Table 2. Comparison results with other excellent algorithms on rPSP index.
PSP is the inverse of PSP. Therefore, the smaller metric is the better the performance. The algorithm runs on 13 test functions and independently runs 30 times, which gets the average and standard deviation. The data in the table are the average of the 30 times. The experiment tested the data obtained by all algorithms for significant differences using the Wilson test. The “+” and “-” in the table indicate statistically superior or inferior to the comparison algorithm, respectively, while “=” indicates statistically similar results. In Table 1, Mean represents the average of 30 results. Std represents the standard deviation on the test problem.
| MO_Ring _PSO_SCD | DN-NSGAII | Omni-optimizer | TriMOEA-TA&R | MMODE_ICD | MMODE_ES | ||
|---|---|---|---|---|---|---|---|
| MMF1 | Mean | 0.0488+ | 0.0969+ | 0.0975+ | 0.0735+ | 0.0493+ | 0.0417 |
| ± Std | 0.0019 | 0.0145 | 0.013 | 0.0108 | 0.0030 | 0.0013 | |
| MMF2 | Mean | 0.0465+ | 0.1742+ | 0.1643+ | 0.0931+ | 0.0247+ | 0.0075 |
| ± Std | 0.0147 | 0.1368 | 0.1305 | 0.0574 | 0.0054 | 0.0010 | |
| MMF3 | Mean | 0.0335+ | 0.1172+ | 0.1351+ | 0.0871+ | 0.0209+ | 0.0068 |
| ± Std | 0.0099 | 0.0759 | 0.0994 | 0.0275 | 0.0037 | 0.0007 | |
| MMF4 | Mean | 0.0273+ | 0.0771+ | 0.0840+ | 0.1537+ | 0.0257+ | 0.0217 |
| ± Std | 0.0020 | 0.0137 | 0.0238 | 0.2262 | 0.0031 | 0.0010 | |
| MMF5 | Mean | 0.0869+ | 0.1773+ | 0.1789+ | 0.1132+ | 0.0853+ | 0.0711 |
| ± Std | 0.0060 | 0.0217 | 0.0245 | 0.0126 | 0.0039 | 0.0035 | |
| MMF6 | Mean | 0.0733+ | 0.1427+ | 0.1523+ | 0.0958+ | 0.0713+ | 0.0637 |
| ± Std | 0.0043 | 0.0150 | 0.0185 | 0.0123 | 0.0043 | 0.0004 | |
| MMF7 | Mean | 0.0267+ | 0.0553+ | 0.0511+ | 0.0672+ | 0.0263+ | 0.0206 |
| ± Std | 0.0015 | 0.0151 | 0.0127 | 0.0520 | 0.0046 | 0.0023 | |
| MMF8 | Mean | 0.0678+ | 0.2799+ | 0.3149+ | 0.3974+ | 0.1303+ | 0.0504 |
| ± Std | 0.0042 | 0.0911 | 0.1326 | 0.1572 | 0.0352 | 0.0048 | |
| MMF9 | Mean | 0.0082+ | 0.0219+ | 0.0316+ | 0.0031- | 0.0047- | 0.0061 |
| ± Std | 0.0008 | 0.0078 | 0.0269 | 0.0001 | 0.0003 | 0.0004 | |
| MMF10 | Mean | 0.1708- | 1.3420+ | 3.1260+ | 0.2014= | 0.2011= | 0.2035 |
| ± Std | 0.0231 | 2.3280 | 3.3420 | 0.0001 | 0.0010 | 0.0024 | |
| SYM-PART simple | Mean | 0.1776+ | 5.4511+ | 6.4608+ | 0.0210- | 0.0427- | 0.0441 |
| ± Std | 0.0226 | 2.576 | 3.0139 | 0.0021 | 0.0055 | 0.0042 | |
| SYM-PART rotated | Mean | 0.2784- | 4.4542+ | 5.4757+ | 2.0975+ | 0.0892+ | 0.0536 |
| ± Std | 0.2500 | 1.4104 | 3.4185 | 1.4432 | 0.0153 | 0.0033 | |
| Omni-test | Mean | 0.4279+ | 1.5563+ | 1.7939+ | 0.7547+ | 0.0512+ | 0.0430 |
| ± Std | 0.0954 | 0.2878 | 0.6207 | 0.2166 | 0.0036 | 0.0020 | |
| +/=/- | 11/0/2 | 13/0/0 | 13/0/0 | 10/1/2 | 10/1/2 |
Notes.
rPSP results and rHV results obtained by these comparison algorithms are underlined and shown in bold when the rPSP results and rHV results obtained by MMODE_ES are better than the comparison algorithms under the same test function.