a single objective by the concept of penalty factor [1]: (7) where FT and ET are fuel cost and emission , k1 and k2 are const...

    work page 2016

  6. [6]

    Partition of frogs into units

    work page

  7. [7]

    Shuffling of frogs The steps 3-5 keeps on iterating until the global evolutionary steps meet the maximum set limit [9]. vii. Evolutionary Programming (EP) EP is inspired by the biological evolution of human beings. It differs from the mainstream genetic algorithms by the fact that offsprings are created by mutation only and cros sover is not considered. T...

    work page 1983

  8. [8]

    REVIEW OF IMPLEMENTATION OF CEED USING AI TECHNIQUES A summary of the efforts made by researchers in CEED case using different AI techniques is given below: i. M. R. AlRashidi et al. [3]: They have done practical implementation of CEED with PSO successfully in 2006 considering almost all important constraints. It is a source reference for further research...

    work page 2006

Show all 27 references
  1. [9]

    SIMULATION RESULTS Combined economic emissio n dispatch using PSO and GA for 500 iterations has been implemented on MATLAB v8.2.0.701(R2013b) with Intel(R) Core(TM) i5 -2410M CPU @ 2.30 GHz 2.30 GHz on 6 generators of IEEE 30 bus system for load demands of 1500 and 2000 MW . T...

    2000

  2. [10]

    and Fig. 2. respectively. TABLE Simulation results of PSO and GA for IEEE 30 bus system with PD = 1500 MW and PD = 2000 MW IEEE 30 PD = 1500 MW PD = 2000 MW PSO GA PSO GA P1 195.79 224 256.32 256 P2 256.55 255 320.57 320 P3 381.25 367 541.95 576 P4 81.69 84 133.10 119 P5 381.8...

    2000

  3. [11]

    CONCLUSION The simulation results dictates that PSO converges faster to global optimal solution as compared to GA due to its simple two equations (velocity and position) mathematical model while GA has to deal with three layers of operators (selection, crossover and mutation) ...

  4. [12]

    Economic dispatch using particle swarm optimization: A review,

    Mahor, V. Prasad, and S. Rangnekar, “Economic dispatch using particle swarm optimization: A review,” Renewable and Sustainable Energy Reviews, vol. 13 , no. 8, pp. 2134–2141, 2009

    2009

  5. [13]

    Investigation on the impact of the penalty factors over solution of the dispatch optimization problem,

    S. Krishnamurthy and R. Tzoneva, “Investigation on the impact of the penalty factors over solution of the dispatch optimization problem,” in International Sci.Int.(Lahore),28(2),1401-1406,2016 ISSN 1013-5316; CODEN: SINTE 8 1406 March-April Conference on Industrial Technology ...

    2016

  6. [14]

    Emission - economic dispatch using a nove l constraint handling particle swarm optimization strategy,

    M. R. AlRashidi and M. E. El -Hawary, “Emission - economic dispatch using a nove l constraint handling particle swarm optimization strategy,” in Canadian Conference on Electrical and Computer Engineering (CCECE), Ottawa, Canada, May 2006, pp. 664–669

    2006

  7. [15]

    An analysis of economic load dispatch using different a lgorithms,

    K. Dasgupta and S. Banerjee, “An analysis of economic load dispatch using different a lgorithms,” in International Conference on Non Conventional Energy (ICONCE), Kalyani, India, Jan. 16 –17, 2014, pp. 216 – 219

    2014

  8. [16]

    Evolutionary programming for generation scheduling,

    D. P. Kothari and J. S. Dhillon, “Evolutionary programming for generation scheduling,” in Power System Optimization. New Delhi, India: Prentice-Hall of India Private Limited, 2006, ch. 7, pp. 501–532

    2006

  9. [17]

    Artificial bee colony optimization for combined economic load and emission dispatch,

