Academia.eduAcademia.edu

Figure 3 – uploaded by Harley R Myler

See full PDFdownloadDownload figure
Fig. 2. Possible scenarios: (a) reverse auction scenario, and (b) auction scenario. According to Fig. 3, at the first decision node, a’ makes an offer regarding the explained process, and then at the second step, and a? makes an acceptance or rejection decision, using the proposed reasoning mechanisms.

Figure 2 Possible scenarios: (a) reverse auction scenario, and (b) auction scenario. According to Fig. 3, at the first decision node, a’ makes an offer regarding the explained process, and then at the second step, and a? makes an acceptance or rejection decision, using the proposed reasoning mechanisms.

Related Figures (14)

Fig. 1. CPM and UPM. 7; and T, are in agent’s CPM and UPM, respectively. T, is beyond agent’s uncertainty boundary. B. Reasoning Mechanism In this subsection, reasoning mechanisms for both hunters and gatherers are developed to establish their behavior during a mission that determines the way that they communicate, negotiate, and cooperate. Since fundamentals of reasoning mechanisms are similar for both type of agents, for the sake of brevity, we consider a general agent defined as a’? with Je {h,g}, where
Fig. 1. CPM and UPM. 7; and T, are in agent’s CPM and UPM, respectively. T, is beyond agent’s uncertainty boundary. B. Reasoning Mechanism In this subsection, reasoning mechanisms for both hunters and gatherers are developed to establish their behavior during a mission that determines the way that they communicate, negotiate, and cooperate. Since fundamentals of reasoning mechanisms are similar for both type of agents, for the sake of brevity, we consider a general agent defined as a’? with Je {h,g}, where
Fig. 4. An example explaining the necessity of multi-tasking capability of gatherer agents.
Fig. 4. An example explaining the necessity of multi-tasking capability of gatherer agents.
Fig. 5. An example of multitask-planning by a gatherer agent Fig. 5 illustrates an example in which a fills out its action plan where q,,,,=5. All sequences happen before accomplishing v,. In sequence |, a* has already v, in its action plan and chooses 7, to negotiate. In sequence 2, a* has reached an agreement for accomplishing 7;., and adds it to its action plan as v,. Moreover, there are two candidates and it chooses T.., to negotiate because T,, is not feasible and fails to satisfy the condition mentioned in step 2. In sequence 3, a’ has reached an agreement for accomplishing T,., and adds it to its action plan as v,. In addition, there are two candidate tasks namely T,, and T,;. T,; cannot be verified in step 3, though it’s feasible itself and passes step 2. Therefore, a’ chooses T.., for negotiation. In sequence 4, a’ has reached an agreement for accomplishing T,,, and adds it to its action plan as v,. Moreover, it chooses T,,, to negotiate because it passes all 3 steps. Although choosing T,., causes change in the path to 7, , but it does not bring a‘ into state 3 for T,. it passes all 3 steps. Although choosing T,., causes change in the path to 7, , but it does not bring a’ into state 3 for T,.
Fig. 5. An example of multitask-planning by a gatherer agent Fig. 5 illustrates an example in which a fills out its action plan where q,,,,=5. All sequences happen before accomplishing v,. In sequence |, a* has already v, in its action plan and chooses 7, to negotiate. In sequence 2, a* has reached an agreement for accomplishing 7;., and adds it to its action plan as v,. Moreover, there are two candidates and it chooses T.., to negotiate because T,, is not feasible and fails to satisfy the condition mentioned in step 2. In sequence 3, a’ has reached an agreement for accomplishing T,., and adds it to its action plan as v,. In addition, there are two candidate tasks namely T,, and T,;. T,; cannot be verified in step 3, though it’s feasible itself and passes step 2. Therefore, a’ chooses T.., for negotiation. In sequence 4, a’ has reached an agreement for accomplishing T,,, and adds it to its action plan as v,. Moreover, it chooses T,,, to negotiate because it passes all 3 steps. Although choosing T,., causes change in the path to 7, , but it does not bring a‘ into state 3 for T,. it passes all 3 steps. Although choosing T,., causes change in the path to 7, , but it does not bring a’ into state 3 for T,.
Algorithm 3 decision-making outer-loop algorithm of a! Algorithm 4 decision-making outer-loop algorithm of a gatherer agent
Algorithm 3 decision-making outer-loop algorithm of a! Algorithm 4 decision-making outer-loop algorithm of a gatherer agent
Fig. 6. investigating the fairness of the HGMP for: (a) hunter agents, and (b) gatherer agents. Statistically, for each type of agents, there is no significant differenc in agents’ effectiveness that demonstrates the fairness of the HGMP.
Fig. 6. investigating the fairness of the HGMP for: (a) hunter agents, and (b) gatherer agents. Statistically, for each type of agents, there is no significant differenc in agents’ effectiveness that demonstrates the fairness of the HGMP.
