A greedy approach to improve pesticide application for precision agriculture using model predictive control
Pests may lead to low crop productivity and profitability. Pesticides are commonly used to protect crops from pests. However, too much pesticide is not only associated with harmful effects to the environment but may also lead to sub-optimal pest management. The existing works focus on the vehicle ro...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Published: |
Elsevier B.V.
2021
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/96489/ http://dx.doi.org/10.1016/j.compag.2021.105984 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Pests may lead to low crop productivity and profitability. Pesticides are commonly used to protect crops from pests. However, too much pesticide is not only associated with harmful effects to the environment but may also lead to sub-optimal pest management. The existing works focus on the vehicle routing problem for pesticide management without giving due consideration to finding the optimal time, amount, and area for pesticide application. To solve this issue, this paper takes an active stance and introduces demand management for pesticide using an active mass-spring suspension system. Moreover, using a controller based on model predictive control that uses the active demand management model, this paper efficiently solves the problem of finding the right time, amount and place for pesticide application in an agricultural field. A greedy algorithm is then proposed to solve the vehicle routing problem after identifying the optimal time, and place for pesticide application. The proposed solution minimizes the risk of pest infestation by considering pest risk prediction models. The simulation results show that the proposed technique can maximize the protection for crops against pests. Moreover, a performance analysis of the proposed technique shows that it has significantly lower computational complexity and can converge to the optimal solution at least 78% faster than existing techniques. |
|---|