    G. P. Dixit, H. M. Dubey, M. Pandit, and B. K. Panigrahi, “Artificial bee colony optimization for combined economic load and emission dispatch,” in International Conference on Sustain able Energy and Intelligent Systems (SEISCON) , Chennai, India, Jul. 20 - 22, 2011, pp. 340-345

    2011

  10. [18]

    Optimization of combined economic and emission dispatch problem - a comparative study,

    R. Bharathi, M. J. Kumar, D. Sunitha, and S. Premalatha, “Optimization of combined economic and emission dispatch problem - a comparative study,” in International Power Engineering Conference (IPEC) , Singapore, Dec. 3–6, 2007, pp. 134–139

    2007

  11. [19]

    Multi- objective environmental/economic dispatch using firefly technique,

    O. Abedinia , N. Amjady , and M. S. Naderi , “ Multi- objective environmental/economic dispatch using firefly technique,” in 11th International Conference on Environment and Electrica l Eng ineering (EEEIC) , Venice, Italy, May 18-25, 2012, pp. 461-466

    2012

  12. [20]

    Combined economic emission dispatch u sing SFLA,

    G. Chen, “Combined economic emission dispatch u sing SFLA,” in International Conference on Information Engineering and Computer Science (ICIECS ), Wuhan, China, Dec. 19-20, 2009, pp. 1-4

    2009

  13. [21]

    Comparison and application of e volutionary programming techniques to combined e conomic emission dispatch with line flow c onstraints,

    P. Venkatesh, R. Gnanadass, and N. P. Padhy , “Comparison and application of e volutionary programming techniques to combined e conomic emission dispatch with line flow c onstraints,” IEEE Transactions on Power Systems, vol. 18, no. 2, pp. 688– 697, 2003

    2003

  14. [22]

    Environmental/economic dispatch using genetic algorithm and simulated annealing,

    P. Erdogmus, A. Ozturk, and S. Duman, “Environmental/economic dispatch using genetic algorithm and simulated annealing,” in International Conference on Electric Power and Energy Conversion Systems (EPECS), Sharjah, Nov. 10–12, 2009, pp. 1–4

    2009

  15. [23]

    Combined economic and emission dispatch solution using gravitational search algorithm ,

    U. Guvenc, Y. Sonmez, S. Duman, and N. Yoru keren, “Combined economic and emission dispatch solution using gravitational search algorithm ,” Scientia Iranica , vol. 19, no. 6, pp. 1754–1762, 2012

    2012

  16. [24]

    Biogeography-based optimization for d ifferent economic load dispatch p roblems,

    A. Bhattacharya and P. K. Chattopadhyay, “Biogeography-based optimization for d ifferent economic load dispatch p roblems,” IEEE Transactions on Power Systems, vol. 25, no. 2, pp. 1064–1077, 2010

    2010

  17. [25]

    A g enetic algorithm based economic d ispatch (GAED) with environmental c onstraint optimisation,

    D. J. King, C. S. Ozveren, and Warsono, “A g enetic algorithm based economic d ispatch (GAED) with environmental c onstraint optimisation,” in Proceedings of 46th International Universities' Power Engineering Conference (UPEC) , Soest, Germany, Sep. 5 -8, 2011, pp. 1-6

    2011

  18. [26]

    A novel seeker optimization approach for s olving combined economic and emission dispatch,

    M. S. Syed, A. R. Syed, and Y. S. Rao, “A novel seeker optimization approach for s olving combined economic and emission dispatch,” in International Conference on Power, Energy and Control (ICPEC) , Dindigul, India, Feb. 6-8, 2013, pp. 371-376

    2013

  19. [27]

    Oppositional biogeography -based optimization for multi-objective economic emission load dispat ch,

    A. Bhattacharya and P. K. Chattopadhyay, “Oppositional biogeography -based optimization for multi-objective economic emission load dispat ch,” in Annual IEEE India Conference (INDICON) , Kolkata, India, Dec. 17-19, 2010, pp. 1-6

    2010

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.