Weighting
Z-AHP: Z-Number Analytic Hierarchy Process
Nuriyev, M. · 2020
Overview
Z-AHP is AHP with two channels per judgment: 'how much more important' (restriction) AND 'how sure I am' (reliability). The reliability TFN is collapsed to a single number α via centroid; then it stretches/shrinks the restriction TFN by √α. After all cells are Z→TFN, Buckley's geometric mean (instead of Saaty's eigenvector) computes row priorities; ratio of each row's geometric mean to the sum gives fuzzy weights; COA defuzzifies; normalize. Hierarchical structure is handled by Saaty's level-by-level composition.
Strengths
- •Hierarchical decomposition enables complex problems (Nuriyev demonstrates goal → 4 main criteria → 9 sub-criteria)
- •Captures both vagueness (TFN) and reliability (second TFN): richer than Buckley fuzzy AHP
- •Buckley's geometric mean is well-established and avoids iterative eigenvector computation
- •Group decision-making built-in via per-cell arithmetic averaging (Eq.25)
Limitations
- •Full n×n pairwise scales poorly: n(n-1)/2 unique comparisons + K experts × levels
- •COA defuzzification ignores TFN shape information (vs GMIR which weights m by 4)
- •Reciprocity assumption may not hold when experts disagree systematically
- •Saaty's classical consistency check (CR) requires post-defuzzification crisp matrix; not natively Z-number aware
Method assistant
Grounded explanations: it explains the method, it does not compute.
Edge cases
- •When all cells are (1,1,1)+(VH,VH), weights are uniform 1/n
- •When K=1, group aggregation is identity: pipeline unchanged
- •Two-level hierarchy: compute weights per level, then composite = main × sub (Nuriyev Table 12)
- •Reciprocity slightly violated: auto-correct or warn
- •n very large (>12): pairwise burden: switch to Z-BWM (2n-3 comparisons)
Common pitfalls
- •Don't apply Buckley geometric mean to raw Z-numbers: do Z→TFN first
- •Don't confuse COA (Eq.30, simple average) with GMIR ((l+4m+u)/6 used in Z-BWM)
- •Don't forget reciprocity Eq.(5) on the lower triangle
- •Restriction TFN and reliability TFN are SEPARATE: never average them
Worked example
- 1.Step 1
- 2.Step 2
- 3.Step 3
- 4.Step 4
- 5.Step 5
- 6.Step 6
- 7.Step 7
- 8.Step 8
- 9.Step 9
How to cite
Nuriyev, M. (2020). Z-numbers Based Hybrid MCDM Approach for Energy Resources Ranking and Selection. International Journal of Energy Economics and Policy. https://doi.org/10.32479/ijeep.9950