We ran the algorithms for all values of a@, and £, that are multiples of 0.025 such that O<a@,<0.5 and 0<£, <0.5, while @, = 8, =0.35. 7, has been calculated for each set of values for a, and f, as illustrated in Fig. 7. According to the results, 7, is vanishingly small when a, + 6, <0.2 that means agents cannot reach an agreement for completing the detected tasks. Further, 7, reaches its maximum when 0.2<a,+,<0.5. Next, 7, falls gradually when a, +, >0.5 because each gatherer’s CMP and UPM are large so that the agent does not fall into the State 3 and easily reach any agreement. As a result, each gatherer accomplishes a significant number of tasks inefficiently which reduces 7, . Wis ghee) eee eee: ee ee t@QIboes cee eee Be ee ey a I ee “Perec ef ase Me” 2 Py SE ee ey Fe ee Ce oe « ren Ses Se ae oo ae Mr: Me. ere « er | C,, as follows:
We ran the algorithms for all values of a@, and £, that are multiples of 0.025 such that O<a@,<0.5 and 0<£, <0.5, while @, = 8, =0.35. 7, has been calculated for each set of values for a, and f, as illustrated in Fig. 7. According to the results, 7, is vanishingly small when a, + 6, <0.2 that means agents cannot reach an agreement for completing the detected tasks. Further, 7, reaches its maximum when 0.2<a,+,<0.5. Next, 7, falls gradually when a, +, >0.5 because each gatherer’s CMP and UPM are large so that the agent does not fall into the State 3 and easily reach any agreement. As a result, each gatherer accomplishes a significant number of tasks inefficiently which reduces 7, . Wis ghee) eee eee: ee ee t@QIboes cee eee Be ee ey a I ee “Perec ef ase Me” 2 Py SE ee ey Fe ee Ce oe « ren Ses Se ae oo ae Mr: Me. ere « er | C,, as follows:
Fig. 7. Investigating the effect of gatherers’ CPM and UPM on the HGMP’s effectiveness 77, : (a) the contour plot of the results using a polynomial curve-fitting, and (b) polynomial curve-fitting.
Fig. 7. Investigating the effect of gatherers’ CPM and UPM on the HGMP’s effectiveness 77, : (a) the contour plot of the results using a polynomial curve-fitting, and (b) polynomial curve-fitting.
Fig. 8. The effect of hunters’ CPM and UPM on the HGMP’s effectiveness 7, : (a) the contour plot of the results using a polynomial curve-fitting, and (b) polynomial curve-fitting.
Fig. 8. The effect of hunters’ CPM and UPM on the HGMP’s effectiveness 7, : (a) the contour plot of the results using a polynomial curve-fitting, and (b) polynomial curve-fitting.
Fig. 9. Investigating the effect of g_, on 77, . The results show a significant increase in 7, when we increase g_., from 1 to 10
Fig. 9. Investigating the effect of g_, on 77, . The results show a significant increase in 7, when we increase g_., from 1 to 10
Fig. 10. fairness of the HGMP for gatherer agents for each value of q,,,, . The results demonstrate that the workload is still distributed equally on both gatherers when g__ increases.
Fig. 10. fairness of the HGMP for gatherer agents for each value of q,,,, . The results demonstrate that the workload is still distributed equally on both gatherers when g__ increases.
Fig. 11. the convergence of the HGMP’s total effectiveness for Gy =1. Gay =4> aNd Pony =10, during T,,. = 1000 iterations D. Functionality Validation of the HGMP by a Comparison
Fig. 11. the convergence of the HGMP’s total effectiveness for Gy =1. Gay =4> aNd Pony =10, during T,,. = 1000 iterations D. Functionality Validation of the HGMP by a Comparison
Fig. 12. Comparing the HGMP with an alternative model where each agent does both exploration and completion together. and maneuvering. Therefore, we first plotted the total effectiveness of the HGMP with respect to 9, / p, Which ranges fron Pr K Py» ie. P| Pe =0,to p, * P,. Le. Pr / Pe =1. Secondly, we ran the explained alternative approach to be able to judge the HGMP’s functionality. Since in this approach there is no hunter and gatherer scheme, we only have one type of agent and the obtained total effectiveness is dependent from this ratio p, / p, - By this comparison we basically wanted to answer this questior that: is the HGMP profitable compared to the alternative method? Fig. 12 shows the results of the implemented simulations fo: that purpose, as explained above. Thus, the answer is it depends to the ratio ,/, and that’s why we ended up in having a criterior
Fig. 12. Comparing the HGMP with an alternative model where each agent does both exploration and completion together. and maneuvering. Therefore, we first plotted the total effectiveness of the HGMP with respect to 9, / p, Which ranges fron Pr K Py» ie. P| Pe =0,to p, * P,. Le. Pr / Pe =1. Secondly, we ran the explained alternative approach to be able to judge the HGMP’s functionality. Since in this approach there is no hunter and gatherer scheme, we only have one type of agent and the obtained total effectiveness is dependent from this ratio p, / p, - By this comparison we basically wanted to answer this questior that: is the HGMP profitable compared to the alternative method? Fig. 12 shows the results of the implemented simulations fo: that purpose, as explained above. Thus, the answer is it depends to the ratio ,/, and that’s why we ended up in having a criterior
Related topics:
Applied MathematicsNumerical Analysis and Computational Mathematics
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved