The manuscript is improved quite a lot, but still include new typo's and new careless mistakes as suggested below. Please don't count on reviewers as a proof-reader. Usually such a manuscript will be rejected without pointing out what/which are specifically wrong.

p. 1 keyword chaotic --- not appropriate as a keyward 
p.43 line 136 (1). Generate => (1) Generate
p.43 line 143 (2). For => (2) For
p.44 line 170 X +Y => X + Y
p.44 line 180 .Suppose => . Suppose
p.45 line 211 between the i-th node and the j-th node => from the jth node to the i-th node
p.45 line 213 as follow => as follows
p.45 line 217 line => linear
p.45 line 221, 223, 224, 225 226 (1). Select => (1) Select etc.
p.47 line 254, 266 "k represent the parameters of neural network for PSO to determine" 
                   => still difficult to understand.
p.48 line 307 then fed => fed 
p.48 line 310 nerve => neural
p.48 line 314 in the literature => not necessary, or if wanted add the [paper]
p.49 line 323 n th => n-th & m th => m-th
p.49 line 328 training simples => training samples
p.49 line 329 simple => sample
p.49 line 331 as the number in the swarm => ?
p.50 line 350 353 in respect with => with respect to
p.52 line 440 Second, => Where is First?
p.52 line 443 etc => etc.
p.54 line 457-458 temperature( => temperature ( etc.
p.54 line 460 ppbv
p.54 line 467 as followed => are as follows
p.54 line 474 1 => one
p.55 line 503 2 => two
p.55 line 512 maxi - mum => muximum
p.55 line 520 Be consitent with the note in the Table 1
p.56 line 531 ppbv
p.58 line 582 not all days in the same trend => not all days were in the same trend.
p.58 line 583 litter -> little
p.58 line 584 will reduce increase => ?
p.58 line 587 indent
p.58 line 599 require => requires

References should be consistent from one item to the next, e.g., all items should end with '.'





As already mentioned, the manuscript is improved quite a lot, but still includes new typo's and new careless mistakes. The explanation in the first half of the text is quite good while the last half is still not easy to understand. The last half is still more like a draft before proof-reading. Newly added parts of the text  with red letters also include mistakes/redundancies/clearness. 

Although the paper could be improved by repeating review and re-edit procedure,  it would be less likely to be done just one more additional time. So, the reviewer recommends the authors to dismiss it now and take a time to refine the paper.