SSDEVOP which allows the size of each simplex to expand or contract can, in certain circumstances, allow them to more readily respond to local contour environments (Nelder and Mead, 1965), though this would possibly conflict with the philosophy of targeting deliberately modest set-point changes, to minimise any risk of process upsets. Furthermore, the data generated from SSDEVOP is not as amenable to modelling and delineation of response surfaces compared to CCRDs. In real experimentation, where there is inherent ‘noise’ in the collected experimental data (and with this, uncertainty in the location of the process contours), SSDEVOP simplexes may not stabilise exactly in regions of optimised performance (Paredes and Ágreda, 2020). Figure 59 shows the same surface response as previously shown in Figure 58, but with experimental ‘noise’ incorporated into the data (defined as SD = 3.5 µm). Figure 59—Grind P80 contours (using data in Chauhan (2014)) with experimental ‘noise’ and the resulting SSDEVOP simplex path. The path of the simplex stabilised in a region where P80 ~ 115 µm; while this was much better than the starting conditions where P80 ~ 160 µm, it was not quite as good as the conditions corresponding to P80 ~ 90 µm which was identified by the error-free simplex shown previously in Figure 58. This is a shortcoming that would be expected in real trials, and which would be exacerbated by larger degrees of measurement uncertainty in the process response. Nevertheless, the process response in Figure 59 was still clearly better at the conclusion of the SSDEVOP, despite the slightly compromised ability of the simplexes to identify the true response minimum. The main benefit of SSDEVOP is that it allows for the seeking out of such improvements via conservative, incremental changes, when the priority is to eliminate the risk of process disruption. It is also suitable for operator involvement in the process, which improves the commitment of operators to the final outcome. CONCLUSIONS Critical decision-making should be driven by rigorous data analysis. The problem for metallurgists is that mineral processing data is complex, owing to the fact that production environments are P80 ~ 90 µm P80 ~ 115 µm P80 ~ 160 µm
RkJQdWJsaXNoZXIy MTM0Mzk2