Universidad Politécnica de Madrid (UPM), 2024

SIMLIN 2.0

Fruit classification line simulator related to mechanical bruises events

Simulador del comportamiento de una línea de clasificación de fruta en relación con la aparición de daños mecánicos

SIMILIN 2.0 is a simulator software of real fruit classification lines predicting mechanical bruises by probabilistic simulation techniques. Classification line simulator shows some specific aspects:

Bruises prediction logistic model: there is the possibility of setting new models using user specific data. An influential data control is performed, as well as the capacity of the prediction model established.
Characterisation statistical of individual fruit loads: this aspect is needed to realistically reproduce the potential damage to the product.
Flexibility in the design and modification of fruit handling lines: from some basic rules, the user is able to build up on the screen a new line, by choosing the elements from a data base.
Impacts data base: by an extensive use of electronic fruits in actual handling lines, we were able to build a data base that characterizes elements and transfers points.
Simulation and validation results: each model must be validated in advance by fruit sampling before and after its line handling.
Line improvements simulation: from a validated model is possible to check the effect of different improvements on classification lines, i.e. replacement of aggressive elements and transfer elements, on the level of damaged fruit.

Innovative aspect

Introduction of decision aid tools in the agricultural sector. In most installations fruit handling and dressing capacity reaches up to 15 t/h, therefore it is not practical to modify in situ without a previous knowledge of its feasibility. Simulation, as an alternative method to experimentation, is specially indicated in these situations. Simulation is based on the replacement of aleatory data from the real world by a sequence of events generated by a computer. When these events are really aleatory, simulation results are not distinguishable from reality.

Main advantages

SIMLIN 2.0 will permit the user to predict the product losses he will have, depending on line conditions and fruit species, variety and state. This kind of tool helps the user to know the actual aggressiveness degree of his machinery, in relation to the product and to his improvement needs.

Authors

  • Margarita Ruiz Altisent, Physical Properties Laboratory of Rural Engineering Department at Technical University of Madrid. mruiz@iru.etsia.upm.es
  • Pilar Barreiro Elorza, Physical Properties Laboratory of Rural Engineering Department at Technical University of Madrid. pbarreiro@iru.etsia.upm.es
  • Concha Bielza Lozoya, Artificial Intelligence Department at Technical University of Madrid. mcbielza@fi.upm.es
  • Francisco García García
  • Francisco Javier García Ramos, Universidad de Cádiz. juanantonio.garcia@uca.es
  • Rubén Heradio Gil, Software Engineering Department at Universidad Nacional de Educación a Distancia. rheradio@issi.uned.es
  • Óscar Pacios Rodríguez
  • Jacinto Martín Jiménez, Mathematics Department at Universidad de Extremadura. jmartin@unex.es
  • Sixto Ríos Insua, Artificial Intelligence Department at Technical University of Madrid. srios@fi.upm.es

AI Website Creator