Finding the optimum solution for dispatching in concrete delivery is computationally intractable because it is a NP-hard （non-deterministic polynomial-time hard） problem. Heuristic methods are required to obtain sat...Finding the optimum solution for dispatching in concrete delivery is computationally intractable because it is a NP-hard （non-deterministic polynomial-time hard） problem. Heuristic methods are required to obtain satisfactory solutions. Inefficiencies in mathematical modeling still make concrete dispatching difficult to solve. In reality, complex dispatching systems are mostly handled by human experts, who are able to manage the assigned tasks well. However, the high dependency on human expertise is a considerable challenge for RMC （ready mixed concrete） companies. In this paper, a logical reconstruction of an expert＇s decision making is achieved by two machine learning techniques： decision tree and rule induction. This paper focuses on the expert dispatcher＇s prioritization of customer orders. The proposed method has been tested on a simulation model consisting of a batch plant and three customers per day. The scenarios generated by the simulation model were given to a dispatch manager who was asked to prioritize the customers in each day. The scenarios and the decisions were then input to the machine learning programs, which created generalizations of the expert＇s decisions. Both decision trees and rules approach 80% accuracy in reproducing the human performance.展开更多
文摘Finding the optimum solution for dispatching in concrete delivery is computationally intractable because it is a NP-hard （non-deterministic polynomial-time hard） problem. Heuristic methods are required to obtain satisfactory solutions. Inefficiencies in mathematical modeling still make concrete dispatching difficult to solve. In reality, complex dispatching systems are mostly handled by human experts, who are able to manage the assigned tasks well. However, the high dependency on human expertise is a considerable challenge for RMC （ready mixed concrete） companies. In this paper, a logical reconstruction of an expert＇s decision making is achieved by two machine learning techniques： decision tree and rule induction. This paper focuses on the expert dispatcher＇s prioritization of customer orders. The proposed method has been tested on a simulation model consisting of a batch plant and three customers per day. The scenarios generated by the simulation model were given to a dispatch manager who was asked to prioritize the customers in each day. The scenarios and the decisions were then input to the machine learning programs, which created generalizations of the expert＇s decisions. Both decision trees and rules approach 80% accuracy in reproducing the human performance.
